Академический Документы
Профессиональный Документы
Культура Документы
John W. Sweetenham
Huntsman Cancer Institute, University of Utah, Salt Lake City, UT, USA
Michael E. Williams
Division of Hematology/Oncology, University of Virginia School of Medicine, Charlottesville, VA, USA
University Printing House, Cambridge CB2 8BS, United Kingdom
Published in the United States of America by Cambridge University Press, New York
Cambridge University Press is part of the University of Cambridge.
It furthers the University’s mission by disseminating knowledge in the pursuit of
education, learning, and research at the highest international levels of excellence.
www.cambridge.org
Information on this title: www.cambridge.org/9781107010598
© Cambridge University Press 2014
This publication is in copyright. Subject to statutory exception
and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without the written
permission of Cambridge University Press.
First published 2014
Printed and bound in Spain by Graphos SA, Arte sobre papel
A catalog record for this publication is available from the British Library
ISBN 978-1-107-01059-8 Hardback
Cambridge University Press has no responsibility for the persistence or accuracy of
URLs for external or third-party internet websites referred to in this publication,
and does not guarantee that any content on such websites is, or will remain,
accurate or appropriate.
.................................................................................................
Every effort has been made in preparing this book to provide accurate and
up-to-date information which is in accord with accepted standards and practice
at the time of publication. Although case histories are drawn from actual cases,
every effort has been made to disguise the identities of the individuals involved.
Nevertheless, the authors, editors, and publishers can make no warranties that the
information contained herein is totally free from error, not least because clinical
standards are constantly changing through research and regulation. The authors,
editors, and publishers therefore disclaim all liability for direct or consequential
damages resulting from the use of material contained in this book. Readers
are strongly advised to pay careful attention to information provided by the
manufacturer of any drugs or equipment that they plan to use.
Contents
Preface to second edition page vii
Preface to first edition ix
List of contributors x
v
Preface to second edition
Since the first edition we have seen a number of impor- and mantle cell lymphoma, and idelalisib in indolent
tant changes in the diagnosis, staging and therapy for B-cell lymphoproliferative disorders. These specifi-
many lymphoma subtypes. We are beginning to cally targeted therapies give us hope that, in the very
observe how molecular and cytogenetic characteristics near future, the management of lymphoma will be
can profoundly influence prognosis and that such tools both more effective and less toxic.
should now be as much a part of our diagnostic arma- In this new edition we have also endeavored to
mentarium as histology and immunophenotyping. rectify some of the omissions in the previous edition;
In the near future we expect that “next generation” in particular, we have now additional chapters on
sequencing will also have a considerable impact not Lymphoplasmacytoid Lymphoma and Atypical
only on our understanding of the pathogenesis of Lymphoproliferative disorders that make the book
lymphoma, but also on our selection of therapy. We more comprehensive.
can also envisage how such techniques will themselves Once again we have incorporated biology,
also lead to less toxic and more targeted therapies. pathogenesis, histopathology, and therapy into each
We have seen the steady increase in the incorpo- disease based chapter, we hope the “joins” do not
ration of PET and PET CT into staging, risk-adapted show too much and are most grateful indeed to
therapy, and reassessment such that the presence of a Drs. Ott, Rosenwald, and Wotherspoon for allowing
PET/CT scanner and radiologic expertise should now their contributions on molecular biology and
be an essential in every major Lymphoma centre. The histopathology, respectively, to be separated and
promise and precautions for this important tool are distributed as before.
summarized within this edition. Our thanks are due too to all the authors and to
In terms of therapy, the steady progress in both the colleagues at CUP for their hard work and forbearance
use of monoclonal antibodies and new treatment regi- in the preparation of this new edition.
mens is reflected in the updated chapters. Lymphoma
specialists are also beginning to see the incorporation Robert Marcus
of agents that block intra-cellular signaling pathways John Sweetenham
into clinical practice, notably ibrutinib in CLL/SLL Michael Williams
vii
Preface to first edition
Why publish a book on lymphoma in 2007? Surely accept that their contributions on histopathology and
there are sufficient reviews, meetings and published molecular cytogenetics would be divided and distrib-
educational symposia to make such a work redundant. uted among the relevant chapters. The editors are most
Furthermore, doesn’t instant access to online informa- grateful for their support.
tion make any work in print out of date before it Each chapter in followed by a select bibliography
appears? rather than an exhaustive list of references. We feel
The editors and authors of this work think not. We that these date very quickly, take up disproportionate
firmly believe that there is still a place for a clear amounts of space and are better found by internet
summary of the diagnosis, staging and therapy of searches or perusal of current journals.
lymphoma in a single volume that reflect the advances This book is intended for senior trainees and
in these areas which have taken place over the past five fellows in hematology and oncology, together with
years. The problem with the plethora of information more experienced practitioners who regularly treat
now available is that it is rarely set in a framework of these disorders. It is not intended for those who may
understanding of the major challenges which still feel they could have written the chapters themselves.
face those involved with the diagnosis and therapy of It is also not a book where the reader will find detailed
non-Hodgkin’s and Hodgkin’s lymphomas. We, and descriptions of rarities seen once in a professional
our patients, are confronted with many facts but lifetime.
little judgment. This work is our modest attempt to The appearance of this volume comes at an oppor-
rectify this. tune time: we have seen over the past five years a
Accordingly the layout of the book should enable profound understanding of the pathology of lym-
the reader to gain an understanding of patterns of phoma, the use of increasingly sophisticated imaging
disease, methods of staging, principles of new techniques, and the integration of monoclonal anti-
approaches to therapy, and interpretation of clinical bodies into standard therapy for NHL. Our hope is
trials and prognostic markers by reading the first part that these radical changes in the way we diagnose and
of the book. In the second part the reader will find treat lymphoma have been reflected in the book and
separate succinct yet comprehensive reviews of the will stand the reader in good stead even when newer
individual disease entities which make up the spectrum data become available.
of diseases comprising the lymphomas. Here we have The editors express their sincere and heartfelt
integrated pathology, molecular biology and therapy thanks to all the authors, who have, in the main,
for each subtype into a single chapter; the reader will responded promptly to our comments and recommen-
not need to flick backwards and forwards to gain dations, and to all those at Cambridge University
a comprehensive understanding of, say, follicular or Press who have helped with this project: Richard
diffuse large B-cell lymphoma. Such an integration Barling, who set the wheels of this vehicle in motion,
has posed significant editorial challenges, and has and especially Betty Fulford and Deborah Russell, who
been made possible by a willingness on the part of kept it moving to its final destination whenever it
Dr Wotherspoon, Dr Rosenwald and co-workers to threatened to stall.
ix
Contributors
x
List of contributors
xi
Chapter
1
Epidemiology
Eve Roman and Alexandra G. Smith
Lymphoma, Second Edition, ed. Robert Marcus, John W. Sweetenham, and Michael E. Williams. Published by 1
Cambridge University Press. © Cambridge University Press 2014.
Chapter 1: Epidemiology
Gathering accurate information about disease bur- than half of all NHL and HL diagnoses occurred in
dens and patterns is central to any successful cancer the three regions with the highest rates – Northern
information strategy. However, whilst cancer registra- America, Europe, and Oceania (Figure 1.1) – largely
tion has a long history in many countries, particularly reflecting the underlying world population distribution.
those in the more affluent regions of the world, nearly The most striking disparity is seen for Asia, which has
80% of the world’s population is not covered by such the lowest rates but more than a third of all of the
systems (www.globocan.iarc.fr/). Furthermore, in estimated worldwide cases. To some extent these
some of these, even enumerating the population at broad regional differences may well reflect, at least in
risk (denominator) through census counts and vital part, underlying lymphoma subtype variations – with a
registration is challenging. Even so, with a view to relatively high proportion of mature T-/natural killer
characterizing the global burden of disease, the cell neoplasms being reported for several Asian popula-
WHO’s International Agency for Research on Cancer tions. These issues are discussed in more depth in the
(IARC) routinely uses the available data to estimate later section on etiology.
worldwide cancer incidence and mortality levels.
Misdiagnosis and underenumeration are recog-
nized as particularly problematic for hematological can- Age and sex
cers. The acute and rapidly fatal presentation of some That lymphoid malignancies are, overall, generally
patients leads to underenumeration in those countries more common in men than in women in all areas of
with less well-developed health service infrastructures the world is evident from Figure 1.1, where for both
and the intermittent and non-specific nature of symp- NHL and HL the age-standardized rates and numbers
toms associated with others poses problems even in are consistently higher for males than for females.
countries with well-developed health service systems Interestingly, these gender differences appear to be
and cancer registration processes. In addition, slightly more pronounced in less developed regions of
population-based data continue to be reported in the the world (http://globocan.iarc.fr/) – the sex rate ratios
broad anatomical-based categories of non-Hodgkin (M:F) for HL, for example, range from 1.8 and 1.6,
lymphoma (NHL), Hodgkin lymphoma (HL), mye- respectively, in Africa and Asia through to 1.2 and 1.1,
loma, and leukemia since, for reasons outlined above, respectively, in North America and Europe. Whether or
the laboratory data required to classify hematological not these gradations reflect genuine underlying inci-
cancers appropriately are hard for cancer registries to dence differences, either generally or within particular
access in a timely and systematic fashion. subtypes, or are in fact caused by enumeration biases
cannot be investigated further from the available cancer
registration data.
Geographic variation The NHL and HL age-specific incidence patterns
Of the 12.68 million new cancers estimated to have seen in more economically developed regions of the
occurred around the world in 2008, 6.64 were in men world are all broadly similar to the pattern shown in
and 6.04 in women. Combined, hematological malig- Figure 1.2, which presents cancer registration data
nancies (lymphomas, leukemias, and myelomas) com- from the UK. For NHL, the age-specific male and
prised 7.5% of the estimated cancers in males and 6.4% female rates increase with increasing age, the diver-
in females, with lymphomas accounting for around gence between the male and female rates becoming
half of all newly diagnosed hematological neoplasms progressively more marked as age increases, but,
in both men and women. despite this, more women than men are diagnosed
Figure 1.1 shows the estimated global regional rates over the age of 80 years. This apparent discrepancy
and numbers of cases of NHL and HL separately for reflects the fact that the UK, like other economically
men and women (www.globocan.iarc.fr/). In general, developed regions of the world, has an aging popula-
estimated lymphoma rates are higher in more econom- tion structure within which more women than men
ically developed regions of the world – North America, survive to reach old age.
Europe, and Australasia – and lower in less affluent The relationship with age and sex is quite different
regions. This relationship with affluence is seen for for NHL and HL. In more affluent regions of the world
both NHL and HL, with one or two interesting pockets HL tends to have a bimodal age distribution that is
2 such as the consistently low overall rates reported characterized by early age peak, within which females
for Japan (www.globocan.iarc.fr/). Nevertheless, less are often in excess but have a deeper following trough,
Chapter 1: Epidemiology
NORTHERN NORTHERN
AMERICA AMERICA
OCEANIA OCEANIA
EUROPE EUROPE
ASIA ASIA
18 16 14 12 10 8 6 4 2 0 2 4 6 8 10 12 14 16 18
B Age Standardized Rate
Male rate Female rate
Male cases Female cases
Number of Cases
20000 18000 16000 14000 12000 10000 8000 6000 4000 2000 0 2000 4000 6000 8000 10000 12000 14000 16000 18000 20000
NORTHERN NORTHERN
AMERICA AMERICA
EUROPE EUROPE
OCEANIA OCEANIA
ASIA ASIA
Figure 1.1 Estimated numbers and age-standardized (world population) incidence rates by region for (A) non-Hodgkin lymphoma and
(B) Hodgkin lymphoma. (Source: GLOBOCAN.)
and a late age peak with a pronounced male excess each year, HMRN comprises an ongoing population-
(Figure 1.2). The earlier age peak is not clearly evident based cohort of all newly diagnosed hematological
in less well developed regions of the world. Inspection malignancies (pediatric and adult). Importantly, as a
of various case-series has revealed that these patterns matter of policy and irrespective of treatment intent,
are the result of differences in the age and gender all diagnoses within HMRN are made and coded to the
frequencies of the various HL subtypes. Likewise, the latest WHO classification by a single specialist hema-
comparatively smooth pattern seen when all NHLs are topathology laboratory.
combined (Figure 1.2) conceals considerable age and HMRN patients newly diagnosed with lymphoma
sex subtype variability. in the 5 years from September 2004 to August 2009 are
The diagnostic challenges posed by hematological proportionately distributed by WHO diagnostic cate-
malignancies means that subtype data can only be gory in Figure 1.3, beginning with the B-cell NHLs,
obtained from specialist registries. Furthermore, moving clockwise through the T-cell NHLs and ending
population-based data with clearly defined numera- with the HLs. Diffuse large B-cell (DLBCL), follicular
tors and denominators (as opposed to hospital-based (FL), and marginal zone lymphomas (MZL) dominate –
case-series) are required for epidemiological research. together accounting for more than 70% of the total. The
One such registry is the UK-based Haematological high proportion of DLBCL and MZL is a common
Malignancy Research Network (www.HMRN.org), feature evident in all reported series, including hospital-
and descriptive age and sex patterns for the lympho- based registers in Asia. By contrast, in certain Asian and
mas diagnosed within HMRN are presented in other population groups, diagnoses of the more indo-
Figures 1.3 and 1.4. Established in 2004 and covering lent FL appear to occur far less frequently, whereas 3
a population of 3.6 million with over 2100 diagnoses diagnoses of T-cell lymphomas are more common.
Chapter 1: Epidemiology
60
400 50
40
30
200
20
10
0 0
4
10 9
15 4
20 9
25 24
30 9
35 4
40 9
45 44
50 9
55 54
60 59
65 4
70 9
75 4
80 9
4
+
0–
5–
–1
–1
–2
–3
–3
–4
–6
–6
–7
–7
–8
85
–
–
–
B
120 6
100 5
80 4
60 3
40 2
20 1
0 0
4
10 9
15 4
20 9
25 4
30 9
35 4
40 9
45 4
50 9
55 4
60 9
65 4
70 9
75 4
80 9
4
+
0–
5–
–1
–1
–2
–2
–3
–3
–4
–4
–5
–5
–6
–6
–7
–7
–8
85
The corresponding HMRN box-and-whisker age lymphoma shows the least variation. By contrast, with a
distributions and sex rate ratios (male rate:female rate), median diagnostic age approaching 71 years, but with
together with their standard errors, are shown in Figure several sporadic pediatric cases, DLBCL covers the larg-
1.4. Whilst most B-cell NHLs have a median diagnostic est age range – the scatter of outliers at younger ages is
age over 70 years, a significant minority tend to be indicative, perhaps, of diagnostic heterogeneity within
diagnosed at younger ages – follicular, Burkitt, and this subtype category.
mediastinal lymphomas in HMRN having median ages Overall, in agreement with other reports, within
of 65, 52, and 36 years, respectively. Furthermore, some HMRN T-cell NHLs tend to be diagnosed at younger
subtypes have broad age ranges whilst others are narrow; ages than B-cell neoplasms. Nonetheless, within T-cell
4 with no patients diagnosed before the age of 48 years and forms of the disease there is considerable subtype
a median of 74 years, the age distribution of mantle cell heterogeneity – the tight age band for enteropathy
Chapter 1: Epidemiology
Enteropathy-type T-cell
Mycosis fungoides
Primary cutaneous CD30 positive T-cell
Anaplastic large cell-T/null type
Peripheral T-cell lymphoma – common; unspecified
Angioimmunoblastic T-cell
0 10 20 30 40 50 60 70 80 90 100 0 1 2 3 4 5
Age at Diagnosis (years) Sex Rate Ratio
Figure 1.4 Lymphoma box-and-whisker plot age distributions, and sex rate ratios with standard errors (M:F). (Haematological Malignancy
Research Network, 2004–2009.) (Legend below).
Outliers
type T-cell contrasting with that seen for anaplastic That males are far more likely to develop lymphoma
large T-cell, for example. Lastly, with a pronounced than females is conspicuously clear in Figures 1.1, 1.2,
pediatric component, HL tends to be diagnosed earlier and 1.4 – but it is also evident that the gender disparity
still, but, again, there are differences between the sub- is much greater for some subtypes, whilst being almost
types – the median ages within the classical Hodgkin absent for others. Among B-cell NHLs, roughly equal
lymphoma (CHL) category, for example, ranging from numbers of males and females were diagnosed with FL
37 years for nodular sclerosis CHL to 60 years for and with extranodal MZL; however, the sex rate ratio
mixed cellularity CHL. These subtype differences are for all other B-cell lymphomas shows a male bias – the
largely responsible for the bimodal HL age and sex most striking being for Burkitt lymphoma which, in 5
distributions shown in Figure 1.2. these and other series, is at least three times as likely to
Chapter 1: Epidemiology
20
Rate per 100,000
15 Female rate
Male rate
10
0
1970 1975 1980 1985 1990 1995 2000 2005 2010
Year
B 4.5
3.5
3
Rate per 100,000
2.5
Female rate
2 Male rate
1.5
0.5
0
1970 1975 1980 1985 1990 1995 2000 2005 2010
Year
be diagnosed in males than in females. T-cell NHLs and emerge as the WHO classification becomes more
HLs also exhibit a male bias, but again there is consid- widely applied in a population-based context..
erable subtype heterogeneity, with angioimmunoblastic
T-cell lymphoma running counter to the trend by being
more common in women. Changes over time
The wide diversity of descriptive patterns shown in Monitoring disease trends over time is a fundamental
Figure 1.4 is indicative of different subtype etiologies, activity of descriptive epidemiology, such analyses
which many new and ongoing epidemiological studies often yielding important etiological clues. Indeed,
are aiming to address. This, coupled with the geo- there are many examples in the field of cancer epi-
graphic variations for T- and B-cell subtypes reported demiology where this has been the case, particularly in
for various case-series, underscore the importance of relation to the identification of hazardous occupa-
using appropriate classifications. Indeed, the contin- tional and environmental exposures. In this context,
ued application of site-based classification systems for the temporal changes reported for NHL in recent
routine cancer registration severely limits the use of decades are unquestionably dramatic, as can be seen
6 their data in epidemiological studies. Hopefully, in the from Figure 1.5, which shows the estimated age-
future, new insights into lymphoma epidemiology will adjusted incidence rates from the Surveillance,
Chapter 1: Epidemiology
Epidemiology and End Results (SEER) Program in the practice will undoubtedly remain challenging to evalu-
United States (www.seer.cancer.gov). ate since hematological oncology is changing rapidly,
The sharp increase seen for NHL in the USA with new approaches to treatment and diagnosis con-
between the 1970s and the 1990s was reported in tinually emerging as diverse patient pathways evolve. In
many countries with population-based cancer regis- this regard, the use of more reliable and sensitive diag-
tration systems (both statutory and voluntary), and the nostic techniques continues to lower the threshold of
magnitude and global scale of the effect naturally disease detection – the ability of flow cytometry to detect
captured both scientific and public attention at the small populations of abnormal cells, for example, as well
time (Figure 1.5). In males, the doubling of the NHL as the introduction of the polymerase chain reaction
rate over a comparatively short 20-year period is par- (PCR) allowing the detection of disease at a molecular
ticularly striking, and contrasts with the modest fall in level. Hence, proportionately more of the patients diag-
HL seen among males over the same time period. nosed today have indolent or asymptomatic disease, and
These registrational trends coincided with a period of it remains possible that the increasing numbers of regis-
significant change in clinical understanding and con- trations may be the result, at least in part, of the recog-
sequent marked taxonomic alterations. nition of new group(s) of patients that would not have
During the 1980s the working formulation domi- been diagnosed in previous decades.
nated in North America, whilst Kiel gained ground in
Europe. In 1994, the Revised European–American Etiology
Lymphoma (REAL) classification was published,
For the reasons discussed above, etiological studies con-
and this was followed by the WHO classifications of
ducted in previous decades have often been hampered
2001 and 2008. Quantification of the likely impact of
by the need to aggregate their data into the broad group-
these diagnostic changes on the patterns seen in
ings of NHL or HL, either because primary source
Figure 1.5 is almost impossible to achieve. Whilst it
information was recorded in that way or because diag-
is generally recognized that shifting clinical diagnos-
nostic standards were inconsistently applied. Hence,
tic practice almost certainly fuelled the trends, there
given the underlying variations in pathologies and prog-
remains debate about whether such changes could
nosis, it is perhaps not surprising that epidemiological
have been responsible for all, or only part, of the
reports on lymphomas have often tended to produce
dramatic increase in NHL registrations. In this con-
inconsistent and conflicting results. Hopefully, this sit-
text, for example, the potential etiological role of
uation will improve in the near future as evidence about
human immunodeficiency virus (HIV) and organ
the pathological and clinical diversity of lymphoma
transplantation, as well as other viruses and environ-
subtypes continues to accumulate, not only revealing
mental exposures, attracted considerable attention at
descriptive differences such as those outlined in the
the time – and these factors are discussed in more
previous section, but also other associations.
detail in the section on causes below.
With respect to the underlying causes of lym-
Whatever the cause, the rate of change in NHL
phoma, it is generally agreed that immune dysregula-
slowed in the USA towards the end of the twentieth
tion plays a pivotal role in lymphogenesis, and most
century, and similar plateauing has now been reported
epidemiological research has concentrated on factors
for other developed regions of the world. Taken at face
and exposures that interact with the immune system –
value, this would seem to support the notion that
particularly infection, immunosuppression, and auto-
changes in disease detection, diagnostic practice, and
immune disease. These interrelated associations are
cancer registration procedures are all likely to have
discussed in the sections that follow.
contributed to the increase seen over the 1970s and
1990s. Indeed, a recent report examining incidence
and survival trends across Europe concluded that ‘the Infection
evolving classification and poor standardization of In the context of immunodeficiency (see following sec-
data collected on hematological malignancies vitiate tion), the strong association between human immuno-
the comparisons of disease incidence and survival over deficiency virus/acquired immunodeficiency syndrome
time and across regions’ (Sant et al., 2009). (HIV/AIDS) and several B-cell lymphomas has been
Importantly, with respect to the examination of examined in numerous studies – the increased risks
future time-trends, the role of changing diagnostic reported range from 10- to 300-fold, the highest 7
Chapter 1: Epidemiology
risks generally been observed for the more aggressive of current investigation. HHV-8 has been found in
NHL subtypes and the lowest risks for the HL subtypes. several comparatively rare tumor subtypes, where it
Unsurprisingly, the potential role of the HIV/AIDS sometimes occurs with EBV. The nature of these
epidemic was one of the factors evaluated in the associations, as well as those with a number of other
context of the striking rise of NHL observed in the viruses, including both hepatitis B and C, are topics
1970s–90s discussed above. In this context it is impor- of much ongoing research.
tant to note that NHL rates had begun to rise before the Albeit by very different mechanisms, bacterial
onset of the HIV/AIDS epidemic in the 1980s infection resulting in chronic inflammation has
(Figure 1.5), and in the era of modern therapies the been consistently linked with a marked increase in
contribution of HIV/AIDS to the totality of lymphomas the risk of some NHL subtypes – one of the best
diagnosed in developed countries remains compara- known associations being that between Helicobacter
tively small. pylori and gastric mucosa-associated lymphoid tissue
In addition to HIV, a number of other viruses are (MALT) extranodal MZL; the presence of the bacteria
either known or thought to impact on lymphoma risk, increase lymphoma risk by about sixfold. As with the
albeit by different mechanisms. The most accepted viral associations (e.g. HTLV-1 and ATLL), bacterial
viral associations are those with human T-cell leuke- infection often occurs many years before MALT lym-
mia virus type 1 (HTLV-1) and the two herpesviruses phoma development, the median age of diagnosis of
Epstein–Barr virus (EBV) and human herpesvirus 8 all extra-marginal zone lymphomas combined in the
(HHV-8) – also known as Kaposi’s sarcoma herpesvi- HMRN series being just under 70 years with no
rus (KSHV). Infection with HTLV-1 is a necessary but obvious sex bias (Figure 1.4). H. pylori infection,
not sufficient cause of adult T-cell leukemia/lym- however, generally occurs in childhood – the preva-
phoma (ATLL), with ATLL developing in around 3% lence being highest in developing countries and fall-
of those infected. ATLL principally occurs in areas ing as socioeconomic status increases. In the future,
where HTLV-1 is endemic, including Japan, the the introduction of effective H. pylori treatments,
Caribbean, and parts of central Africa and, like most coupled with increasing affluence in developed
other lymphomas, it is more common in males than regions of the world, may well lead to a reduction in
females. incidence of H. pylori-positive gastric MZLs.
In contrast to the specific nature of the HTLV-1/ Interestingly, with respect to the timing of infection
ATLL association, the globally ubiquitous EBV fea- and subsequent lymphoma development, detailed anal-
tures in several lymphoma subtypes, including ysis of medical records collected during the course of a
Burkitt lymphoma and CHL, as well as those lym- UK case-control study revealed significant increases in
phomas occurring in immunosuppressed individu- the frequency of non-specific infectious illness episodes
als. EBV is invariably associated with the endemic more than 10 years before the diagnosis of CHL, but not
form of Burkitt lymphoma that occurs in equatorial of DLBCL and FL. No associations with specific infec-
Africa and New Guinea, where, with a median age of tions were, however, found, and whether or not the
around 7 years and a male to female ratio of 2:1, it is excess seen for CHL was a consequence of an under-
the commonest pediatric malignancy. However, as lying immune abnormality or whether infection played
can be seen from Figure 1.4, the median age at onset a causal role could not be determined.
of the comparatively rare sporadic form of Burkitt
lymphoma seen in other areas of the world is much
later, and the male predominance is considerably Immunosuppression
greater. In contrast to endemic Burkitt’s, EBV is In addition to HIV/AIDS, it has been known for some
only associated with around a third of all sporadic time that both inherited and acquired immunodefi-
cases, and the role of EBV in these tumors remains a ciency syndromes predispose towards lymphoprolifer-
continuing area of research. Likewise, the signifi- ative disorders. A few rare inherited disorders of the
cance of the well-established association between immune system, including Wiskott–Aldrich syn-
EBV and CHL, observed most frequently in children drome and ataxia telangiectasia, are well known to be
and the elderly in around a third of all cases in associated with marked increases in the risk of lym-
resource-rich countries (but even more commonly phoproliferative disorders. Such conditions are, how-
8 in less affluent parts of the world), remains an area ever, exceedingly rare in the general population, and
Chapter 1: Epidemiology
the proportion of total lymphoma diagnoses for which Taken as a whole, there is broad consensus that
they account is correspondingly small. elucidating the mechanisms underpinning the link
In addition to these rare inherited conditions, lym- between autoimmune disease and the lymphomas
phoma risks in organ transplant recipients have been could lead to fundamental insights into the biology of
extensively investigated, and there are many ongoing both groups of disorders, and there is considerable
patient cohorts. The spectrum of lymphoma subtypes speculation about potential mechanisms in the litera-
that occur in immunosuppressed individuals tends to ture. In this regard routine examination of sex-specific
differ in important respects from those seen in the non- associations could provide valuable information since it
immunosuppressed, being more aggressive, less respon- has long been known that most autoimmune conditions
sive to therapy, often EBV-associated, and more likely are considerably more common in women than in men,
to occur at extranodal sites. Indeed, post-transplant whereas the reverse is true for the majority of lympho-
lymphoproliferative disorders (PTLD) are generally proliferative malignancies. Furthermore, there are well-
considered separately, the latest WHO classification documented sex-specific variations in immune response
distinguishing several different malignant and benign that may well impact on the risk of these diseases.
PTLD disease forms. Immunosuppression is recognized
as the major causal factor in PTLD, the likelihood of
development varying with the patient’s age and type of Other environmental factors
transplant – the incidence being lowest following renal The dramatic increase in lymphoma registrations seen
transplant and highest following heart and lung. at the end of the last century (Figure 1.5) resulted
Organ transplantation is an increasingly successful in a massive increase in epidemiological research.
therapy and, as with HIV/AIDS, it is one of the factors Accordingly, in addition to investigating the explicit
often considered to have contributed to the temporal immunological factors discussed above, many other
increase in NHLs seen at the end of the last century putative risk factors have been examined, including
(Figure 1.5). Nonetheless, organ transplantation genetic predisposition, associations with other comor-
remains comparatively rare and the proportion of bidities, and a wide range of occupational and other
the increase it could account for remains small. potentially hazardous environmental exposures.
However, despite concentrated effort, many of the
potentially hazardous associations identified have
Autoimmune disease been weak or contradictory, with few consistent asso-
Several strong associations between autoimmune dis- ciations emerging. Indeed, at the present time, the lack
ease and subsequent lymphoma development have of accepted environmental determinant(s) for lym-
been described. In particular, links with autoimmune phomas sits starkly against the accumulated knowl-
disease-based chronic inflammation and MALT lym- edge about other cancers such as lung, stomach, skin,
phoma are well recognized – two of the strongest and bladder, and breast.
most well-known relationships being those between With respect to the continued investigation of lym-
Hashimoto thyroiditis and thyroid lymphoma and phoma determinants, the InterLymph consortium – a
between Sjögren’s syndrome and salivary gland lym- scientific forum for epidemiologic research on lym-
phoma. Consistent associations have also been phoma – was established in 2001. InterLymph members
reported for other chronic inflammatory rheumatic have reviewed and published on several health-related
diseases; for example, rheumatoid arthritis and sys- states/events and environmental exposures – and their
temic lupus erythematosus with both MZL and website includes all their reports and provides up-to-
DLBCL, and gastrointestinal inflammatory autoim- date information on many putative risk factors (http://
mune conditions, such as celiac disease and Crohn’s epi.grants.cancer.gov/InterLymph/). Thus far, topics
disease, with the T-cell lymphomas. Indeed, evidence investigated by InterLymph have included self-reported
linking lymphoproliferative disorders with autoim- obesity (rising levels being another potential factor sug-
munity continues to accumulate, the strongest associ- gested as contributing to temporal trends), smoking
ations being seen for MZL, DLBCL, and T-cell and alcohol (no consistent effects found); sun exposure
lymphoma, and the weakest with the relatively indo- and history of atopic infections (marginally reduced
lent FL. Associations between chronic inflammatory risks found); and family history of hematological can-
rheumatic diseases and CHL, but not nodular sclerosis cers (increased risks confirmed). Finally, with a view to 9
HL, have also been reported. investigating genetic susceptibility, the majority of
Chapter 1: Epidemiology
InterLymph studies collected constitutional biological Marcos-Gragera R, Pollán M, Chirlaque MD, et al. Attenuation
material from cases and corresponding controls, and of the epidemic increase in non-Hodgkin’s lymphomas in
findings here have, perhaps, been more informative. Spain. Ann Oncol, 2010;21 Suppl 3:iii90–96.
Thus far polymorphisms in two immune system- Mozaheb Z, Aledavood A, Farzad F. Distributions of major
related genes have been identified using the candidate sub-types of lymphoid malignancies among adults in
gene approach, associations being seen for both DLBCL Mashhad, Iran. Cancer Epidemiol [Internet]. 2010, Oct.
29 [cited Jan. 10, 2011]; Available from: http://www.ncbi.
and MZL, and genome-wide association studies are nlm.nih.gov/pubmed/21036690
currently ongoing.
Orem J, Mbidde EK, Lambert B, de Sanjose S, Weiderpass E.
Burkitt’s lymphoma in Africa, a review of the epidemiology
and etiology. Afr Health Sci, 2007;7(3):166–175.
Further reading
Relander T, Johnson NA, Farinha P, et al. Prognostic
Anderson LA, Gadalla S, Morton LM, et al. Population-
factors in follicular lymphoma. J Clin Oncol,
based study of autoimmune conditions and the risk of
2010;28(17):2902–2913.
specific lymphoid malignancies. Int J Cancer, 2009;
125(2):398–405. Sagaert X, Van Cutsem E, De Hertogh G, Geboes K,
Tousseyn T. Gastric MALT lymphoma: a model of
Ansell P, Simpson J, Lightfoot T, et al. Non-Hodgkin
chronic inflammation-induced tumor development. Nat
lymphoma and autoimmunity: does gender matter? Int J
Rev Gastroenterol Hepatol, 2010;7(6):336–346.
Cancer, 2011;129(2):460–466.
Sant M, Allemani C, Santaquilani M, et al. EUROCARE-4.
Barrans S, Crouch S, Smith A, et al. Rearrangement of MYC
Survival of cancer patients diagnosed in 1995–1999.
is associated with poor prognosis in patients with diffuse
Results and commentary. Eur J Cancer,
large B-cell lymphoma treated in the era of rituximab. J
2009;45(6):931–991.
Clin Oncol, 2010;28(20):3360–3365.
Sant M, Allemani C, Tereanu C, et al. Incidence of
Boffetta P, Armstrong B, Linet M, et al. Consortia in cancer
hematologic malignancies in Europe by morphologic
epidemiology: lessons from InterLymph. Cancer
subtype: results of the HAEMACARE project. Blood,
Epidemiol. Biomarkers Prev, 2007;16(2):197–199.
2010;116(19):3724–3734.
Bosetti C, Levi F, Ferlay J, et al. Incidence and mortality from
Smedby KE, Hjalgrim H, Askling J, et al. Autoimmune
non-Hodgkin lymphoma in Europe: the end of an
and chronic inflammatory disorders and risk of
epidemic? Int J Cancer, 2008;123(8):1917–1923.
non-Hodgkin lymphoma by subtype. J Natl Cancer Inst,
Boyle P. International agency for research on cancer. World 2006;98(1):51–60.
Cancer Report 2008. Lyon: IARC Press; 2008.
Smith A, Roman E, Howell D, et al. The Haematological
Crouch S, Simpson J, Ansell P, et al. Illness patterns prior to Malignancy Research Network (HMRN): a new
diagnosis of lymphoma: analysis of UK medical records. information strategy for population based
Cancer Epidemiol [Internet]. 2010, Sept. 8 [cited Sept. 15, epidemiology and health service research. Br J
2010]; Available from: http://www.ncbi.nlm.nih.gov/ Haematol, 2010;148(5):739–753.
pubmed/20832384
Swerdlow S. International Agency for Research on Cancer,
Dal Maso L, Serraino D, Franceschi S. Epidemiology of World Health Organization. WHO Classification of
AIDS-related tumours in developed and developing Tumours of Haematopoietic and Lymphoid Tissues. 4th
countries. Eur J Cancer, 2001;37(10):1188–1201. ed. Lyon, France: International Agency for Research on
Ferlay J, Shin H, Bray F, et al. Estimates of worldwide Cancer, 2008.
burden of cancer in 2008: GLOBOCAN 2008. Int J Végső G, Hajdu M, Sebestyén A. Lymphoproliferative
Cancer [Internet]. 2010, June 17 [cited Aug. 26, 2010]; disorders after solid organ transplantation–classification,
Available from: http://www.ncbi.nlm.nih.gov/pubmed/ incidence, risk factors, early detection and treatment
20560135 options. Pathol Oncol Res [Internet]. 2010, Dec. 31 [cited
Jarrett RF. Viruses and lymphoma/leukaemia. J Pathol, Jan 12, 2011];Available from: http://www.ncbi.nlm.nih.
2006;208(2):176–186. gov/pubmed/21193979
Jemal A, Siegel R, Xu J, Ward E. Cancer statistics, 2010. CA Westlake S. Cancer incidence and mortality in the United
Cancer J Clin, 2010;60(5):277–300. Kingdom and constituent countries, 2004–06. Health Stat
Q, 2009;43:56–62.
Luminari S, Cesaretti M, Marcheselli L, et al. Decreasing
incidence of gastric MALT lymphomas in the era of anti- Yoon SO, Suh C, Lee DH, et al. Distribution of lymphoid
Helicobacter pylori interventions: results from a neoplasms in the Republic of Korea: analysis of 5318 cases
population-based study on extranodal marginal zone according to the World Health Organization
10 lymphomas. Ann Oncol, 2010;21(4):855–859. classification. Am J Hematol, 2010;85(10):760–764.
Chapter
2
Prognostic factors for lymphomas
Guillaume Cartron and Philippe Solal-Céligny
Lymphoma, Second Edition, ed. Robert Marcus, John W. Sweetenham, and Michael E. Williams. Published by 11
Cambridge University Press. © Cambridge University Press 2014.
Chapter 2: Prognostic factors for lymphomas
and five) risk groups. However, the widely in other patients who may differ by age, distribution,
expanding use of “pocket computers” and smart and treatment modalities. This external validation is
phones will permit us to use more complex indices mandatory before widespread adoption of the PI. A
and/or algorithmic programs, e.g. the MIPI univariate analysis is the first step. Parameters here are
(Mantle Cell Lymphoma International Prognostic selected from the literature and will include demo-
Index). graphic data, clinical and biological markers of tumor
burden, and the effect of the lymphoma upon the host.
This univariate analysis may, however, generate bias:
It must ideally rely on overall survival (OS) when creating the FLIPI, it appeared that including the
OS is the most convenient endpoint. PIs have been cell subtype of follicular lymphoma would require a
developed from retrospective analyses of large groups pathology review since the distribution was obviously
of patients. In these retrospective studies, determining center-dependent. Another example is the absence of
the time to progression (TTP) or time to treatment inclusion in all indices of comorbidities, since the pres-
failure (TTF) may lead to uncertainties and is a poten- ence and the severity of a concomitant disease will
tial source of bias. influence prognosis. This is especially true in the eld-
However, OS cannot be used in analyses of patients erly. New scores for evaluating these comorbidities and
with some lymphoma subtypes, as in Hodgkin’s dis- their severity have been proposed and should be inte-
ease or indolent lymphomas where the OS is very grated in new clinical PIs.
good, events are rare and an analysis would require There is no consensus on the optimal method for
many years of follow-up. It is now widely accepted – including continuous variables in univariate analyses.
although not definitively proved – that significant They may be included as a continuous variable –
improvement in progression-free survival (PFS) is a which is the most accurate but the most complex
good surrogate of an improvement in OS that can be option – or dichotomized according to a threshold
used in the evaluation of the influence of a new treat- change in prognosis beyond a certain point (for
ment in different prognostic subgroups. instance, the threshold of 60 years was chosen for the
IPI since it was considered a maximal age for autolo-
It must be discriminant gous bone marrow transplantation). The expansion of
To be useful to clinicians, a PI must separate patients computer software in daily use by physicians in their
into risk groups with significantly different OS or PFS day-to-day practice will allow more and more use of
when compared by log-rank test and hazard ratio for continuous variables in PIs. Parameters that signifi-
either of these endpoints. cantly influence the OS of patients in the univariate
Furthermore, except in some common subtypes analysis should then be included in the multivariate
such as large cell NHL or Hodgkin’s disease, there analysis. However, if the number of patients is very
must be an even distribution between risk groups to high, a selection of parameters has to be made accord-
allow assessment of a new treatment in an adequate ing to statistical (P value in the univariate analysis) and
number of patients. In developing the PI, it may also clinical considerations. Parameters that significantly
be useful in some frequently occurring subtypes to influence OS in the multivariate analysis are chosen
identify a small group of patients, either with a very for building the PI. Detailed statistical considerations
good or a very poor prognosis. are beyond the scope of this chapter.
classification was originally developed for Hodgkin’s Table 2.1 Prognostic factors for survival in International
disease, which commonly spreads via contiguous Index Patients.
lymph node groups. Because the patterns of spread are
Relative P
somewhat different between NHL and HL, it is therefore risk value
not surprising that the Ann Arbor classification was less
(A) Patients of all ages
efficient in identifying prognostic subgroups of patients Age (≤60 years versus >60 years) 1.96 <0.001
with aggressive NHL. Age at diagnosis was also com- LDH (≤normal versus >normal) 1.85 <0.001
monly identified as a prognostic factor in numerous Performance status (0,1 versus 1.80 <0.001
2–4) 1.47 <0.001
studies. However, most publications that included large Ann Arbor stage (I/II versus III/IV) 1.48 <0.001
numbers of older patients, demonstrated a tendency to Extranodal involvement (≤1 site
treat elderly patients with lower doses of chemotherapy versus >1 site)
to reduce treatment-related toxicity, while others sug- (B) Patients ≤60 years of age
gested that older patients could benefit from full-dose Ann Arbor stage (I/II versus III/IV) 2.17 <0.001
LDH (≤normal versus >normal 1.95 <0.001
therapy. In this context, many investigators attempted to Performance status (0, 1 versus 1.81 <0.001
identify pre-treatment prognostic factors from their own 2–4)
series of patients with aggressive NHL. Features inde-
pendently associated with outcome were often used to
develop prognostic-factor models in which the individ-
ual patient’s risk correlated to the number of adverse death was determined by adding the number of adverse
factors present at diagnosis. Although the clinical fea- prognostic factors present at diagnosis. It was therefore
tures used in these models differed, they were thought to possible to identify four groups that had significantly
reflect three basic features: different predicted 5-year OS: 73% (low-risk group
(1) the tumor’s growth and invasive potential: lactate with either no or one adverse factor); 51% (low–
dehydrogenase (LDH), Ann Arbor stage, tumor intermediate risk group with two adverse factors); 43%
size, bone marrow (BM) involvement, and number (high–intermediate risk group with three adverse
of nodal and extranodal sites; factors); and 26% (high-risk group with four or five
adverse factors). Since age was an independent predictive
(2) the patient’s reaction to the tumor: performance
factor for outcome and patients younger than 60 years
status (PS), B symptoms;
were often candidates for intensive treatment, an age-
(3) the patient’s likely ability to tolerate intensive
adjusted model was also developed. From 1274 patients
therapy: PS, BM involvement, age.
aged 60 years or younger, three adverse factors were
The models were similarly predictive for outcome in identified: tumor stage, serum concentration of LDH,
large series of patients uniformly treated, suggesting and PS. By adding the number of adverse prognostic
that, although they relied on different parameters, they factors, this age-adjusted IPI (aaIPI) identified four risk
identified the same patient subpopulations. groups with a predicted 5-year OS of 83% (low-risk
In 1993, institutions and cooperative groups from the group with no adverse factor), 69% (low–intermediate
United States, Canada, and Europe participated in the risk group with one adverse factor), 46% (high–inter-
International Non-Hodgkin’s Lymphoma Prognostic mediate risk group with two adverse factors), and 32%
Factors Project to attempt to develop a predictive (high-risk group with three adverse factors).
model (the IPI) based on a large cohort of patients with When clinical characteristics of patients above
aggressive NHL. This model was designed to allow com- and below 60 years were compared, it was seen that
parisons of published clinical trials and to identify equal percentages of younger and older patients pre-
patients at high risk for relapse after anthracycline- sented with elevated LDH level, advanced stage, poor
based chemotherapy. From a study of 2031 patients performance status PS (2–4), or had multiple extra-
treated with an anthracycline-based chemotherapy nodal sites. These demonstrated that the worst out-
between 1982 and 1987, clinical features independently come of older patients was not because of more
predictive for OS and relapse were age, PS, LDH levels, extensive disease. The comparison of complete
Ann Arbor stage, and the number of extranodal sites. remission (CR) and relapse-free survival (RFS) rates
These five factors were incorporated into a model between older and younger patients (Table 2.2) also
(Table 2.1) and an individual patient’s relative risk of indicated that they had similar CR rates but lower 13
Chapter 2: Prognostic factors for lymphomas
Risk group Risk factors Distribution CR rate (%) RFS of CR (%) Survival (%)
of cases (%)
2-year 5-year 2-year 5-year
rate rate rate rate
(A) International Index
(patients of all ages)
Low (L) 0,1 35 87 79 70 84 73
Low–intermediate (LI) 2 27 67 66 50 66 51
High–intermediate (HI) 3 22 55 59 49 54 43
High (H) 4,5 16 44 58 40 34 26
(B) Age-adjusted Index
(patients ≤60 years)
Low (L) 0 22 92 88 86 90 83
Low–intermediate (LI) 1 32 78 74 66 79 69
High–intermediate (HI) 2 32 57 62 53 59 46
High (H) 3 14 46 61 58 37 32
(B) Age-adjusted Index
(patients >60 years)
Low (L) 0 18 91 75 46 80 56
Low–intermediate (LI) 1 31 71 64 45 68 44
High–intermediate (HI) 2 35 56 60 41 48 37
High (H) 3 16 36 47 37 31 21
RFS rates that translated into significant age-related groups: low and low–intermediate groups with 4-year
differences in survival. When survival of patients OS of 82% and 81%, respectively, and high and high–
with different Ann Arbor stages was compared intermediate groups with 4-year OS of 49% and 59%,
according to risk groups defined by the IPI, it was respectively. They proposed to redistribute these pre-
possible to determine in each different stage signifi- dictive factors into a Revised International Prognostic
cantly different outcomes, indicating that the IPI was Index (R-IPI) and identified three distinct prognostic
more predictive than the Ann Arbor classification. groups with a very good (no adverse prognostic factor,
The IPI was derived from a cohort of patients with 4-year PFS: 94%, OS: 94%), good (1 or 2 adverse
aggressive lymphoma, including both T-cell lym- factors, 4-year PFS: 80%, OS: 79%), and poor (3 or
phoma and different subtypes of aggressive B lym- more adverse factors; 4-year PFS: 53%, OS: 55%) out-
phoma. In addition, none of the patients received the come, respectively (Table 2.3).
current gold standard treatment of rituximab and The IPI or R-IPI are now used to design therapeu-
CHOP (cyclophosphamide, doxorubicin, vincristine, tic trials and to select and compare treatment strat-
and prednisone)-like chemotherapy. Subsequently, the egies, but these clinical variables clearly represent
predictive value of IPI was reassessed in a retrospective surrogates for the intrinsic cellular and molecular het-
study in 365 patients with DLBCL who received six to erogeneity within aggressive lymphoma and highlight
14 eight cycles of R-CHOP. The authors found that IPI the need for patient-specific and biologically based risk
remained predictive but they identified only two risk factors.
Chapter 2: Prognostic factors for lymphomas
Table 2.4 Immunohistochemical markers having prognostic value in diffuse large B-cell NHL.
subgroups. Cases with CD10 expression by 30% of cells expressed genes found in in vitro activated peripheral
are regarded as GC type as well as cases that are CD10–, blood B-cells as well as some genes expressed normally
BCL-6+, and IRF4/MUM1–. All other cases are regarded by plasma cells, suggesting their post-GC origin.
as of non-GC type. However, this immunophenotyping Initially, a “type 3” group was defined, including a
division does not correlate well with gene expression- collection of DLBCL which cannot be classified as the
based subgrouping of DLBCL (see Gene expression GC and ABC groups, but this does not represent a
profiling, below). distinct group. This stratification related to stages of
B-lymphocyte ontogeny has been found to be correlated
Genetic abnormalities with OS independently of IPI, GC-like DLBCL having
Several recurrent chromosomal abnormalities have significantly better OS rates when compared to those
been described in DLBCL, including those involving with ABC-like DLBCL. However, the microarrays used
BCL-6 (3q27), BCL-2 (18q21), P53 (17p), and REL in these studies were different and there was no overlap
(2p14). None of them are pathognomonic of DLBCL. between the genes identified in these studies. The clin-
BCL-6 is located in 3q27 and is involved in chromo- ical utility of such technologies is also limited by the cost
somal translocations or mutations leading to deregu- and the requirement for fresh or optimally cryopre-
lation of BCL-6 expression. The chromosomal served tumor. A simplified 6-gene predictive model
translocations involving 3q27 are the most frequent using quantitative real-time polymerase chain reaction
genetic abnormalities and are detected in 30–40% of was subsequently developed, demonstrating that the
DLBCL. Contrary to the predictive value of increased expression of three genes (LMO2, BCL-6, FN1) corre-
BCL-6 mRNA and BCL-6 protein expression, the clin- lated with prolonged survival, whereas the expression of
ical importance of such translocations has been con- another set of three genes (BCL-2, CCND2, SCYA3) was
troversial and most studies have failed to demonstrate associated with a shorter survival. The predictive value
a significant effect on survival. Translocation t(14;18) of the gene expression profile was later confirmed using
(q32;q21) is found in up to 30% of DLBCL, and most tissue microarray (TMA), suggesting the potential of
early studies have not shown a prognostic impact of this technology to replace gene expression profiling in
this translocation in de novo DLBCL. clinical practice. Gene expression profiling was origi-
Mutations and deletion of P53 are reported in up to nally obtained with tumor samples from patients trea-
20% of DLBCL and this has been associated with a ted without rituximab. The predictive value of gene
more aggressive clinical course. However, p53 protein expression profiling has been subsequently demonstra-
expression is imperfectly correlated with P53 mutation ted in a more recent cohort of patients receiving
and it has been suggested that the analysis of p53 R-CHOP.
with its downstream target p21 may identify a sub-
group of DLBCL with a particularly poor outcome Treatment-related prognostic factors
(p53+/p21–). MYC rearrangement was observed in up Although parameters associated directly with response
to 10% of cases of DLBCL. The MYC break partner is to treatment cannot be used to define induction ther-
IGH (60%) or non-IGH (40%) genes. In some, MYC apy, they provide information regarding the chemo-
rearrangements are associated with a concurrent BCL- sensitivity of the tumor. The recent introduction
2–IGH translocation and/or BCL-6 breakpoint or both. of positron emission tomography (PET) using
18
These cases are usually characterized by a high prolif- F-fluorodeoxyglucose (18F-FDG) has improved our
eration index and have a very poor prognosis. ability to evaluate the prognostic relevance of
response.
Gene expression profiling
Studies using gene expression profiling have identified Time to complete remission
different patterns of gene expression allowing the sep- The survival of patients with DLBCL is closely related
aration of DLBCL subtypes deriving from distinct to achieving CR at the end of treatment, whereas stable
stages of lymphocyte ontogeny. The first group, or progressive disease (refractory disease) are both
named GC-like DLBCL, expresses genes characteristic associated with poor survival. The time required to
of normal B-cells in the germinal center, whereas non- obtain CR has also been identified as an important
16 GC-like DLBCL consists of cells with the character- prognostic factor of OS, with patients achieving CR
istics of activated B-cells (ABC). The ABC-like tumors early having the best survival. However, evaluation of
Chapter 2: Prognostic factors for lymphomas
Table 2.5 Predictive value of whole-body PET–18F-FDG for treatment evaluation in NHL.
response by clinical examination and computer correlated better with PFS than those observed after
tomography scans is often limited to evaluating completion of chemotherapy. Results of prospective
patients with bulky or disseminated tumors who fre- studies evaluating the impact of interim PET–18F-FDG
quently (30–60%) have residual abnormalities of in patient management are awaited.
uncertain significance.
Dose intensity and schedule
Positron emission tomography (PET) using Until the advent of rituximab, the CHOP regimen
18
F-fluorodeoxyglucose (18FDG) (cyclophosphamide, doxorubicin, vincristine, and pred-
PET–18F-FDG is now considered as the non-invasive nisone) was considered as the gold standard for the
imaging technique of choice for the detection of resid- treatment of DLBCL. Early phase II studies have high-
ual disease after treatment. The positive predictive lighted the potential role of dose intensity and schedule
value of PET–18F-FDG for relapse is close to 100% of treatment on survival rates. These second and third
(Table 2.5) and persistent PET–18F-FDG uptake after generation regimens failed, however, to demonstrate an
first line therapy is associated with a high risk of impact on survival. Several recent randomized trials have
progression. The follow-up time and the small size of focused again on increasing frequency and intensity of
the studies that included both DLBCL and Hodgkin’s CHOP. They demonstrated that CHOP given every 2
lymphoma are confounding factors, and the prognos- weeks (CHOP-14) or an intensified regimen (ACVBP)
tic impact of PET–18F-FDG at the end of treatment could improve OS rates. More intensive approaches
needs to be confirmed in future studies. Despite these using high-dose chemotherapy followed by ASCT
limitations, PET–18F-FDG has now been incorporated have also been tested in first line therapy. Results are
into the revised response criteria for DLBCL. often conflicting, but could suggest that intensive early
The prognostic value of 18F-FDG uptake has also therapy might benefit age-adjusted IPI high–inter-
been evaluated before autologous stem cell transplanta- mediate (HI)-risk patients. The benefit of adding rit-
tion (ASCT), and positive PET–18F-FDG scan was also uximab to CHOP on survival has been demonstrated
correlated with PFS and OS. PET–18F-FDG before trans- in both elderly patients with advanced DLBCL and in
plantation seems to be an even stronger prognostic fac- young low-risk patients (defined by IPI ≤1). The
tor than IPI and its prognostic significance is higher impact of intensified chemotherapy regimens in the
before rather than after transplantation. Early prediction era of rituximab (R-CHOP-14) has been studied in
of response to therapy could also distinguish those three prospective trials; two have failed to demonstrate
patients who might benefit from standard therapy any survival advantage compared to R-CHOP-21. The
from those for whom intensive strategies may be role of high-dose chemotherapy with rituximab as
more beneficial. A rapid decrease of 18F-FDG uptake part of initial therapy is currently under investigation.
by tumoral tissue has been observed as early as 7 days
after the first administration of chemotherapy in patients Prognostic factors at relapse
with NHL, and it might therefore be possible to separate
early patients with or without residual 18F-FDG uptake. Clinical and treatment-related prognostic factors
Nevertheless, optimal timing to assess response remains Initial therapy of DLBCL with anthracycline-based
unknown. Some authors found that the predictive value chemotherapy cures approximately 40–50% of
of PET–18F-FDG was independent of IPI, and patients and therefore 50–60% of patients will either 17
PET–18F-FDG findings obtained after the first cycle have disease refractory to initial therapy or will relapse
Chapter 2: Prognostic factors for lymphomas
after a complete response. For these patients, high- Prognostic factors at diagnosis
dose chemotherapy followed by ASCT is the most
successful approach, obtaining 40–45% 5-year event- Clinical factors
free survival (EFS) as demonstrated in the PARMA Several retrospective analyses have looked for prog-
trial which compared salvage regimen (DHAP) alone nostic factors in FLs. Those that have been reported
or followed by ASCT. Early studies of high-dose che- as having a significant influence on prognosis are
motherapy have demonstrated that chemosensitivity listed in Table 2.6. The IPI proposed for aggressive
to initial second line therapy was an overwhelming lymphomas has also been tested in FL, where sev-
predictor of outcome. Other additional adverse factors eral studies have shown that the IPI can separate
identified in many of these studies include a remission patients into groups with significantly different
duration shorter than 12 months, an elevated lactate prognoses.
dehydrogenase (LDH), bulk disease greater than However, there are many drawbacks in using the
10 cm, or three or more chemotherapy regimens IPI in FLs: (1) the statistical methodology is inad-
before ASCT. The IPI has also been evaluated at the equate since some factors that were not significant in
initiation of second line therapy as a predictor of out- aggressive lymphomas might influence the prognosis
come. In a retrospective analysis of the PARMA trial, of FL; (2) the IPI is poorly discriminant in FL, with less
aaIPI was highly predictive of both response to DHAP than 15% of patients in the high–intermediate risk and
and OS for the entire cohort of patients. For chemo- high-risk groups.
sensitive patients randomized to autologous stem cell This prompted a group of specialists in 1998 to
transplantation, aaIPI failed, however, to be predictive initiate a collection of data from approximately 5000
of OS. Furthermore, no difference in PFS or OS was patients with FL. After univariate and multivariate
found between the two arms for patients with an aaIPI analyses, the Follicular Lymphoma International
score of 0. Prognostic Index (FLIPI) was devised. It relies on
five parameters (Table 2.7): age, serum LDH, Ann
Gene expression profiling Arbor stage, hemoglobin level, and number of
There are very few studies including gene expression nodal sites. From these, three risk groups were
profiling at the time of relapse and its predictive value established.
is under investigation. The importance of phenotype The characteristics of these groups, distribution of
assessed by TMA in predicting survival in relapse or patients in the test group, 5-year and 10-year survivals,
refractory DLBCL treated with ASCT has recently and relative risks of deaths are seen in Table 2.8.
been reported. No difference in survival was observed The FLIPI has been validated by analysis of other
for patients with either the GC phenotype or the non- series and has been designed with patients treated
GC phenotype. with conventional chemotherapy before the era of
anti-CD20 monoclonal antibodies (MoAb). OS can-
not be used for validation in patients treated with
Follicular lymphomas anti-CD20 MoAbs because of the scarcity of events.
Follicular lymphomas (FLs) account for 25–30% The usefulness of the FLIPI has thus been tested on
of all NHLs. Until 20 years ago, treatment of FLs PFS and has been validated on different series of
was palliative and did not significantly modify the patients treated with cyclophosphamide, vincristine,
natural history of the disease. Since the development and prednisone (CVP) followed by rituximab, CVP
of immunotherapies such as interferon alpha, anti- combined with rituximab, CHOP combined with
CD20 monoclonal antibodies, and combinations rituximab, or treatment with radiolabeled anti-
of these with chemotherapy, overall survival has CD20 MoAb.
improved. However, these treatments may have Since the FLIPI did not include some parameters
immediate and/or long-term toxicity, may have that were not measured routinely at the time of
untoward effects on quality of life, and have variable the analysis (tumor bulk, serum β2 microglobulin
costs. level) and relied on the analysis of patients initially
It has thus become essential for clinicians to have treated without anti-CD20 MoAbs, a new PI has
detailed information on the prognosis of the disease at recently been proposed. The FLIPI2 uses PFS as
18 diagnosis in order to select treatment with the optimal the endpoint and prospectively included a larger
efficacy/toxicity/cost ratios.
Chapter 2: Prognostic factors for lymphomas
Table 2.6 Adverse prognostic factors in follicular lymphomas (from retrospective analyses of the literature).
Characteristics
Demographic Advanced age
Male gender
Pathological Follicular diffuse architecture
Cell type grade 3
Cytogenetic Absence of t(14;18)
Additional cytogenetic abnormalities
6q deletion, 17p, 1p abnormalities
p53 gene mutations
Molecular Overexpression of macrophage and dendritic cell genes
Overexpression of p53
Tumor mass Ann Arbor stage III–IV
Number of nodal sites involved
Massive marrow infiltration
Presence of bulky tumors
Increased serum LDH level
Increased serum β2 microglobulin level
Increased serum CA125
Consequences upon the host Poor performance status
B symptom(s)
Anemia
Hypoalbuminemia
Table 2.8 Outcome and relative risk of death according to risk group as defined by the FLIPI.
Table 2.9 Adverse prognostic factors for progression-free survival in the FLIPI2.
Table 2.10 Outcome and relative risk of progression according to risk group as defined by the FLIPI2.
cyclin-dependent kinase inhibitors located on chro- patients treated with conventional chemotherapy
mosome 9p21. However, these abnormalities have but not in patients treated with rituximab-based
not been reported consistently and their relative con- regimens;
tribution remains unknown. – a high number of CD4-positive T-cells in the
lymphomatous follicles has a negative influence
Gene expression profiling on outcome. However, subpopulations of
Studies have now been carried out on the gene CD4-positive T-cells have opposite effects. FOX
profile of 191 patients with FL. Two signatures were P3-positive Tregs may inhibit FL cell
distinguished, one associated with overexpression of proliferation and thereby positively influence
T-lymphocyte genes and the other with macrophage prognosis. Programmed cell death 1-positive CD4
and/or dendritic cell genes. The former signature was T-cells could also have a positive impact on FL
associated with a good prognosis while the latter was outcome.
associated with a poor prognosis, and survival differ- However, these results should be cautiously inter-
ences were highly significant and independent of clin- preted since discrepant results between studies have
ical factors. been reported.
The role of the microenvironment in FLs has been
extensively studied and many cells play a role: Treatment-related prognostic factors
20 – macrophages in the follicular and interfollicular Several factors may allow the prediction of response
areas have a negative influence on prognosis only in to treatment or duration of response. The significance
Chapter 2: Prognostic factors for lymphomas
of persistence after treatment of t(14;18)-positive cells discriminant since most patients with MCL have
in the blood and/or bone marrow is controversial. adverse prognostic factors according to the IPI (age
It does not seem to influence the risk of relapse >60, Ann Arbor stage IV disease, more than one extra-
after radiation therapy for localized disease, while it nodal involvement site) and are in the high–inter-
was associated with a short PFS after autologous mediate and high-risk groups.
bone marrow transplantation. Some studies suggest a In 2008, The German Low Grade Lymphoma
negative influence for these cells, others an absence Study Group and the European Mantle Cell
of prognostic value. According to a few gene profile Network specifically designed a new prognostic
analyses, some signatures are associated with an index for overall survival of patients with MCL: the
improved response to CHOP or to rituximab. Mantle Cell Lymphoma International Prognostic
Index (MIPI). Based on the data of 455 patients
Prognostic factors at relapse with advanced stage MCL, the simplified MIPI is
based on four parameters: age, Eastern Cooperative
Histological transformation into DLBCL is by far the
Oncology Group (ECOG) performance status, serum
most important prognostic factor at relapse, and jus-
LDH level, and white blood cell count (Table 2.12).
tifies nodal biopsy whenever possible. Few studies have
Weighting each parameter with 0 to 3 points allowed
tested the prognostic factors at relapse. The FLIPI was
the identification of patients in three risk group (low
tested and showed different OS according to risk
risk: 0–3 points; intermediate risk: 4–5 points; high
group. Sensitivity to initial chemotherapy and delay
risk: >5 points) with significantly different OS. This
between first and second episodes also materially
index has been validated in a few retrospective anal-
affects prognosis.
yses, especially in patients treated with high-dose
therapy.
Mantle cell lymphomas Some authors have advocated the addition of Ki 67
expression in addition to the MIPI since high Ki 67
Prognostic factors at diagnosis expression is a poor prognostic factor. However, con-
cordance between pathologists in evaluating Ki 67
Clinical factors
expression is poor.
Among NHL subtypes, mantle cell lymphomas (MCL)
have the poorest long-term survival, for several
reasons: Histological factors
MCL initially infiltrates the mantle zone surrounding
(1) The median age is around 65 and few patients can normal germinal centers, generating a pattern called
be treated with intensive therapies. “nodular” MCL. Progressively, the germinal centers
(2) Although response rates are high after initial are destroyed by the neoplastic infiltrate, and the
therapy, complete response (CR) rates are low and MCL becomes of the “diffuse” type. In both nodular
most patients have residual disease. and diffuse types, the largest neoplastic compartment
(3) Even in CR patients, the relapse rate is high after a is made of small- to medium-size lymphoid cells,
median duration of response of around 18 months. with few large cells. There is no demonstration of a
(4) After several relapses, most patients have blastic difference in prognosis between the nodular and dif-
transformation, which is highly resistant to all fuse types. MCL is sometimes characterized by
treatments. slightly larger cells with a high mitotic index and is
(5) Anti-CD20 MoAbs are less active than in other therefore termed “blastoid” variant. This latter var-
B-cell NHLs. iant is associated with a more aggressive clinical
Because of the low incidence of the disease and its behavior.
heterogeneity in presentation, there has been no con-
sensus on a clinical prognostic index in MCL patients. Genetic abnormalities
Several retrospective analyses have reported a number The genetic hallmark of MCL is the chromosomal
of adverse prognostic factors. These are referred to in translocation t(11;14) (q13;q32) that can be detected
Table 2.11. The IPI designed for aggressive NHLs has in virtually all cases of MCL. This translocation leads
also been applied to patients with MCL. IPI is poorly to overexpression of cyclin D1, which plays a key role 21
Chapter 2: Prognostic factors for lymphomas
Table 2.11 Clinical prognostic factors in mantle cell lymphomas from retrospective
analyses.
in the control of the G1 phase of the cell cycle. An additional cytogenetic alterations which may contribute
elevated level of cyclin D1 expression in MCL cells to proliferation and poor prognosis. Other abnormalities
accelerates G1/S phase transition and thus tumor cell of the cell cycle regulatory pathway that have been
proliferation. The level of cyclin D1 expression – as reported in MCL are associated with an aggressive
measured by mRNA translation – is directly correlated behavior:
with tumor cell proliferation rate, as measured by Ki
67 expression, and clinical aggressiveness. – deletions of the cyclin D kinase (CDK) inhibitor
Cyclin D1 is a key regulator of the cell cycle, and gene (CDK p16INK4a) on chromosome 9;
22 elevated levels of cyclin D1 in MCL cells promote – gene amplification of CDK4;
Chapter 2: Prognostic factors for lymphomas
– decreased expression of P27KIP1 which is an prognostic value of IPI has also been assessed for
inhibitor of cyclin D/CDK complexes; PTCL. Most of these studies included patients with
– inactivation of P53, mainly observed in blastoid nodal PTCL and found that IPI had predictive value.
MCL types with a high proliferation index. Another model has been developed in a large study
and included some IPI factors (age, LDH, PS) in
These genetic alterations share the fact that they all
addition to BM involvement. PTCL-NOS patients
predominantly target cell cycle regulation and lead to
commonly present with unfavorable characteristics,
an uncontrolled proliferation of MCL lymphoma
including B symptoms, elevated LDH, bulky tumor,
cells. This has been confirmed by gene profile studies,
poor PS, and extranodal disease. The majority of
which have demonstrated that a “proliferation sig-
patients (50%) thus fall into the unfavorable IPI cat-
nature,” i.e. an overexpression of proliferation-
egory. IPI failed to correlate with outcome in patients
associated genes, was a strong predictor of poor
with angioimmunoblastic T-cell lymphoma (AIL) in a
survival. This “proliferation signature” probably
large retrospective study, whereas male sex, mediasti-
reflects the combination of the genetic abnormalities
nal lymphadenopathy, and anemia adversely affected
mentioned above and was found to be associated
overall survival. IPI has also been evaluated in primary
with overexpression of genes involved in drug resist-
nasal natural killer cell lymphoma and was found to be
ance. In conclusion, all MCLs have a poor prognosis
correlated with survival. Application of IPI in the
with conventional treatments. As cell cycle distur-
remaining subtypes has not been evaluated because
bances are the main factor in the development and
of their scarcity, such as hepatosplenic γδ T-cell lym-
clinical course of MCL, abnormalities of genes and
phomas that present almost exclusively with high-risk
proteins of the cell cycle will most probably provide
IPI scores, and it is unlikely that IPI has any clinical
the most useful prognostic information. This contri-
utility in this setting.
bution will be further improved when the biological
tools can be integrated with new therapeutic
approaches. Histological factors
Most PTCLs are predominantly nodal NHLs that
include AIL, PTCL-NOS, and systemic anaplastic
Peripheral T-cell lymphomas large-cell lymphomas (ALCL). Systemic ALCL has a
superior survival compared to other PTCL, whereas
Prognostic factors at diagnosis the prognostic impact of PTCL-NOS compared to AIL
Clinical factors remains controversial, with 5-year overall survival
around 30–35%. It has also been demonstrated that
The importance of T phenotype has been recognized
systemic ALCL were heterogeneous according to
by the Revised European American Lymphoma
expression of ALK protein. ALK-negative cases often
(REAL) and the subsequent updated WHO classifica-
present in elderly patients, with an advanced stage,
tion. These classifications have also allowed us to
elevated serum LDH levels, B-symptoms, and extra-
appreciate fully the wide heterogeneity of peripheral
nodal involvement, and have a prognosis similar to
T-cell lymphoma (PTCL), which accounts for 10–15%
that of other PTCL, whereas ALK-positive cases have a
of all NHLs in western countries. The WHO classifi-
5-year survival close to 90%.
cation divides them into predominantly leukemic,
nodal, and extranodal types. PTCL not otherwise
specified (PTCL-NOS) represents the most common Treatment-related prognostic factors
T-NHL in western countries, accounting for 60–70% Because of the low incidence and the wide heterogeneity
of PTCLs. Because of this heterogeneity and the com- of PTCL, the evaluation of treatment regimens has been
paratively recent histopathological description, retro- limited to non-randomized and retrospective studies. In
spective analyses of PTCL prognostic factors have to addition, most of them included different subtypes of
be analyzed with caution. For example, earlier reports PTCL known to have different outcomes. As in DLBCL,
based on older classifications failed to find clinical reaching a CR at the end of treatment probably
impact of phenotype on outcome, whereas several improves the clinical outcome. However, optimal treat-
studies using the updated classification indicated that ment is not defined here and it is therefore difficult to
T-phenotype has a negative impact on survival. The identify prognostic factors related to treatment. 23
Chapter 2: Prognostic factors for lymphomas
Table 2.13 Therapeutic groups of Burkitt’s lymphoma according to the Société Française d’Oncologie Pédiatrique (SFOP)
and the Berlin–Frankfurt–Münster (BFM) Group.
SFOP BFM
Low-risk group
Stage I Initial complete resection of lymphoma manifestations
Abdominal stage II
Intermediate-risk group
Unresected stage I No or incomplete resection of lymphoma manifestations with at least one of
Non-abdominal stage II the following criteria:
Any stage III or IV or L3ALL – only extra-abdominal sites
(<70% blasts) – abdominal site and LDH <500 UI/L
High-risk group
CNS involvement No or incomplete resection of abdominal lymphoma and LDH ≥500 UI/L
L3ALL (>70% blasts) BM involvement and/or CNS disease and/or multifocal bone involvment
cell proliferation and a tumour mass ≥10 cm have also relapse. Burkitt’s lymphoma refractory to a second
been found to have a poorer prognosis. CNS and BM line of chemotherapy has a low probability of cure,
involvement are recognized to be the most significant whatever the treatment.
adverse prognostic factors and justify an intensified
approach. Using these prognostic factors, different
therapeutic groups have been defined, allowing the Primary central nervous system
development of strategies based upon individual risk lymphoma
(Table 2.13).
Prognostic factors at diagnosis
Treatment-related prognostic factors Clinical factors
Survival of patients with Burkitt’s lymphoma requires Primary central nervous system lymphomas (PCNSL)
patients to enter CR. Stable or progressive disease represent 2% of NHL, making it difficult to identify
(refractory disease) is almost always rapidly fatal. robust clinical prognostic factors. A primary brain
Some other treatment-related prognostic factors have localization is an adverse prognostic factor compared
been demonstrated, especially in children. First, the to localized non-CNSL. Immunodeficiency, whether
absence of complete response after three multiagent related to inherited disorders, to HIV, or to organ
chemotherapy courses correlated with an adverse out- transplantation, increases the risk of PCNSL. Even if
come, and few patients who achieved a partial these patients are not usually included in prospective
response at this time could be salvaged with high- trials, they also seem to have a bad prognosis compared
dose chemotherapy with ASCT. Second, patients to immunocompetent PCNSL. Age and PS are the only
whose tumors did not respond to the prephase regi- two universally accepted prognostic factors for PCNSL.
men (including cyclophosphamide, vincristine, and It has been suggested that they are factors influencing
prednisone) ultimately failed. The importance of therapeutic choice rather than independent survival
early intensification of treatment has also been indicators, but critical reviews of published data and a
shown, with significantly lower EFS if the second more recent large study have confirmed that they are
course of chemotherapy (COPADM2) started after independent variables. The prognostic value of IPI has
day 21. It has also been established that relapse does been tested in some retrospective studies. The IPI is not
not occur after 1 year in CR, and patients without early accurate for PCNSL, which are localized and frequently
relapse or progression can be considered cured. At associated with a poor PS.
relapse, it has been demonstrated that the ability to It is also likely that other specific prognostic factors
cure patients with high-dose chemotherapy followed influence survival. The influence of lymphoma local-
by ASCT was closely related to the chemosensitivity of ization or number of tumors and of CSF protein level 25
Chapter 2: Prognostic factors for lymphomas
has therefore been retrospectively evaluated with of accurate evaluation of these tumors by CT scan and
conflicting results. More recently, predictors of MRI. The impact of PET–18F-FDG in PCNSL evalua-
response and survival were analyzed in an interna- tion has not yet been validated.
tional multicenter retrospective series of 378 immu- PCNSL treatment influences the outcome of
nocompetent patients with PCNSL in an attempt to patients. Prospective randomized studies failed to
develop a prognostic index. More than 90% of patients demonstrate an impact of choice of treatment on sur-
were treated with radiotherapy alone or with com- vival but retrospective analyses have shown that radio-
bined radio-chemotherapy. In the multivariate analy- therapy or chemotherapy alone give lower survival
sis, five factors were identified to have an impact on rates compared to combined chemo-radiotherapy.
survival (Table 2.12): age, PS, LDH serum level, pro- HD-MTX-based chemotherapy has been found to
tein CSF concentration, and involvement of the deep have a significant impact on survival in several studies
structures of the brain. Each variable was assigned a using multivariate analysis. Concomitant intrathecal
value of either 0 if favorable, or 1 if unfavorable, and chemotherapy did not show any impact..
final score was obtained by the addition of the values of
these five variables. The number of adverse features Hodgkin’s lymphoma
was significantly correlated to survival, and prognostic
value of this index was also found in patients treated Prognostic factors at diagnosis
with high-dose methotrexate (HD-MTX)-based che-
motherapy, with or without radiotherapy. Clinical factors
There are numerous studies analyzing the impact of
Histological factors clinical characteristics at diagnosis on outcome in
patients with Hodgkin’s lymphoma (HL). However,
Over 90% of B-PCNSL are DLBCL, Burkitt’s, and low-
most of them included few patients, were heterogeneous
grade NHL, each representing 5%. The T phenotype is
and retrospective, and the prognostic factor models
found in less than 5% of PCNSL. The phenotype and
developed should be considered carefully before they
histological subtypes do not seem to influence survival
are used as a basis for deciding on treatment for an
in PCNSL. BCL-6 and BCL-2 protein expression has
individual patient with HL. Moreover, prognostic factors
been found to have a prognostic influence in DLBCL. In
are closely related to the treatment used and, with the
a recent study including 83 patients with PCNSL, GCB
improvement in outcome for patients with limited-stage
and ABC phenotypes were assessed by TMA; most
HL, these models are no longer clinically relevant. In
patients express an ABC phenotype and have survival
advanced stage HL, prognostic factors can also be helpful
identical to DLBCL expressing the same phenotype.
in identifying patients at high risk of relapse and devel-
oping new therapeutic strategies. All prognostic factors
Treatment-related prognostic factors described are related to tumor spread, tumor burden,
An important difference from NHL located outside patient characteristics, or interaction of tumor and host.
the CNS is that, in the case of PCNSL, obtaining a
CR does not seem to have a positive prognostic Related to tumor spread and tumor burden
impact. In a retrospective study, patients with CR The Ann Arbor classification describes the anatomic
26 and PR did not have statistically different 5-year sur- spread of tumor cells via contiguous lymph node
vivals. However, such results highlighted the difficulty groups and has been demonstrated to be of prognostic
Chapter 2: Prognostic factors for lymphomas
relevance for disease-free survival (DFS) and OS in adverse prognostic factor in many malignant diseases,
many studies testing different therapeutic regimens but increased morbidity and mortality in older
in HL. It is therefore possible to separate “early dis- patients from other causes has to be taken into account
ease” (i.e. stages I/II) which represents two-thirds of and results have to be matched to the general popula-
patients with an estimated 5-year OS close to 90% and tion. Secondly, increased comorbidity leads to
“advanced disease” (i.e. stages III/IV) which represents increased therapy-related mortality. Nevertheless,
one-third of patients, with 5-year OS of around 70%. when matching patients to the general population in
The Ann Arbor classification has been used for more multivariate analyses, age has been shown to have an
than 30 years to influence therapeutic strategies, but adverse prognostic impact, especially in advanced
clinical studies clearly demonstrated differences of stage disease. Gender also seems to influence outcome
survival among patients with same stage, demonstrat- in HL. This could be due in part to the association of
ing the need for additional prognostic factors to adapt female gender with other good parameters, including
the therapeutic strategy to individual risk. localized stage, lack of B symptoms, or nodular scle-
The total body burden count of tumor is an impor- rosis. In a large cohort of patients, multivariate analy-
tant prognostic factor and the presence of bulk disease, sis demonstrated that male gender was independently
especially in the mediastinum, is associated with a poor associated with a reduced OS.
prognosis. Bulky disease is defined as either a single
mass of tumor tissue exceeding 10 cm in the largest Related to interaction of tumor and host
diameter or a mediastinal mass with a ratio of largest B-symptoms (fever, weight loss, and night sweats)
transverse diameter of the mass to the transverse diam- have been known to have clinical relevance for
eter of the thorax at T5–6 of ≥0.35 on a chest radio- many years and have been included in the Ann
graph. This definition has been included in the Ann Arbor classification. They are more frequent in
Arbor classification and is often referred to as the advanced stage disease compared to early stages and
Cotswold modification. Bulky mediastinum, found in correlate with other biological parameters, including
15–20% of patients with early stage disease, correlated increased erythrocyte sedimentation rate (ESR) and
with a reduced DFS after radiotherapy or chemotherapy decreased serum albumin or hemoglobin. Despite
alone as well as after combined treatment, whereas in this, their independent prognostic value has not
advanced-stage HL treated with more aggressive thera- been found in multivariate analysis, but the presence
pies, there are less consistent data on the prognostic of B-symptoms is used by most cooperative groups to
significance of mediastinal bulk. The prognostic value define prognostic groups and inform therapeutic
of bulk disease outside the mediastinum is more uncer- strategies. An elevated ESR was shown to have inde-
tain. Some have calculated tumor burden according to pendent prognostic significance for DFS and OS for
both the number and the size of involved nodes. These different stages and several regimens, especially in
authors demonstrated that, in early stages, tumor bur- early stage disease. The European Organization for
den correlated with both DFS and OS. Other specific Research and Treatment of Cancer (EORTC) lym-
sites of HL, such as BM, spleen, or pleural involvement, phoma group defined a group of patients with either
were shown to be of prognostic significance. Their ESR more than 50 without B-symptoms or ESR more
prognostic value is controversial and seems to have than 30 with B-symptoms who have a worse outcome
only a little impact on survival. The number of organs and an increased rate of relapse. Anemia has also
involved has been correlated with OS and DFS by some been reported to be associated with a decreased sur-
but was not found in larger studies. LDH could be vival in both early and advanced stages. In large
viewed as a surrogate of tumor burden, and an elevated retrospective analyses, lymphopenia lower than
level of LDH has been found to have a prognostic value 0.6 × 109/L was independently associated with
in HL associated with involvement of marrow and at decreased survival. Many other biological para-
least two extranodal sites. meters easily available in peripheral blood, such as
ferritin, C-reactive protein, β2-microglobulin, ceru-
loplasmin, or thymidine kinase, have been described
Related to patient characteristics to influence outcome in HD, but their significance fell
Some patient characteristics, including age and sex, to insignificant levels when multiparametric analysis
are important in prognosis. Age is considered as an was used. 27
Chapter 2: Prognostic factors for lymphomas
28 Marrow involvement
Chapter 2: Prognostic factors for lymphomas
Table 2.16 Prognostic factors for freedom from progression in International Index Patients.
Table 2.17 Prognostic and therapeutic groups defined by the German Hodgkin’s Lymphoma Study Group.
Stage
Risk factors IA, IB, IIA IIB III, IV
None Early stages
≥3 lymph node areas
High ESR* Early unfavorable stages
Extranodal involvement
Massive mediastinal tumor Advanced stages
* Erythrocyte sedimentation rate ≥50 without or ≥30 with B-symptoms.
Table 2.18 Prognostic and therapeutic groups for subdiaphragmatic stages I–II defined by the European
Organization for Research and Treatment of Cancer Group.
Stage
Risk factors IA, IB IIA, IIB
None Favorable stages
≥4 lymph node areas
High ESR
Unfavorable stages
Age ≥50 years
Massive mediastinal tumor
* Erythrocyte sedimentation rate ≥50 without or ≥30 with B-symptoms.
prognostic scores were established from patients depleted subtype was associated with a worse out-
treated before the last decade, and recent data have come. The number of patients suffering from this
demonstrated that they lose predictive power with subtype is very low, and histologic classification of
improved treatment, emphasizing the need for new HL has been reconsidered in the REAL classification,
prognostic factors. separating lymphocyte-predominant HL from classi-
cal HL. Grading according to the number of Hodgkin
Histological factors Reed–Sternberg (HRS) cells or the characteristics of
The prognostic value of the classification established the background infiltrate may predict prognosis in
in 1966 by Lukes and Butler has been analyzed in some settings, but is not required for routine clinical
many studies. They demonstrated that lymphocyte- purposes. 29
Chapter 2: Prognostic factors for lymphomas
Some markers, such as Ki 67, P53, Rb, BCL-2 the impact of interim PET–18F-FDG in patient manage-
or LMP-1, expressed by HRS have been analyzed and ment are now awaited. However, because ASCT is not
correlated with outcome. Such studies have yielded usually used in front line therapy, the predictive value of
conflicting results; they were retrospective and PET–18F-FDG before high-dose chemotherapy and
included too small a number of patients to support ASCT in HL has not been systematically evaluated.
any definitive conclusion. Some groups have focused
on non-malignant cells and demonstrated a negative Prognostic factors at relapse
prognostic impact of CD8 granzyme B-positive T-cells
Many studies have analyzed prognostic factors of
or eosinophilic infiltrates.
patients with refractory or relapsed HL. The length
of remission after first-line chemotherapy has been
Other biological factors shown to be an important prognostic factor. Using
In the past decade, new biological parameters have conventional salvage therapy, virtually no patient
been investigated to predict prognosis. CD30 is con- with primary progressive disease (10% of patients)
sistently expressed by HRS cells and soluble CD30 was survives after 8 years, contrasting with a projected
found to have independent prognostic significance in 20-year survival of 11% and 22%, respectively, for
different studies. Soluble CD30 seemed to correlate patients relapsing early within 12 months of CR
with tumor burden and lymphocyte-depleted HL. (15% of patients) and for patients relapsing more
VCAM-1 and ICAM-1 are adhesion molecules and than 12 months after CR (15% of patients).
soluble forms could have prognostic significance.
Several cytokines, such as IL-2, IL-6, IL-7, IL-8, IL-
10, IL-12, IL-13, and TNFα, secreted by both HRS cells Further reading
and lymphoid cells surrounding HRS cells, have also Arcaini L, Lazzarino M, Colombo N, et al. Splenic marginal
been evaluated but their prognostic value remains to zone lymphoma: a prognostic model for clinical use.
be defined with any certainty. Blood, 2006;107(12):4643–4649.
Cheson BD, Pfistner B, Juweid ME, et al. Revised response
criteria for malignant lymphoma. J Clin Oncol,
Treatment-related prognostic factors 2007;25(5):579–586.
Dose intensity and time to complete remission Dave SS, Wright G, Tan B, et al. Prediction of survival in
Survival of patients with HL is closely related follicular lymphoma based on molecular features of
tumor-infiltrating immune cells. N Engl J Med,
with achieving CR at the end of treatment, and 2004;351:2159–2169.
stable or progressive disease (refractory disease)
Federico M, Bellei M, Marcheselli L, et al. Follicular
are both associated with inferior survival. In a
lymphoma international prognostic index 2: a new
retrospective analysis, the dose delivered in the prognostic index for follicular lymphoma developed by the
first three cycles of chemotherapy was the main international follicular lymphoma prognostic factor
factor predicting CR and outcome, demonstrating project. J Clin Oncol, 2009;27:4555–4562.
the importance of dose intensity in HL treatment. Hasenclever D, Diehl V. A prognostic score for advanced
Moreover, early CR, usually evaluated after three or Hodgkin’s disease. International Prognostic Factors
four cycles, was also shown to be predictive of a Project on Advanced Hodgkin’s disease. N Engl J Med,
better outcome. 1998;339:1506–1514.
Hoster E, Dreyling M, Klapper W, et al. A new prognostic
Positron emission tomography (PET) using index (MIPI) for patients with advanced stage mantle cell
18
F-fluorodeoxyglucose (18FDG) lymphoma. Blood, 2008;111:558–565.
Most studies demonstrating PET–18F-FDG prognostic Lenz G, Wright G, Dave SS, et al. Stromal gene signatures in
large-B-cell lymphomas. N Engl J Med,
value for post-treatment evaluation included both 2008;359:2313–2323.
DLBCL and HL, and in this situation negative predictive
Patte C, Auperin A, Michon J, et al. The Societe Francaise
value seemed to be higher in HL than in HD. Early
d’Oncologie Pediatrique LMB89 protocol: highly
prognostic value of PET–18F-FDG has also been dem- effective multiagent chemotherapy tailored to the tumor
onstrated in early response assessment and correlated burden and initial response in 561 unselected children
30 independently with PFS and OS, especially for with B-cell lymphomas and L3 leukemia. Blood,
advanced-stage patients. Prospective studies evaluating 2001;97:3370–3379.
Chapter 2: Prognostic factors for lymphomas
Philip T, Biron P, Maraninchi D, et al. Role of massive chemotherapy in non-Hodgkin’s lymphoma: is [18F]
chemotherapy and autologous bone-marrow FDG-PET a valid alternative to conventional diagnostic
transplantation in non-Hodgkin’s malignant lymphoma. methods? J Clin Oncol, 2001;19:414–419.
Lancet, 1984;i:391. Spaepen K, Stroobants S, Dupont P, et al. Prognostic value of
Rosenwald A, Wright G, Chan WC, et al. The molecular use pretransplantation positron emission tomography using
of profiling to predict survival after chemotherapy for fluorine 18-fluorodeoxyglucose in patients with
diffuse large-B cell lymphoma. N Engl J Med, aggressive lymphoma treated with high-dose
2002;346:1937–1947. chemotherapy and stem cell transplantation. Blood,
Sehn LH, Berry B, Chhanabhai B, et al. The Revised 2003;102:53–59.
International Prognosis Index (R-IPI) is a better Swerdlow SH, Campo E, Harris NL (eds). WHO
predictor factor of outcome than the standard IPI for Classification of Tumours of Hematopietic and Lymphoid
patients with diffuse large B-cell lymphoma treated with Tissues. Lyon: IARC press, 2008.
R-CHOP. Blood, 2007;109:1857–1861. The International Non-Hodgkin’s Lymphoma Prognostic
Solal-Celigny P, Roy P, Colombat P, et al. Follicular Factors Project. A predictive model for aggressive non-
lymphoma international prognostic index. Blood, Hodgkin’s lymphoma. N Engl J Med, 1993;329:987–992.
2004;104:1258–1265. Thieblemont C, Bastion Y, Berger F, et al. Mucosa-associated
Spaepen K, Stroobants S, Dupont P, et al. Prognostic value of lymphoid tissue gastrointestinal and nongastrointestinal
positron emission tomography (PET) with fluorine-18 lymphoma behavior: analysis of 108 patients. J Clin
fluorodeoxyglucose ([18F]FDG) after first-line Oncol, 15:1624–1630.
31
Chapter
3
Imaging
Heok K. Cheow and Ashley S. Shaw
32 Lymphoma, Second Edition, ed. Robert Marcus, John W. Sweetenham, and Michael E. Williams. Published by
Cambridge University Press. © Cambridge University Press 2014.
Chapter 3: Imaging
Figure 3.5 Ill-defined increase in opacification of the small bowel Figure 3.6 Multiple low attenuation lesions throughout the liver
mesentery (misty mesentery) in a patient with follicular lymphoma. and spleen were demonstrated by CT with small bilateral pleural
effusions. Biopsy demonstrated this to be due to diffuse large B-cell
lymphoma.
Figure 3.7 Marked thickening of the small bowel loops within the
pelvis resulting from small bowel lymphoma.
Figure 3.8 Lytic lesions are demonstrated throughout the left ilium
with destruction of the cortex and an associated soft tissue mass,
contradictory. Using a node-by-node analysis with sur- proven to be DLBCL at biopsy.
gical correlation, one study found CT to have a sensi-
tivity of only 20%, correctly staging only 42% of patients Extranodal sites of disease can also be difficult to
with Hodgkin’s lymphoma (HL). Conversely, other identify with CT. Gastrointestinal wall thickness is
workers have reported CT to have 92% sensitivity in difficult to estimate, particularly in the non-distended
staging HL and 75% sensitivity in staging NHL. The stomach, and will only be visible when gross
majority of published work lies between these results, (Figure 3.7). In the stomach, an interrupted rugal
suggesting an accuracy of approximately 70%. pattern and focal thinning may provide the only clue.
Splenic involvement may be manifest by either Bone marrow disease will only be evident when a lytic
splenomegaly, for which there are many causes, or or sclerotic lesion has developed or when there is
the presence of focal lesions (Figure 3.6). Depending extension into the adjacent soft tissues (Figure 3.8).
on the size and number of lesions, together with the Conversely, pulmonary disease is usually readily appa-
technical parameters relating to intravenous enhance- rent but can be very difficult to differentiate from
34 ment, these may easily be overlooked and/or difficult other causes of nodules and consolidation, particularly
to characterize with any confidence. infection and drug toxicity (Figures 3.9 and 3.10).
Chapter 3: Imaging
Figure 3.9 Multiple pulmonary nodules bilaterally measuring up to Figure 3.10 Consolidation in the right upper lobe with an air
1 cm in diameter. Biopsy proved these to be DLBCL. bronchogram demonstrated at CT. Biopsy proved this to be mantle
cell lymphoma.
B
Figure 3.11 (B) Axial images through the lesion demonstrate the
extradural mass extending through the neural foramina bilaterally.
A
Similarly, imaging focal deposits in the limbs with MRI
Figure 3.11 (A) Sagittal T2-weighted MRI of the thoracic spine
demonstrating an extradural mass posterior to, and compressing,
can provide additional information as it offers superior
the spinal cord. contrast between different soft tissues over CT. MRI of
the bone marrow may be useful in identifying patients
Magnetic resonance imaging (MRI) is primarily with focal lymphoma deposits, particularly when bone
used to image the central nervous system (CNS), soft marrow biopsy is normal. Lymphoma deposits are seen
tissue masses, and bone marrow. MRI has a greater as low signal on T1 and high signal on fat-saturated
sensitivity than CT in identifying CNS lesions and pro- T2-weighted images as the normal fatty marrow is dis-
vides excellent multiplanar images (Figures 3.11 and placed, often with disease spilling out into the extradural 35
3.12), which may help in radiotherapy planning. or paravertebral spaces. The identification of focal
Chapter 3: Imaging
A B
Figure 3.12 (A) Axial T1-weighted image following intravenous Figure 3.12 (B) Coronal T1-weighted image following intravenous
gadolinium demonstrates a uniformly enhancing mass within the gadolinium demonstrates the lesion extending up into the right
pons, extending posteriorly into the right middle cerebellar midbrain toward the cerebral peduncle. Image courtesy of Dr J. Cross.
peduncle. Image courtesy of Dr J. Cross.
HL
Nodular sclerosis type High
Mixed cellularity type Moderate to high
Lymphocyte-depleted type Moderate to high
Lymphocyte-predominant type Low
B-cell NHL
Diffuse large B-cell lymphoma High
Burkitt lymphoma High
Large cell and anaplastic lymphoma High
Follicular lymphoma (grade 3) Moderate to high
Follicular lymphoma (grade 1–2) Low to moderate
Mantle cell lymphoma Low to moderate
Marginal zone (inc. MALT) lymphoma None to high
Small lymphocytic lymphoma None to low
T-cell lymphoma
Extranodal NK/ T-cell lymphoma High
Peripheral T-cell lymphoma High
Adult T-cell leukemia–lymphoma High
Cutaneous T-cell lymphoma Moderate
Mycosis fungoides and Sézary syndrome Low
However, in high-grade disease it is in the order of a involved field radiotherapy, although the use of radia-
few millimeters. In this respect, PET–CT may be tion may alter when results of prospective studies
particularly useful in evaluating pulmonary disease, utilizing the results of FDG–PET become available.
as tiny nodules may be beyond the resolution of PET The presence of disease below the diaphragm will
alone. As the evidence base grows and PET–CT tech- increase the number of planned cycles of chemother-
nology becomes more widely available, it will become apy. Additionally, the phrase “bulky lymph node
the imaging modality of choice for staging and man- mass” in a radiological report should be reserved for
aging high-grade lymphoma. those lesions ≥10 cm in diameter, to avoid confusion.
Novel PET tracers are under development and However, in indolent follicular lymphoma, a
have been used in early phase clinical studies for watch-and-wait policy may be adopted for many
assessing tumor characteristics such as hypoxia, apop- patients, even with advanced stage disease. There
tosis, and proliferation. One example is 3-deoxy- are a number of radiological indications for chemo-
3–18F-fluorothymidine (18FLT), a thymidine analog therapy, including nodal or extranodal mass ≥7 cm in
that demonstrates tumor proliferation. It has been diameter, ≥3 nodal sites with a diameter ≥3 cm,
used in assessing early treatment response in lym- splenic enlargement, compression syndrome, pleural
phoma, but it is still too early to see whether it can or pericardial effusion, and the radiologist must be
replace 18FDG in lymphoma imaging. Future imaging aware of the potential therapeutic implications of the
may involve an array of tracers tailored to answer report.
specific clinical questions. For childhood NHL and adult patients with
Burkitt’s lymphoma, the St Jude classification
Staging classifications (Table 3.3) is employed. This differs significantly
from the Ann Arbor system in the way that extrano-
The Ann Arbor staging classification with Cotswold
dal disease is staged, reflecting the different patterns
revision (Table 3.2) is used to provide the radiologic
of disease and the prognostic implications for these
staging of HD and most types of NHL. It is important
two groups of patients.
for the radiologist to understand the impact of staging
on management so that the report is tailored appro-
priately. For example, in HL, early stage asymptomatic Response assessment
38 supradiaphragmatic disease is currently treated by a The International Workshop Criteria (IWC), pub-
combination of short-course chemotherapy and lished in 1999, have become the widely accepted
Chapter 3: Imaging
Table 3.2 Ann Arbor staging classification and Cotswold Table 3.3 St Jude staging classification.
revision.
I Single tumor (extranodal)
Stage Area of involvement Single anatomic area (nodal)
Excluding mediastinum or abdomen
I A single lymph node region or a
single lymph localized II Single tumor (extranodal) with regional node
involvement of an extralymphatic involvement
site Two single (extranodal) tumors without regional
lymph node involvement on the same side of the
IE Localized involvement of a single diaphragm
extralymphatic organ or site Two or more nodal areas on the same side of the
II Two or more lymph node regions diaphragm
on the same side of the Primary gastrointestinal tumor with or without
diaphragm involvement of associated mesenteric nodes only,
grossly completely resected
IIE Localized involvement of a single
extralymphatic organ or site and III Two single (extranodal) tumors, one on either side of
of one or more lymph node the diaphragm
regions on the same side of the Two or more nodal areas, involving above and below
diaphragm the diaphragm
Primary intrathoracic tumors (mediastinal, pleural,
III Lymph node regions on both thymic)
sides of the diaphragm Extensive primary intra-abdominal disease
IIIE Lymph node regions on both Paraspinal or epidural tumors irrespective of other sites
sides of the diaphragm IV Any of the above with initial CNS or bone marrow
accompanied by localized involvement
involvement of an extralymphatic
organ or site
IV Diffuse involvement of one or
more extranodal organs with or Table 3.4 International Workshop Criteria.
without lymph node involvement
Complete response (CR): complete disappearance of all
Localized involvement of a single detectable disease by imaging, with nodes previously >1.5 cm
extralymphatic organ or site with regressing to <1.5 cm and nodes of 1.0–1.5 cm to <1.0 cm. In
non-regional lymph node addition, resolution of disease-related symptoms,
involvement normalization of biochemical abnormalities, and normal bone
Additional qualifiers marrow biopsy.
A Absence of systemic symptoms Complete response unconfirmed (CRu): as for CR, but with a
residual mass >1.5 cm in diameter that has regressed by >75%.
B Presence of systemic symptoms
Partial response (PR): at least 50% reduction in the sum of the
X Bulky disease, defined as a nodal product of the greatest diameters (SPD) of the six largest nodes.
mass >10 cm in maximum There should be neither increase in the size of other nodes nor
diameter or mediastinal mass any new sites of disease. Hepatic and splenic lesions should also
>one-third of the internal reduce >50% in the SPD.
diameter of the thorax on a PA
chest radiograph Stable disease (SD): where response is less than a PR but there
is not progressive disease.
Progressive disease (PD): more than 50% increase in the
product of the diameters in any previously abnormal node or
the development of new disease sites either during or at the
standard for the assessment of disease response in end of therapy.
NHL. These have enabled comparability of clinical Relapsed disease (RD): the appearance of any new disease or
trials as well as the day-to-day management of an increase in size of over 50% of residual lesions in patients
who had previously achieved CR or CRu.
patients. Although based primarily on CT findings,
the IWC take bone marrow biopsy, clinical, and bio-
chemical information into account and are outlined in
Table 3.4. However, there are a number of drawbacks will be positive for lymphoma on re-biopsy.
with the IWC criteria that need to be addressed. Moreover, if a patient has been imaged too soon
Following treatment for lymphoma, up to 60% of following treatment, there may have been insufficient
patients with nodal disease will have a residual mass time for any significant volume reduction. As
outlined above, whilst CT provides excellent
39
on CT. Of these, only a small proportion (up to 20%)
Chapter 3: Imaging
Table 3.5 IWC + PET description. Table 3.6 Manifestations of pulmonary drug toxicity in
lymphoma.
CR CR by IWC with a completely negative PET
CRu, PR or SD by IWC with a completely negative PET Diffuse alveolar Bleomycin, carmustine,
and a negative bone marrow biopsy (BMB) if damage cyclophosphamide
positive prior to therapy
PD by IWC with a completely negative PET and CT Organizing pneumonia Bleomycin, cyclophosphamide
abnormalities ≥1.5 cm (≥1.0 cm in the lungs) and Pulmonary Cyclophosphamide, cytarabine,
negative BMB if positive prior to therapy hemorrhage amphotericin B
CRu CRu by IWC with a completely negative PET but with Non-specific interstitial Chlorambucil, carmustine
an indeterminate BMB pneumonia
PR CR, CRu, or PR by IWC with a positive PET at the site
of a previously involved node/nodal mass
CR, CRu, PR, or SD by IWC with a positive PET outside
the site of a previously involved node/nodal mass
SD by IWC with a positive PET at the site of a Complications of therapy
previously involved node/nodal mass that Patients with lymphoma may be treated with an array
regressed to <1.5 cm if previously >1.5 cm, or
<1.0 cm if previously 1.1–1.5 cm of cytotoxic agents and radiation therapy, which can
SD SD by IWC with a positive PET at the site of a
themselves result in significant morbidity and occa-
previously involved node/nodal mass (i.e. residual sionally death. The lungs appear to be particularly
mass) sensitive to these insults and this can manifest itself
PD PD by IWC with a positive PET finding corresponding in a number of ways, the more common of which are
to the CT abnormality (new lesion, increasing size listed in Table 3.6.
of lesion)
PD by IWC with a negative PET and a CT abnormality Diffuse alveolar damage (DAD) is the histopatho-
(new lesion, increasing size of lesion) of <1.5 cm logical pattern that results from necrosis of type II
(1.0 cm in the lungs) pneumocytes and alveolar endothelial cells. The clin-
ical correlate of this is the acute respiratory distress
syndrome (ARDS). In the acute phase (first week after
insult), there is interstitial and alveolar edema and
hyaline membranes. Following this is a reparative
phase, with proliferation of type II pneumocytes and
interstitial fibrosis (usually at 1–2 weeks). The chest
radiograph may demonstrate bilateral focal areas of
consolidation, often in the mid and lower zones. At
this stage, CT often demonstrates a combination of
consolidation, ground glass opacification, and fibrosis
(Figure 3.17). In the chronic phase, fibrosis may
improve, remain stable, or progress. The fibrotic
changes depicted at CT are predominantly peripheral
and subpleural. A number of drugs have been impli-
cated, with cyclophosphamide, carmustine, and bleo-
mycin being the commonest in the context of
lymphoma.
Figure 3.16 Axial CT and PET images obtained before and after Cyclophosphamide-related injury is not dose-
chemotherapy demonstrate an enlarged right cervical node that has
reduced in size to <1 cm following chemotherapy. By IWC, this related and has been reported up to 13 years after
represents complete response (CR). However, the PET study shows administration. Conversely, carmustine has a clear
residual high tracer activity seen in this region, indicating active dose-related toxicity, with the overall incidence rising
residual disease.
from 20–30% to 50% once a cumulative dose of 1.5 g/
m2 has been exceeded. Bleomycin pulmonary toxicity
completion of radiotherapy. The use of PET to eval- is reported to occur in up to 18% of patients, with a
uate residual lesions in indolent disease is less clear and mortality of 4.2%. There appear to be a number of
studies should be interpreted with caution, particu- risk factors associated, including age >40 years, cumu-
larly if PET was not used during initial staging. lative dose >450 units, and use of granulocyte 41
Chapter 3: Imaging
Figure 3.17 Extensive ground glass opacification bilaterally with Figure 3.18 Multiple foci of consolidation, some with a rim of
areas of consolidation in the left lower lobe. A presumptive diagnosis ground glass opacification typical of angioinvasive aspergillosis.
of bleomycin toxicity was made, with the patient responding to drug
withdrawal and a course of corticosteroids.
A B
Figure 3.19 (A) A patient with Hodgkin disease and HIV presented Figure 3.19 (B) Eleven days later, the patient deteriorated
with acute dyspnea. Initial CT demonstrates multiple foci of ground further and CT demonstrated a pneumomediastinum and left
glass opacification and bilateral pleural effusions. This was proven to pneumothorax. The ground glass opacification within the lung
be PCP following bronchoscopy. parenchyma is now more confluent and the septal thickening is
more evident.
Juweid ME, Wiseman GA, Vose JM, et al. Response lymphoma and Hodgkin disease. Radiographics,
assessment of aggressive non-Hodgkin’s lymphoma by 2010;30:269–291.
integrated International Workshop Criteria and fluorine- Rimmer MJ, Dixon AK, Flower CDR, Sikora K. Bleomycin
18-fluorodeoxyglucose positron emission tomography. J lung: computed tomographic observations. Br J Radiol,
Clin Oncol, 2005;23:4652–4661. 1985;58:1041–1045.
Kazama T, Faria SC, Varavithya V, et al. FDG PET in the Rossi SE, Erasmus JJ, McAdams HP, Sporn TA,
evaluation of treatment for lymphoma: clinical usefulness Goodman PC. Pulmonary drug toxicity: radiologic and
and pitfalls. Radiographics, 2005;25:191–207. pathologic manifestations. Radiographics,
Lister TA, Crowther D, Sutcliffe SB, et al. Report of a 2000;20:1245–1259.
committee convened to discuss the evaluation and Surbone A, Longo DL, DeVita VT Jr, et al. Residual
staging of patients with Hodgkin’s disease: Cotswolds abdominal masses in aggressive non-Hodgkin’s
meeting. J Clin Oncol, 1989;7:1630–6. Erratum in: J Clin lymphoma after combination chemotherapy:
Oncol 1990;8:1602. significance and management. J Clin Oncol,
Meyers JL. Pathology of drug-induced lung disease. In 1988;6:1832–1837.
Katzenstein AA, Askin FB, (eds). Katzenstein and Askin’s Terasawa T, Lau, J, Bardet S, et al. Fluorine-18-
Surgical Pathology of Non-neoplastic Lung Disease, 3rd fluorodeoxyglucose positron emission tomography
edn. Philadelphia, PA: Saunders, 1997;4:81–111. for interim response assessment of advanced stage
Moon JH, Kim EA, Lee KS, et al. Cytomegalovirus Hodgkin’s lymphoma and diffuse large B-cell
pneumonia: high-resolution CT findings in ten lymphoma: a systematic review. J Clin Oncol,
non-AIDS immunocompromised patients. Korean J 2009;27:1906–1914.
Radiol, 2000;1:73–78. Zissin R, Metser U, Hain D, Even-Sapir E. Mesenteric
Paes FM, Kalkanis DG, Sideras PA, Serafini AN. FDG PET/ panniculitis in oncologic patients: PET-CT findings.
CT of extranodal involvement in non-Hodgkin Br J Radiol, 2006;79:37–43.
44
Chapter
4
Clinical trials in lymphoma
Thomas M. Habermann and Matthew Maurer
Introduction
Clinical trials are the backbone of the development and was that of Galen, who attained an authority that
advancement of therapeutic approaches. The pace of remained unchallenged. Medicine had a long tradi-
development of new agents, the regulatory overhead, tion of physicians as trusted advisors to patients.
the costs of data management, inclusion of quality-of- Treatment choices were based not on investigation
life assessments, radiologic assessments and central but rather on uncontrolled experimentation.
review, central pathology review, biospecimen acquis- Therapies were primarily evaluated on individual
ition, symptom-control assessments, and the evaluation patients in an ad hoc manner. One of the earliest
of biologic correlates all present significant challenges and most commonly cited examples of clinical
for the future development of new therapeutic agents experiments to evaluate therapeutic interventions
and regimens. As these therapeutic agents and was the treatment of scurvy with oranges and lemons.
approaches become increasingly targeted and the basic The comparison of results from these trials was pri-
science, clinical risk factors, toxicities, and natural his- marily to those in the literature and other historical
tory of lymphomas broadens, the identification controls. The first strictly controlled clinical trial
and assessment of relevant endpoints, be they clinical through random assignment of patients to treatment
(e.g. physical examination, radiologic) or biologic (e.g. groups was not reported until 1931, in the examina-
immunologic, genetic, metabolic, etc.), also becomes tion of tuberculosis patients. A subsequent trial in
increasingly complex. In turn, the more complex the tuberculosis patients was also the first to use random
endpoints and the more targeted the regimens, the numbers to assign patients to experimental versus
more challenges are presented when designing and control groups, and it was demonstrated that strep-
conducting clinical trials and analyzing and interpreting tomycin plus bed rest was superior to bed rest
the results. While this field of clinical trial design is too alone. Since then, the practice and development of
broad to give full discussion adequately in this chapter, clinical trials has expanded greatly. In the United
the reader is referred to more extensive references on States, controlled clinical trials were first under the
this area. Given the complexity of issues when design- National Cancer Institute (NCI) under Gordon
ing, monitoring, interpreting, and analyzing data for a Zubrod. Zubrod was one of the key individuals in
clinical trial, statistician input and collaboration is of the formation of the Eastern Solid Tumor Group
paramount importance. This chapter will focus on the (now the Eastern Cooperative Oncology Group,
principles and details of clinical trials from the clinician ECOG), which published the first randomized trial
perspective. As such, the fundamental considerations of in solid malignancies in the USA comparing nitrogen
clinical trial design as well as the types of trials con- mustard and thiophosphamide.
ducted in clinical research will be reviewed. In today’s research environment, a primary focus
in the development and conduct of clinical trials is the
safety and protection of the patient involved in such
Background research. This is of primary importance as we continue
The practice and science of clinical trials and research to refine and explore new methods for clinical trial
is in its relative infancy. Until 1750, the thinking design and analysis.
Lymphoma, Second Edition, ed. Robert Marcus, John W. Sweetenham, and Michael E. Williams. Published by 45
Cambridge University Press. © Cambridge University Press 2014.
Chapter 4: Clinical trials in lymphoma
modulators of signal transduction pathways, angiogen- depending on the endpoints and phase of the trial.
esis inhibitors, growth factor receptor inhibitors, poten- These measures are on qualitative (categorical), quan-
tiators of apoptosis, modulators of the immune titative (measurement), or time-to-event data. An
response, and modulators of gene expression. They example of a qualitative measure is the classification
may include cellular therapy, gene therapy, and vaccine of response, where tumor measurements define com-
therapy, and the application of clinical trials to these plete response (CR), partial response (PR), stable dis-
interventions presents new challenges and opportuni- ease (SD), and progressive disease (PD). For those
ties. With cytostatic agents, the standard endpoint of refractory and/or aggressive patient populations, the
clinical response may not be sufficient to determine overall response rate may be of interest. For newly
efficacy. Instead, the proportion of patients who have diagnosed and/or more indolent patient populations
not progressed or relapsed within a specified timeframe where the overall response (CR and PR) is typically
may be used, or other endpoints more appropriate to high with standard treatment and thus where a com-
capture the effects of a cytostatic agent. In addition, plete response is more interesting, the CR rate may
evaluation of immunotherapeutic regimens may require instead be used as a primary measure of efficacy.
similar alternate endpoints to evaluating efficacy. The Quantitative endpoints focus on continuous meas-
timing of the study, initial therapy versus treatment in ures of effect. Examples of this are outcomes such as
relapse, and histology subtype are often the most rele- changes or percentages of changes in quality-of-life
vant issues in determining the endpoints. scores, immunologic parameters, or other biologic
data that can measure clinical effect or adverse reac-
tions. Time-to-event data can also be defined in many
Response criteria different ways, depending on the needs of the trial, and
As histology has been further refined, response criteria it is important to define at the outset, not retrospec-
have been standardized. The Modified Cheson criteria tively, what constitutes an event as well as the time from
incorporate positron emission tomography (PET) scans which this is measured (e.g. study entry and diagnosis
into diffuse large B-cell lymphoma (DLBCL) and need to be defined). Time-to-event measures include
Hodgkin lymphoma (HL). OS is defined as death from EFS (i.e. time to an event defined as progression, relapse,
any cause from the time of randomization. Lymphoma- and/or death), OS (i.e. time to death from any cause),
specific survival is defined from the time of entry to the and even time to engraftment in stem cell transplanta-
study to death as a result of lymphoma or death of tion. Time to event measures must take into account the
unknown causes attributed to therapy. Progression fact that not all patients will have had an event at the time
base endpoints included the following: PFS is defined of analysis, but, having followed the patient for a certain
as lymphoma progression or death from any cause. time, there is the evidence that they are event-free at least
However, in studies in which failure to respond without until that time. Statistical methods are used in the esti-
progression is considered an indication for another mation and analysis of this type of data to account for
therapy, such patients should be censored at that point some patients not reaching defined endpoints.
for the progression analysis. Event-free survival (EFS) is Overall, outcome data can take many different
defined as time from study entry to any treatment fail- forms to accommodate the endpoints and goals of
ure and defined as disease progression or discontinua- the clinical trial. What is important is to define these
tion of therapy for any reason (progression, toxicity, endpoints adequately when designing the trial and to
refusal, or initiation of new treatment). Time to pro- consider carefully how effectively these outcome data
gression (TTP) is defined as documented lymphoma can be collected on all patients who will enroll onto the
progression or death due to lymphoma. Non-lym- trial. This is a more significant logistical concern in the
phoma deaths are censored at the time of death. context of multi-institution clinical trials, where
standards and care may differ and multiple pieces of
information are being collected.
Outcome data
After determining the primary and secondary end-
points of interest for the clinical trial, it is important Evaluating treatment effect
to identify those variables and resulting outcome data In assessing and determining efficacy in clinical trials,
necessary to answer the relevant research questions at the classification of the treatment effect on patients is 47
the end of the trial. Outcome measures will vary critical for not only identifying the clinical response
Chapter 4: Clinical trials in lymphoma
rate, but also in defining progression and thereby time after the initial pre-treatment PET to assess response
to progression or PFS. The definition of response is and predict outcome.
disease- and even subtype-dependent and needs to be
appropriate for the study. These definitions also evolve
over time as new technologies develop to evaluate the Precision and error control
existence of and changes in the disease. The addition of Two key aspects related to the precision of clinical trial
computerized axial tomography (CT scan) changed design are interval estimates of the primary endpoint,
how lymphoma was assessed and measured, and func- referred to as confidence intervals (CI), and the simi-
tional imaging – such as PET scans – is now replacing larity of the study sample to the population to which the
and/or complementing the CT scan in lymphoma. results apply. A CI is a point estimate ± a margin of
PET/CT fusion scans are more readily available; how- error that depends on the degree of confidence one
ever, the use of oral and intravenous contrast agents desires to place on the true value being within the
that are routinely utilized in CT scanning impairs the interval. This is calculated from the data that will
quality of the PET scans. include the true parameter of interest (e.g. response
In 1999, an International Working Group (IWG) rate, mean change in quality of life) for the population
developed recommendations for response assessment with a specified probability. Calculation of the confi-
for non-Hodgkin lymphoma (NHL) that were adopted dence interval is dependent on the primary endpoint
internationally by study groups, industry, regulatory and corresponding outcome measure, but is based on
agencies, and, subsequently, by clinical trial groups for the statistic (e.g. proportion, mean), a margin of error
HL. Despite the publication of standardization of that is calculated using a cut-off value for the corre-
response criteria for NHL, an examination of current sponding level of confidence based on the distribution
trial parameters in nine international lymphoma study of the statistics, as well as the standard error (i.e. stand-
groups by the working group for quality management ard deviation of the statistic) calculated from the sample
(WG-QM) of the Competence Network Malignant of patients enrolled on the clinical trial. Typically, 95%
Lymphoma found significant differences and missing CI are calculated, but other confidence levels are also
definitions. This finding led to an international project useful. These intervals may be obtained from tables
to harmonize trial parameters for malignant lym- designed for the binomial distribution. There is a
phoma. Revised response criteria for malignant lym- trade-off between level of confidence and the width of
phomas from the members of the International the interval, where the wider the interval, the higher the
Harmonization Project (IHP) of the Competence confidence level that the true population parameter is in
Network Malignant Lymphoma were first reported at that interval. Another determinant of CI width (and
the American Society of Hematology in December thereby precision of the estimate) is sample size; the
2005. These response criteria incorporated functional larger the sample size, the greater the precision and thus
imaging via the PET scan. Pre-treatment PET was the narrower the CI. It is important to note that, in the
recommended but not required in DLBCL, HL, follic- context of clinical trials, multistage designs can require
ular lymphoma (FL), and mantle cell lymphoma different methods of calculating CI to avoid bias.
(MCL). Response assessment was recommended for An essential aspect of precision is the similarity of
DLBCL and HL and in FL and MCL only when overall the patients eligible for the study versus the target
response rate (ORR)/CR was a primary endpoint of population of interest. The eligibility criteria deter-
the study. Routine post-treatment follow-up was not mine the study population, which is typically a subset
recommended. A CR was defined as all lesions nega- of the target population. A trade-off exists based on
tive on the PET scan, a normal lactate dehydrogenase restrictive versus broad criteria. The advantages of
(LDH), and negative flow cytometry. Complete remis- restrictive eligibility criteria are that the patients are
sion undefined (CRu) was no longer a response clas- more homogeneous, smaller sample sizes may be
sification. Relapse should include PET-positive results. required, and there is a reduced likelihood of con-
By 2011, multiple groups internationally have been founding factors. Disadvantages are that the results
working to further define when PET scans should be may not be able to be generalized to the target pop-
performed, how to incorporate the maximal standard- ulation, and fewer patients may be eligible, which may
ized uptake value (SUVmax), to which types of lym- translate to a longer study duration or non-feasibility
48 phoma PET is most applicable, and how to utilize PET of the enrollment goal. In general, eligibility criteria
Chapter 4: Clinical trials in lymphoma
should be loosened as the phase of the trial increases defined the lymphoma populations of aggressive lym-
and trials become larger. Phase I trials are usually phomas. Subsequent studies have demonstrated the
more restrictive, because patients should not be utility of the IPI in other lymphoma histologies, such
enrolled in a dose-finding trial with a single agent if as peripheral T-cell lymphoma, not otherwise speci-
there is an alternative treatment that is known to be fied. Modifications of the IPI have been demonstrated
effective. Such trials may span multiple histologies if to have value in virtually all NHLs. Initially adopted in
tolerability is the key determinant of dose level, but the the FLs, an adaptation, the FLIPI score and modified
criteria can be restrictive in terms of who would be FLIPI score, have replaced the IPI in FL and better
eligible to receive this experimental regimen. Phase II differentiate groups of patients.
trials are often conducted in the target population of Closely related to the issue of precision is that of
interest, but are often more restricted to ensure a base- error rates and how to control these errors. In clinical
line level of health and some homogeneity, to better trials there are two types of error rates to control: type I
evaluate initial evidence of activity. This is an impor- and type II errors. A type I error is defined as the
tant reason why results from phase II trials should not probability of concluding that a treatment effect or
be used to change clinical practice. Phase III studies difference exists when in truth it does not. The type
are typically the most liberal in terms of eligibility II error is the probability of concluding that a treat-
criteria as these are usually large trials that are ment effect or difference does not exist when in truth it
intended to evaluate regimens against standard of does. Both types of errors should be controlled in most
care, and their results can affect clinical practice. clinical trials. The level of acceptable error rates will
Because of this, it is critical to design the eligibility depend on the phase of the trial. The closer the study is
criteria to ensure that patients entered into the clinical to results that will be generalized to the target popula-
trial are similar to, or representative of, the target tion, the stricter the acceptable levels are for the rates
population so that the results can be generalized to of type I and type II errors. Typically, type I and type II
the patient population of interest. error rates are constrained at 10% each or even 5% type
In lymphoma, histology is the most important I and 20% type II in the phase II setting. However, in
eligibility criterion. In 1982, the non-Hodgkin’s the phase III setting the typical constraints are for 5%
Lymphoma Pathologic Classification Project proposed type I and 10% type II error rates. These predefined
a morphologic system that was representative of constraints on error rates affect the resulting sample
at least six classification systems. The Working size for the clinical trial, and by definition affect the
Formulation had three major subdivisions: low power of the study. Power is defined as the probability
grade, intermediate grade, and high grade. A classifi- of concluding that a treatment effect or difference
cation that was based on immunohistochemistry, exists when it truly does exist, and is calculated as 1 –
cytogenetics, and molecular genetic characteristics type II error rate (e.g. type II error rate of 0.10 trans-
was developed, the Revised European–American lates to 90% power). The constraint on the type I error
Lymphoma classification (REAL). The World Health rate also affects the decision rule to be used when
Organization (WHO) classification, fourth edition, is determining if the results are significant. This signifi-
now the accepted classification system internationally. cance level is the threshold for the resulting P values
This allows for cross-interpretation of study popula- derived using the calculated test statistic for the end-
tions in different studies. point(s) of interest. This is intuitive as the P value is
A second factor is the heterogeneity of the individ- technically the conditional probability of seeing an
ual study population. Two phase III USA trials in the effect as large or larger by chance alone given that
1990s in aggressive lymphoma demonstrated that the null hypothesis is true. Therefore, the significance
phase II trials did not predict superiority of other level establishes the cutpoint, at which point it is
agents over CHOP (cyclophosphamide [C], adriamy- believed that the probability of seeing as extreme a
cin [H], vincristine [O], and prednisone [P]). At least result is due to chance alone under the assumption
one factor was the heterogeneity in the individual that the null hypothesis is true versus our rejection of
study populations. Shipp and colleagues developed the assumption that the null hypothesis is true. In
the International Prognostic Factor Index (IPI) that other words, this translates to the strength of evidence
encompassed DLBCL to address the issue of hetero- that it is necessary to see to reject the assumption that
geneity in patient study populations. This study the null hypothesis is true. 49
Chapter 4: Clinical trials in lymphoma
Confidence A range of values calculated from the sample observations that are believed with a particular probability to
interval contain the parameter value. These convey information that the P values do not.
Cox regression The most widely used statistical model in clinical research. This is used when the analysis of a clinical trial is based
model on the time to an event. It models the hazard function as a set of explanatory variables rather than the mean.
Event-free survival Time to event as determined by relapse, progression, or death.
Hazard function The probability that an individual experiences an event (death, impairment, etc.) in a small time interval. It is a
measure of how likely an individual is to experience an event, or a function of the age of the individual.
Intent-to-treat All patients entered in a clinical trial are analyzed together as reflecting that treatment, whether or not they
analysis completed or ever received the treatment. This prevents possible bias from many components.
Interim analysis An analysis, preferably conducted by an independent data-monitoring committee, to assess whether the
ongoing trial can realistically be expected to answer the primary question, taking into account the inclusion rate,
adverse events, previous experience, and the statistical significance as compared to the defined statistical
guidelines in the trial.
Kaplan–Meier A procedure for estimating the survival function for a set of survival times, some of which may be censored
estimator observations. For example, the probability of a patient surviving 2 years after a treatment can be calculated as the
probability of surviving 1 year given that the patient survived the first year.
Overall survival Time to death from any cause.
Power The probability of rejecting the null hypothesis when it is false.
P value An index measuring the strength of the evidence against the null hypothesis. It is not the probability that the
specified hypothesis is true.
Primary endpoint In lymphoma, the primary endpoints are overall survival and progression-free survival.
Randomization The key features of randomization are that the treatment assignment is based on chance alone, the patient
characteristics are balanced or equivalent, and the randomization process provides the foundation for the
statistical tests that are used.
Sample size The number of patients to be included in a study so that the study has a particular power to detect an effect.
Secondary The secondary endpoints in lymphoma include event-free survival, time to progression, etc.
endpoints
Significance level The level of probability at which it is agreed that the null hypothesis will be rejected. This is conveniently set at
0.05. P values need to be much smaller than 0.05 before they can be considered to provide strong or conclusive
evidence against the null hypothesis.
Preclinical Initial studies on in vitro samples and/or animal studies to obtain information Figure 4.1 Progression of clinical
on dosing and mechanisms of action trial research.
Phase II Trials used to obtain initial evidence of efficacy, and typically incorporate a
decision rule of what evidence is required to declare it promising vs. not. These are not
definitive trials on efficacy of the regimen. Also used to obtain more information on
toxicity profile of the regimen.
Phase III Trials that compare experimental regimen(s) to the standard of care; these trials
are typically large, randomized trials whose outcomes can influence clinical practice.
and thus affect the dose escalation decisions. It is pru- If no DLT is observed, then the next cohort of three
dent to restrict what constitutes a DLT carefully; often patients is treated at the next dose level. If moderate
these are restricted to non-hematologic toxicities, but toxicity (usually one patient with DLT) is observed,
they can include time to engraftment (e.g. failure to then another cohort of three patients may be treated
engraft by the absolute neutrophil count [ANC] by 28 at that dose level for a maximum of six patients. If
days could be considered a DLT). The timeframe for significant toxicity is observed (usually defined as two
which observed toxicity will be considered in the dose or more patients with DLT), the MTD will have been
escalation decisions should also be defined. When exceeded. A subsequent cohort of three patients may be
determining the study duration, it is important to con- accrued to the previous dose level if six patients had not
sider not only how long it takes to accrue a cohort of yet been treated to ensure it is the MTD. Accrual must
patients, but also for how long each cohort needs to be be stopped after each cohort of three patients is enrolled
evaluated, as defined by the specified toxicity observa- at a dose level to assess toxicity before proceeding to the
tion period. Even though a regimen may be given con- next dose level. This type of trial design was utilized to
tinuously until a patient progresses, or even when the study dose escalation of ProMACE-CytoBOM in
regimen is given for 6 months, it is not uncommon for DLBCL, resulting in a phase II study utilizing a 200%
only the first one to two cycles to be considered when dose level. More recently, this design was utilized in a
deciding whether or not a DLT was observed, and thus trial of R2-CHOP (lenalidomide, rituximab-CHOP)
whether or not the next cohort can be accrued to the where the standard dose of CHOP was utilized and
next dose level. the lenalidomide dose was incrementally increased.
The most commonly used phase I trial design is the Alternatives to the standard cohort-of-three design
standard cohort-of-three design. Patients are entered in for phase I trials include the continual reassessment
a stepwise fashion in cohorts of three, with up to six method (CRM) and the accelerated titration method. 51
patients per dose level, in order to determine the MTD. In brief, the CRM approach utilizes the investigator’s
Chapter 4: Clinical trials in lymphoma
prior beliefs and the observed data to develop a model or analyzed at the end of the study. An alternative is to
for determining the probability of unacceptable toxicity conduct the study with dose escalation decisions deter-
at a given dose level. This method uses data from each mined by MTD rules, and utilize the biologic efficacy
treated cohort to choose the dose for each successive data at the end of the study to help determine the
cohort. Advantages of this method include that the overall recommended dose level. This may require a
MTD may be determined more quickly and accurately, specified number of patients at the MTD and sub-
but disadvantages are that this approach is more com- MTD dose levels to ensure sufficient samples for this
plicated and potentially difficult to implement. Also, the type of evaluation by dose level.
performance of this design depends on the accuracy of Finally, there is the scenario where it is of interest
the model, and an incorrect model can cause problems to determine the recommended dose level only based
with too many patients treated at high dose levels or an upon some biologic or immunologic outcomes. This is
inaccurate estimate of the MTD. In the accelerated often the case in regimens where previous studies have
titration method there are different schedules for accel- determined the MTD to be higher than the dose levels
erated dose escalation. In brief, one schedule evaluates of interest in the proposed dose-finding trial, and only
one patient at a dose level until a certain (sub-DLT) level changes in these correlative outcomes are of interest in
of toxicity is observed, after which it reverts to a stand- determining the recommended dose level. This is the
ard cohort-of-three design. Another schedule allows for only situation in which multiple dose levels may be
intra-patient dose escalation using either a standard or simultaneously evaluated and where patients may be
CRM escalation plan. Advantages of this approach are randomized to dose levels. In all other phase I scenar-
that it can require fewer patients and, specifically, fewer ios, it is completely inappropriate to do so, and each
patients at potentially suboptimal levels, and that when dose level must be evaluated sequentially. However,
dose escalation is done within patients no patient is there are other approaches that can be taken that
undertreated. Disadvantages are that dose levels are reflect those of the tolerability-based phase I dose-
potentially evaluated with very few patients, and for finding trials but in which the trial takes into account
the intra-patient dose escalation there are ethical con- a previously defined measure of biologic response.
siderations for dosing patients until toxicity. Overall, there are many issues to consider when
The designs that have been discussed focus on the designing phase I clinical trials. Since these are dose-
determination of a recommended dose level based on finding trials, often there is no data yet on efficacy, and
MTD. In the current research environment, with more there is no promise that these regimens will be effective.
targeted therapies and/or immunotherapies, higher This affects the types of patients who are ethically
dose levels may not necessarily be associated with eligible for these types of trials. Also, it is important to
greater probability of response; i.e. the dose–response note that phase I trials are for determining recommen-
curve may not be consistently increasing, but rather ded dose levels, not for concluding initial evidence of
may be parabolic or some other function where the efficacy. While phase I trials can provide useful prelimi-
curve may reflect decreasing efficacy at higher dose nary data on efficacy, there are still limited numbers of
levels. In such a case, the phase I trial may be designed patients treated at a dose level. Finally, it is important to
to determine a recommended dose level based on note that even though individual agents have been
tolerability (as assessed through determination of the evaluated in the phase I setting, if their combination
MTD) and also some measure of efficacy or biologic has not, and this is the regimen of interest, this too must
response. Recent designs have been developed that use be evaluated in the context of a phase I trial. At a
a CRM approach, where the possible outcomes are no minimum, this tolerability assessment can be combined
efficacy and no toxicity, efficacy and no toxicity, or with a phase II trial, where the phase II portion of the
severe toxicity. These types of models are attractive in trial would be accrued at the determined MTD.
that they directly incorporate some measure of efficacy
(usually as measured through some biologic param-
eter) in the dose escalation decisions. However, these Phase II trials
designs also require that the biologic efficacy informa- The goals of a phase II trial are to assess therapeutic
tion be obtained relatively quickly so that it can be activity, to further characterize toxicity of the regimen,
used in the dose escalation decision-making process. and to determine if the treatment warrants further inves-
52 In practice, many of these endpoints are batched and/ tigation in the context of a phase III trial. Phase II trials
Chapter 4: Clinical trials in lymphoma
are typically the first formal evaluation of efficacy of an design is a concern when a large number of patients
experimental therapeutic regimen. With limited resour- are required, where many patients may be treated with
ces and the number of new agents available for evalua- a regimen that may not have any evidence to date of
tion in lymphoma, well-designed phase II trials are even potential therapeutic benefit.
more important. Therefore, sample sizes are constrained Because of these considerations, a multistage trial
to reduce the number of patients who are exposed to a is often the most appropriate method for evaluating
potentially ineffective or even toxic regimen, but a suffi- the efficacy of these experimental regimens. Most
cient number of patients are required to effectively eval- often, two-stage phase II designs are used, in which
uate evidence of efficacy. Phase II trials are not definitive one interim analysis is conducted to permit early stop-
evaluations of efficacy, but rather identify those regimens ping for futility and/or toxicity. These designs require
worthy of further study in the phase III setting. that accrual be temporarily suspended at each stage to
To evaluate efficacy in the phase II setting, decision evaluate those patients fully before proceeding to the
rules are established for the level of evidence required next stage. This can cause concern regarding momen-
to indicate that the regimen is promising. Typically, tum of the trial and timeliness of the study duration,
the outcome measures for documentation of efficacy and modifications to these designs are sometimes used
are dichotomized as “successes” or “failures,” where where the interim analyses are conducted but where
each is defined dependent on characteristics of the accrual is not halted (note that only the first n1 patients
regimen and/or target patient population. In the con- required for the interim analyses are included to eval-
text of lymphoma trials, this is most often the inci- uate the corresponding decision criteria at that stage).
dence of a clinical response (partial or complete) to If accrual is expected to be rapid, there is not much
treatment, although in some settings it may be more information on the toxicity profile for this regimen,
relevant to focus specifically on the incidence of com- and/or the primary endpoint can be evaluated in a
plete responses or other definitions of “success.” relatively short timeframe, then it is recommended
Accounting for constraints on type I and type II that accrual be temporarily halted between stages to
error rates and what true success rate one would expect allow for the interim analyses. A general rule of thumb
if the regimen was not effective (P0) versus what one is to estimate how useful an interim analysis would be
would expect if it truly was effective (Pa), a fixed if accrual was not halted; if all patients required for the
required sample size and decision criteria are trial will be accrued by the time results from an interim
established. analysis are available, then it is prudent simply to make
Multiple possible methods for developing phase II it a one-stage design (if there are a reasonably small
trials exist. The focus in this review is on the broader number of patients [40 or less]) or to impose a halting
issues to be considered when designing a phase II of accrual between stages.
study. A primary consideration is whether or not to Designs such as the Simon optimal or Simon min–
include one or more interim analyses, or “stages.” A max design are commonly used, and these provide a
one-stage phase II trial accrues patients as rapidly as rule for stopping in the event of early evidence of lack
possible and then evaluates all endpoints at the end of of efficacy. The Fleming multistage design also pro-
the trial. Resulting data are analyzed using the typical vides decision criteria to stop accrual when sufficient
binomial estimators (assuming a “success”/“failure” evidence indicates lack of activity, but in addition
variable) and exact binomial confidence intervals. provides an upper stopping limit, where accrual may
The primary drawback to the one-stage design is that be stopped for positive results. If there is sufficient
there is no formal opportunity for stopping the trial early evidence that the regimen is effective, it is often
early. This can be more problematic if results on the not of interest to halt accrual; therefore, accrual can be
first patients are available and investigators make early permitted to continue. This can allow a more complete
conclusions based on these observations without the evaluation of the toxicity profile in these patients as
benefit of a-priori criteria based on probabilistic rules. well as more precision in the estimated primary end-
One-stage designs are best used when the sample sizes point. What this upper stopping limit can provide is a
are moderately small and/or when the primary end- rule whereby it is possible to report out the early
point requires a relatively long follow-up to evaluate it evidence of the positive results. Without such a rule,
fully (e.g. response rate by 1 year, if late responses are it is inappropriate to report out efficacy results before
typical and should not be missed). However, this the study has completed accrual. 53
Chapter 4: Clinical trials in lymphoma
Randomized phase II trials type of trial is useful when there is concern regarding
the true success rate in the target population and/or
When more than one experimental therapeutic regimen
when it is unclear what the success rate is for the
is to be evaluated, a randomized phase II study design
standard of care.
provides a venue for simultaneously assessing the effi-
It should be noted that for these phase II screening
cacy of these regimens in the same patient population.
trials, the calculation of sample sizes has special consid-
Since phase II trials are not comparative trials, the
erations depending on the primary efficacy endpoints
purpose of a randomized phase II trial is to simulta-
(e.g. response, survival). The biggest difference in cal-
neously evaluate multiple regimens independently. A
culating sample sizes for standard versus screening
study design and decision criteria are generated for each
phase II trials is that, in the screening trial setting, the
treatment arm, where multiple treatment arms can
type I error is typically not constrained. By definition,
utilize the same study design and criteria as appropriate.
the type I error is the probability of concluding that one
The benefit with these trials is not only evaluating
is more promising than the other(s) when in truth they
multiple experimental treatments simultaneously but
are equivalent. In the context of a screening trial, this
also reducing potential patient selection biases.
type of error is not seen as a problem since it does not
Despite the fact that randomized phase II trials are
necessarily matter which one goes forward to the phase
underpowered for direct comparisons between arms,
III setting if the regimens are truly equivalent.
the success rates between arms can be informally eval-
Just as the hybrid study of phase I and phase II
uated together to identify the most promising regimen
endpoints can be studied in a phase I/II design, a phase
to be carried forward to the phase III setting. This
II/III design can also be utilized in special circum-
screening design approach simultaneously screens
stances. This randomized design utilizes an upfront
two or more regimens to identify the most promising.
phase II element where patients are also randomized to
The criteria for many such designs is based on the
the standard of care arm. Various strategies can be
regimen that results in the highest observed success
used in the first portion of this type of trial to deter-
rate by any amount. This design is appropriate for
mine which regimen(s) is (are) carried forward. Once
prioritizing between two experimental regimens
identified, the study continues accrual to the standard
when there is no a-priori reason to prefer one treat-
of care control arm and any regimens brought forward
ment over the other. Again, since this type of trial does
from the phase II portion of the trial. In the final
not support a direct comparison of efficacy endpoints
analysis, these treatment arms are directly compared
and is underpowered to do so, this is not a definitive
based on the defined efficacy endpoint(s). Like
trial for establishing efficacy of an experimental regi-
randomized phase II studies, statistical use of the
men but rather to identify the most promising regi-
reference arm is unclear and problematic. Different
men. In addition, a flexible screening design has been
strategies in this design have been proposed.
developed that further requires that, for one experi-
mental treatment regimen to be considered more
promising than another, its success rate must be
greater than the other by a prespecified differential. If Phase III trials
the success rates are not different by at least this While efficacy may be formally evaluated in the con-
amount, this design formalizes the ability to use text of phase II trials, these should not be considered
other factors such as quality of life, cost, immunologic definitive trials but simply preludes to the phase III
parameters, or other factors in the decision of which trial. In the phase III setting, a randomized controlled
regimen is most promising and should be carried clinical trial can affect clinical practice in the care of
forward to the phase III setting for a more definitive patients. The major purpose of a randomized con-
evaluation. There is the possibility in the context of trolled trial is to reduce the risks of bias. The phase
this type of design also to include a standard of care III setting is where the standard of care is evaluated,
treatment arm, where there is not direct comparison typically through the comparison of standard treat-
but rather to ensure that the experimental regimen is ment to one or more experimental regimens of inter-
considered more promising. This type of trial is more est. If the standard of care for a target population is no
problematic since there is an inclination and tempta- treatment or a “wait and watch” approach, the control
tion to make these direct comparisons, which are arm can be no treatment or a placebo control. The
54 inappropriate in the phase II setting. However, this decision on whether or not to use a placebo and/or
Chapter 4: Clinical trials in lymphoma
blinding of the assigned treatment depends on the phase II setting. These regimens included MACOP-B
potential for bias in the results owing to potential (methotrexate [M], adriamycin [A], cyclophosphamide
differences caused by patient and/or treating physician [C], vincristine [O], prednisolone [P], and bleomycin
attitudes. If a placebo is used, it needs to look the same [B]); m-BACOD (methotrexate [m], bleomycin [B],
and be able to be given in the same manner as the adriamycin [A], cyclophosphamide [C], vincristine
active intervention. The patient and/or treating physi- [O], and dexamethasone [D]); and ProMACE-
cian may be blinded to the true treatment received, as CytaBOM (prednisone [P], methotrexate [M],
necessary. adriamycin [A], cyclophosphamide [C], and etoposide
The design of the phase III trial will be determined [E], combined with cytarabine [C], bleomycin [B], vin-
first of all by the overall goal, whether the intention is cristine [O], and methotrexate [M]). This trial demon-
to assess superiority, equivalence, or non-inferiority. strated that all arms were equivalent, and therefore
In the most common type of phase III clinical trial an CHOP remained the standard of care. Subsequently,
experimental regimen is evaluated to determine if it is phase III trials demonstrated that rituximab-CHOP reg-
superior to another regimen, typically the standard of imens resulted in superior OS rates compared to CHOP.
care or control arm. In this case, it is of interest to test The issue of OS as an outcome for randomized clinical
the null hypothesis that the treatment arms are equiv- trials with effective subsequent therapies is a complex
alent. However, there is increasing interest in evaluat- issue. A subsequent therapeutic intervention in a patient
ing the equivalence or non-inferiority of one regimen who has relapsed that works better in the standard treat-
to another. This is often the case when an active stand- ment arm than in the experimental arm will lead to a
ard of care exists, where the goal may be to ensure that smaller OS difference.
an experimental regimen that is possibly less toxic, In the crossover design, every patient gets each
easier to administer, and/or is more cost-effective is intervention but they are randomized in terms of the
at least as efficacious as the standard treatment. In order in which they receive these interventions. These
equivalence trials, the trial tests the null hypothesis types of designs are best used when there is a short
that there is a difference in the efficacy between treat- duration of effect of the treatments (to avoid carry-
ment arms, and in a sense the null and alternate over effects and bias), and the patients are in relatively
hypotheses are the reverse of those in the superiority stable condition where they can complete both treat-
setting. For non-inferiority trials, it is of primary ments. An advantage of this type of design is that each
interest to ensure that the experimental regimen(s) is patient serves as their own control. This can poten-
(are) not worse than the standard of care. The exper- tially reduce the required sample size. This design is
imental regimen(s) may be superior, but this design most suited to chronic conditions, but for lymphoma
focuses on evaluating only if it is at least as effective as patients the response may depend on the order in
the standard treatment. which the different therapies are given. It is possible
Once the overall goals of the phase III trial have been to use crossover studies to evaluate interventions to
defined (superiority versus equivalence or non-inferior- ameliorate side effects or conditions that affect lym-
ity), then there are still many other factors to consider in phoma patients but generally they are not used when
the design of this type of trial. The most common is the attempting to improve OS or other measures of dis-
single-factor trial, where the primary endpoint between ease control. The disadvantage is that in trials in which
two treatment arms is evaluated and compared. This is the control arm patients cross over to the experimental
the most straightforward type of design, with patients treatment after clinical deterioration, a smaller OS
randomized to one of two treatment groups, and the difference could be observed.
endpoints of interest compared at the end of the trial. The factorial design can also be used when there
There are other special designs that can be used in the are two or more different comparisons to be made and
phase III setting, including multifactor (or multiple the treatments can be combined. For example, one
treatment arm), crossover, and factorial designs. The may wish to determine if the addition of an agent
advantage of a multifactor design is that it addresses makes therapy more effective, and also to assess
multiple treatment questions in the context of one clin- whether a shorter duration of therapy is as effective
ical trial. A specific example is the SWOG 8516 study, as the standard duration (e.g. 4 versus 6 months). For
which compared CHOP to three other regimens that this hypothetical example, there are two treatment
had sufficient evidence of promising efficacy in the questions of interest, and they can be combined as in 55
Chapter 4: Clinical trials in lymphoma
a two-by-two table, with patients randomized to seeing as extreme a result if the null hypothesis was
receive one of the four combined treatment possibil- indeed true. The earlier the interim analysis, the
ities. This is the only clinical trial that is designed to smaller the P value cut-offs; i.e. the fewer the patients
formally evaluate interactions between two treatment accrued and data analyzed, the stronger the evidence
interventions. In this setting, an interaction means that has to be to reject the null hypothesis.
the effect of one intervention depends on the level of As discussed above, the purpose of the phase III trial
another intervention. While the ability to assess the is a comparative randomized controlled trial to evaluate
existence of a treatment interaction can be of interest, two or more treatment interventions. The choice of
this type of trial design also introduces a level of com- primary endpoint to formally compare between treat-
plexity that can make implementation and conduct ment groups will depend on the overall goals of the
logistically difficult. In addition, the existence of a study. These primary endpoints can focus on efficacy,
treatment interaction can require a much larger sam- such as PFS, OS, or even response rates. However, it is
ple size for full evaluation. possible also to focus on tolerability-based endpoints
Another major consideration in the design of such as the incidence of a specific severe toxicity, if that
phase III clinical trials is the need for interim is what is truly needed to differentiate and determine
“looks” or analysis of the data. Formal interim anal- the best treatment intervention. Although the choice of
yses of the primary endpoint need to be defined a primary endpoint is flexible, the most commonly used
priori when designing the trial. Beyond the decision endpoint in the phase III setting for lymphoma studies
of the number of interim analyses, as well as the is PFS or OS. Given that phase III trials are used to help
timing of these analyses, it is important to determine define clinical practice for these patients, it is thus
in the design how to control for the potential inflation important to focus on the endpoints of ultimate interest
of the type I error rate arising from these interim when caring for these patients. One can argue that OS as
looks at the data (i.e. the increased possibility of well as quality of life is the ultimate goal when treating
rejecting the null hypothesis when it is in fact true, patients, but endpoints such as PFS are useful surro-
because of the additional analyses of the hypotheses). gates. The choice of endpoint will depend on the target
To account for this in the development of the trial patient population and the ability to conduct and ana-
design, we choose what is called an α (or type I error) lyze the trial in a reasonable timeframe. For example,
spending function; this essentially takes into account OS may best be used when evaluating treatment options
the multiple tests of the primary endpoint. These for aggressive lymphoma histologies, but for indolent
spending functions range from simple to complex. histologies this endpoint may take too long to evaluate
One of the more straightforward spending functions adequately with sufficient power. In that setting, it is
is that proposed by Pocock, which essentially divides likely better to focus instead on PFS. It is of interest that
the targeted type I error rate, α, by the number the immunochemotherapy protocols incorporating rit-
of looks, k. This approach is similar to a simple uximab eventually demonstrated an improvement in
Bonferroni type of correction seen in multiple com- OS in the rituximab arms.
parison corrections. This approach uses the same P The important part of this process is ultimately to
value cut-off for all analyses: if the P value is less than define fully the primary endpoint to be used in compar-
α/k then it is statistically significant. This approach is ing the effect of the treatment groups. For example, OS
easy to use and understand, but the cut-off at the final is typically defined as the time from study entry to death
analysis can be quite small. Other more complex from any cause. If it is of interest to evaluate survival
approaches are more commonly used, the most com- from time of diagnosis; this is valid for studies focused
mon being the O’Brien–Fleming method. This on newly diagnosed patients where lead-time biases and
approach uses a more complicated statistical algo- the potential impact of prior therapies are unlikely to
rithm to determine the P value cut-offs to be used present a problem. In defining PFS, the event of interest
for each analysis, where the cut-offs to determine is often inclusive of both disease progression as well as a
statistical significance increase as more patients are true relapse (disease recurrence after a complete
accrued and one gets closer to the final analysis. This response). Again, the key to time-to-event analyses is
approach, although more complicated in its calcula- to define effectively and consistently what constitutes an
tions, is the more intuitive. Recall that the P value event as well as the date from which the time will be
56 from the test statistic of interest is the probability of calculated (e.g. study entry or diagnosis). In the analysis
Chapter 4: Clinical trials in lymphoma
of these time-to-event endpoints between treatment In a phase III setting, rather than randomizing
arms, the differences are typically expressed in terms patients to different initial treatments, all patients
of a hazard ratio, which is the ratio of the event rates on may be given the same initial regimen and are only
each arm. Many considerations are required when randomized to different therapies at the maintenance
determining sample size based on a time-to-event end- phase. This second randomization asks a different
point such as the anticipated accrual period, accrual question. Patients who are eligible to be randomized
period, minimum length of follow-up on all patients, to maintenance therapy are separated in time and by
expected loss to follow-up, and, of course, the number response, usually CR or PR, therefore not all of the
of events expected or median time to the event of patients who entered the study initially will undergo
interest. Kaplan–Meier methods and/or Cox regression the second randomization, as they may need to be
models and corresponding log-rank or Wald statistics responding to be eligible. With the time separation,
are most often used to evaluate and compare time-to- long-term questions about the first randomization
event endpoints between treatment arms. These meth- may be difficult to answer. Patient selection biases
ods formally account for the possibility that not all are introduced because patients in the second random-
patients will have had the event of interest at the time ization can be different. Interactions of the different
of the analysis, but the important information that they treatments make the interpretations complex. If
have been event-free up until the last evaluation is patients are randomized up-front, then patients drop-
incorporated. ping out can present major issues since patients are
Randomization is a key component of the phase analyzed according to the initial randomization on an
III trial. There are different methods for randomiza- intent-to-treat basis. Comparing only those patients
tion. Other relevant issues include stratification (IPI who received the assigned treatment is not valid since
in initial randomization, response if a second ran- baseline characteristics are balanced only at the time of
domization, etc.) and the different timepoints at the randomization and the benefits of randomization
which patients are randomized to various treatments. are lost.
Randomization is a controlled assignment of patients A modification of this second approach has been
to treatment arms, which removes the potential for applied. To compare induction treatments without the
bias in treatment allocation, thus supporting the confounding effect of maintenance, analyses cannot
validity of statistical tests of significance. Patients simply exclude all patients randomized to a mainte-
are randomized at a time as close as possible to nance treatment because the proportion of non-
the initiation of treatment so as to avoid death, clin- responding patients relative to the whole population
ical deterioration, complications, ineligibility, or a would be higher and would underestimate the failure-
change of mind regarding participation in a clinical free survival (FFS) and OS. To achieve an unbiased
trial. Since clinical trials are based upon an intent-to- estimate, an approach may be applied (weighted Cox
treat analysis, this limits further variability. regression) that approximately doubles the information
There are different methods for randomization, for patients randomized to observation and uses a
where the method of interest will also depend largely modified variance estimator that is valid for the analy-
on whether or not it is necessary to stratify. Stratification sis. As described for weighted Cox regression, the
is a process used to insure treatment arms are well robust variance estimator provides a proper estimate
balanced with respect to important baseline prognostic of the variance of the relative risk estimate in this setting
factors and reduce the potential for confounding varia- and can be implemented using statistical software such
bles when analyzing the data. Stratification prevents as S-Plus. The concept of the weighted analysis, to
chance imbalances on important factors that can affect remove the bias that can result from analyzing only a
the interpretation of the primary endpoint between subset of the patients in two-stage randomized designs,
treatment arms. Post hoc methods to adjust for these is consistent with previously proposed methods for the
imbalances are less credible. Examples of stratification missing data problem. This approach was recently
factors in lymphoma clinical trials are IPI and histology applied to the E4494 trial, a DLBCL trial which initially
(if conducted across multiple histologies). In multicen- randomized patients to R-CHOP (rituximab CHOP)
ter studies, it is also useful to stratify on center to versus CHOP followed by a second randomization
account for inherent differences between treating sites. rituximab versus observation.
57
Chapter 4: Clinical trials in lymphoma
Habermann TM, Weller EA, Morrison VA, et al. Green S, Dahlberg S. Planned versus attained design in phase
Rituximab-CHOP versus CHOP alone or with II clinical trials. Stat Med, 1992;11:853–862.
maintenance rituximab in older patients with Herson J, Carter SK. Calibrated phase II clinical trials in
diffuse large B-cell lymphoma. J Clin Oncol, oncology. Stat Med, 1986;5:441–447.
2006;24:3121–3127.
Hunsberger S, Rubinstein LV, Dancey J, Korn EL. Dose
Juweid ME, Stroobants S, Hoekstra OS, et al. Use of positron escalation trial designs based on a molecularly targeted
emission tomography for response assessment of endpoint. Stat Med, 2005;24:2171–2181.
lymphoma: consensus of the imaging subcommittee of
international harmonization project in lymphoma. J Clin Korn EL. Nontoxicity endpoints in phase I trial designs for
Oncol, 2007;25:571–578. targeted, non-cytotoxic agents. J Natl Cancer Inst,
2004;96(13):977–978.
Nowakowski GS, Laplant B, Habermann TM, et al.
Lenalidomide can be safely combined with R-CHOP Liu P-Y, Dahlberg S, Crowley J. Selection of designs for pilot
(R2CHOP) in the initial chemotherapy for aggressive studies based on survival endpoints. Biometrics,
B-cell lymphomas: phase I study. Leukemia, 2011, Jul 01. 1993;49:391–308.
[Epub ahead of print] O’Quiqley J, Pepe M, Fisher L. Continual reassessment
Pfister B, Diehl V, Greb A, et al. International harmonization method: a practical design for phase 1 clinical trials in
of trial parameters in malignant lymphoma. Eur J cancer. Biometrics, 2002;58(1):245–249.
Haematol, 2005;75:53–54. Sargent DJ, Goldberg RM. A flexible design for multiple
Solal-Céligny P, Roy P, Colombat P, et al. Follicular armed screening trials. Stat Med, 2001;20(7):1051–1060.
lymphoma international index. Blood, Simon R, Wittes RE, Ellengerg SS. Randomized phase II
2004;104:1258–1265. clinical trials. Cancer Treatment Rep, 1985;69:1375–1381.
Swerdlow SH, Campo E, Harris HL, et al. (eds). WHO Simon R, Freidlin B, Rubinstein L, et al. Accelerated titration
Classification of Tumours of Haematopoietic and designs for phase I clinical trials in oncology. J Natl
Lymphoid Tissues. IARC: Lyon 2008. Cancer Inst, 1997;89(15):1138–1147.
The International Non-Hodgkin’s Lymphoma Prognostic Storer B. Design and analysis of phase I clinical trials.
Factors Project. A predictive model for aggressive Biometrics, 1989;45:925–938.
non-Hodgkin’s lymphoma. New Engl J Med,
Thall PF, Russel KE. A strategy for dose-finding and safety
1993;329:987–994.
monitoring based on efficacy and adverse outcomes in
The Non-Hodgkin’s Lymphoma Pathologic phase I/II clinical trials. Biometrics, 1998;54:251–264.
Classification Project. National Cancer Institute Zhang W, Sargent DJ, Mandrekar S. An adaptive dose-
sponsored study of classifications of non-Hodgkin’s finding design incorporating both toxicity and efficacy.
lymphomas: summary and description of a working Stat Med, Published online 11 October 2005.
formulation for clinical usage. Cancer,
1982;49:2112–2135.
Interpretation
Phase I, II, III trials Cox DR. Regression models and life-tables. J R Stat Soc Series
B, 1972;34:187–200.
Bryant J, Day R. Incorporating toxicity considerations into
the design of two-stage phase II clinical trials. Biometrics, Fleming TR. Clinical trials: discerning hype from substance.
1995;51:1372–1383. Ann Int Med, 2010;153:400–406.
Conaway MR, Petroni GR. Bivariate sequential designs for Kaplan E, Meier P. Nonparametric estimation from
phase II trials. Biometrics, 1995;51:656–664. incomplete observations. J Am Stat Assoc,
1958;53:457–481.
Fan SK, Chaloner K. Optimal designs and limiting optimal
designs for a trinomial response. J Stat Plan Inference, Lunceford JK, Davidian M, Tsiatis AA, et al. Estimation of
2004;126:347–360. survival distributions of treatment policies in two-stage
randomization designs in clinical trials. Biometrics,
Garrett-Mayer E. The continual reassessment method
2002;58:48–57.
for dose-finding studies: a tutorial. Clin Trials,
2006;3(1):57–71. O’Brien PC, Fleming TR. A multiple testing procedure for
clinical trials. Biometrics, 1979;35:549–556.
Gooley T, Martin P, Fisher L, et al. Simulation and design
tool for phase I/II clinical trials: an example from bone Therneau TM, Grambsch PM. Modeling Survival Data:
marrow transplantation. Control Clin Trials, Extending the Cox Model. New York, Springer-Verlag,
1994;15:45–462. 2000;161–163.
59
Chapter 4: Clinical trials in lymphoma
Wahed AS, Tsiatis AA. Optimal estimator for the survival effective subsequent therapies. J Clin Oncol,
distribution and related quantities for treatment policies 2011;29:2439–2442.
in two-stage randomization designs in clinical trials.
Lenz G, Wright G, Dave SS, et al. Stromal gene signatures
Biometrics, 2004;60:124–133.
in large-B-cell lymphomas. N Engl J Med,
2008;359:2313–2323.
Meyer RM. Generalizing the results of cancer clinical trials.
Future issues J Clin Oncol, 2009;27:187–189.
Cerhan SM, Ansell SM, Fredericksen ZS, et al. Genetic
Moss AJ, Francis CW, Ryan D. Collaborative clinical trials.
variation in 1253 immune and inflammation genes
N Engl J Med, 2011;364:789–791.
and risk of non-Hodgkin lymphoma. Blood,
2007;110:4455–4463. Rosenwald A, Wright G, Chan WC, et al. The use of
DeMets DL, Califf RM. A historical perspective on clinical molecular profiling to predict survival after
trials innovation and leadership. Where have academics chemotherapy for diffuse large-B-cell lymphoma.
gone? JAMA, 2011;305:713–714. N Engl J Med, 2002;346:1937–1947.
Korn EL, Freidlin B, Abrams JS. Overall survival as Ware JH, Hamel MB. Pragmatic trials-guides to better
the outcome for randomized clinical trials with health care? N Engl J Med, 2011;364:1685–1687.
60
Chapter
5
Hodgkin lymphoma
Stephanie Sasse and Andreas Engert
Pathology: Andrew Wotherspoon Molecular cytogenetics: Andreas Rosenwald and German Ott
Lymphoma, Second Edition, ed. Robert Marcus, John W. Sweetenham, and Michael E. Williams. Published by 61
Cambridge University Press. © Cambridge University Press 2014.
Chapter 5: Hodgkin lymphoma
Figure 5.5 Classical Hodgkin’s lymphoma. Reed–Sternberg cells are Figure 5.6 Nodular lymphocyte-rich classical Hodgkin’s lymphoma
large with abundant cytoplasm. Classical variants have a bilobed (lymph node). The lymphoid population has a modular/follicular
nucleus with prominent large inclusion-type nucleoli. growth pattern. The nodules contain a background of small
lymphocytes with scattered large cells.
Figure 5.7 Nodular lymphocyte-rich classical Hodgkin’s lymphoma Figure 5.8 Nodular lymphocyte-rich classical Hodgkin’s lymphoma
(lymph node). The large cells have the morphology of Reed– (lymph node). The Reed–Sternberg cells are negative for CD20.
Sternberg cells rather than popcorn cells, helping to distinguish this
from nodular lymphocyte-predominant Hodgkin’s lymphoma.
Figure 5.11 Nodular sclerosing Hodgkin’s lymphoma (lymph Figure 5.12 Lymphocyte-depleted Hodgkin’s lymphoma. A
node). The node architecture is effaced. There are cellular nodules proliferation of pleomorphic cells including occasional Reed–
that are separated from each other by sclerotic bands. The nodules Sternberg cells in a background that shows a paucity of lymphocytes.
contain Reed–Sternberg cells in a background of small lymphocytes,
histiocytes, eosinophils, and plasma cells.
and/or immunophenotyic features that make differen- confirmed in gene expression profiling studies that
tiation between DLBCL and cHL impossible. revealed global downregulation of the B-cell lineage
Morphologically the cells are frequently seen in gene repertoire, including surface molecules, kinase
sheets and may be associated with sclerosis. The cells signaling, and B-cell-specific transcription factors
are pleomorphic and there are often areas that have a (OCT2, BOB1, and PU1). Recently, it was demonstra-
distinctly cHL appearance adjacent to other areas that ted that the loss of B-cell identity in cHL results from
are more typical of DLBCL. Immunophenotypically, the the aberrant expression of the transcription factors
cells demonstrate a DLBCL/cHL overlap, with expres- ABF-1 and Id2, leading to downregulation of B-cell
sion of CD45 and preservation of the B-cell repertoire genes, but also to expression of genes not usually asso-
(CD20 and CD79a) associated with expression of CD30 ciated with B-cell lineage. A characteristic feature of
and, in most cases, CD15. The B-cell transcription fac- HRS cells is constitutive NFκB activity, which is medi-
tors OCT2 and BOB1 are frequently expressed together. ated by inactivating mutations of the NFκB inhibitor
IκBα in 30% of the cases, or, to a lesser extent, of IκBε.
Additionally, genomic amplification of a component of
Molecular pathology NFκB (c-Rel) is frequently demonstrated in cHL.
and cytogenetics HL cases have been demonstrated to harbor com-
The molecular and genetic analysis of the malignant plex clonal chromosomal abnormalities with fre-
tumor cells in this lymphoma, the HRS cells, has long quently hyperdiploid karyotypes. Some recurrent
been hampered by the paucity of the neoplastic cells chromosomal gains in cHL involve JAK-2 (9p23–24),
(1–2%) in the lymphoma specimen. However, advan- REL/BCL11a (2p13–16) and MDM2 (12q14), provid-
ces in tissue microdissection techniques have enabled a ing possible explanations for the constitutive NF-κB
more detailed molecular genetic analysis of HRS cells. and STAT member family expression in HL. In a
HRS cells in cHL have a pronounced lineage infi- thorough investigation of NLP-HL involving micro-
delity concerning the expression of antigen markers dissection, degenerate oligonucleotide-primed poly-
and frequently express myeloid (CD15), B-cell (Pax- merase chain reaction (DOP-PCR), and comparative
5), plasma cell (MUM-1 and syndecan), and activation genome hybridization (CGH) analysis chromosomal
markers (CD30). The analysis of IgVH genes, however, regions with a gain included 1p, 1q, 2q, 3p, 3q, 4q, 5q,
demonstrated that the vast majority of HRS cells har- 6p, 6q, 8q, 11q, 12q and X, and losses were demon-
bor clonal Ig rearrangements that represent a molec- strated at 17/17p. Chromosomal translocations occur
ular marker of B-cell derivation. Interestingly, a in a fraction of HL; in NLP-HL, BCL-6 rearrangements
proportion of these cases (25%) harbors destructive were identified in up to 48% of cases.
Ig rearrangements with crippling mutations. Since To identify potential transforming events able to
such crippling mutations would result in apoptotic rescue HRS cell precursors from apoptosis, candidate
cell death in reactive germinal center (GC) B-cells, it genes like TP53, CD95 (Fas), and N-RAS were inves-
has been postulated that HRS cells might be derived tigated, but no mutations were found. HRS cells,
from preapoptotic GC B-cells. Additional evidence nevertheless, are resistant to Fas-mediated apoptosis,
came from the analysis of composite lymphomas con- but no alterations of downstream effectors of the Fas
sisting of simultaneous HL and non-HL tumors that cascade were found. Very recently, inactivating muta-
demonstrated a clonal relationship between the two tions of A20, an NFkB inhibitory protein, were
lymphomas and a mutational pattern characteristic of described in a significant subset of classical HL.
GC B-cells. Expression of T-cell markers also occurs in Approximately 40% of cHL cases in the western
a subset of cHL and a proportion of these cases harbor world are associated with infection of EBV, whereas in
clonal T-cell receptor rearrangements and germline Ig some geographic regions like Latin America such
configuration, pointing to the existence of rare T-cell- involvement can reach up to 90% of cases depending
derived HL. In contrast to cHL, in NLP-HL, L&H cells on age and histological subtype. The transforming
show an immunophenotype consistent with B-cell capabilities of the viral latent membrane protein 1
derivation and harbor clonal and somatically mutated (LMP-1) activating NFκB can be considered a poten-
IgVH genes with ongoing mutations. tial transforming event after infection of HRS cell
Although derived from B-cells, HRS cells usually precursors by rescuing them from apoptosis. It is note-
lack B-cell lineage marker expression. This fact was worthy that the viral latent membrane protein 2 67
Chapter 5: Hodgkin lymphoma
Diagnosis, staging, and risk Stage III Involvement of lymph node regions on both
sides of the diaphragm (III), which may also be
stratification accompanied by involvement of the spleen
(IIIS) or by localized contiguous involvement
of only one extranodal organ site (IIIE) or both
Diagnosis and staging (IIISE)
III1 With or without involvement of splenic, hilar,
An excisional biopsy of a suspicious lymph node is celiac, or portal nodes
indispensable for the diagnosis of HL. HL patients are III2 With involvement of para-aortic, iliac, and
treated according to stage and risk factors. The histo- mesenteric nodes
logical subtype – except NLP-HL – does not influence Stage IV Diffuse or disseminated involvement of one or
the treatment decision. The stage of disease is assessed more extranodal organs or tissues, with or
without associated lymph node involvement
with the Cotswolds classification, a modified version
of the Ann Arbor classification, which was published Designations applicable to any disease stage.
by Lister (Table 5.2). A No symptoms
The clinical staging includes chest X-ray, abdomi- B Fever (temperature, >38°C), drenching night
nal sonography, CT scans of the neck, thorax, abdo- sweats, unexplained loss of 10% of body weight
within the preceding 6 months
men and pelvis, bone marrow biopsy, and bone
marrow or skeletal radionuclide imaging. In some X Bulky disease (a widening of the mediastinum
by more than one-third of the presence of a
cases, additional tests such as MRI, PET, or a liver nodal mass with a maximal dimension greater
biopsy might be indicated. So far, the role of 18FDG- than 10 cm)
68 PET in the initial diagnostic procedure has not been E Involvement of a single extranodal site that is
clearly defined. contiguous or proximal to the known nodal site
Chapter 5: Hodgkin lymphoma
Table 5.3 Definition of treatment groups according to the EORTC/GELA and GHSG.
(Groupe d’Etudes des Lymphomes de l’Adulte), and based chemotherapy. Residual disease in the CT scan
the GHSG classify patients into early favorable, after completion of chemotherapy was further
early unfavorable, and advanced stages depending assessed by 18FDG–PET scan. Only those patients
on the clinical factors listed in Table 5.3. A recently with 18FDG–PET-positive residual lesions received
performed retrospective analysis of the GHSG con- consolidation radiotherapy. For those patients with a
firmed the need to differentiate an early favorable negative 18FDG–PET scan, a progression-free survival
and unfavorable risk group, and identified the risk (PFS) of 92.1% could be documented after a follow-up
scores of the GHSG, the EORTC and of the of 36 months; this was about the same as the PFS of
National Comprehensive Cancer Network (NCCN) those patients with complete remission in the CT
as equally suitable to define the early unfavorable scan. The negative predictive value of 18FDG–PET
risk group. scan determined in this analysis was 94.6% (95% CI,
To define the risk to patients with advanced disease 92.7–96.6%). Thus, this preliminary analysis indicates
more precisely, the “International Prognostic Score” that consolidation radiotherapy can be omitted in
(IPS) was developed. The IPS consists of seven factors, advanced stage HL patients with 18FDG–PET-negative
which were significantly related to an unfavorable prog- residual disease after treatment with BEACOPP-
nosis when present at initial diagnosis of HL: serum escalated. The negative predictive value of 18FDG–
albumin <4 g/dL, hemoglobin <10.5 g/dL, male sex, age PET scan after completion of chemotherapy in early
≥45 years, stage IV disease, leukocytosis >15 000/mm³, favorable and unfavorable HL is still being evaluated;
lymphocytopenia <600/mm³ and/or <8% of white cells. so far there are data supporting a PET-guided radio-
therapy approach in early favorable and unfavorable
HL out of a clinical trial.
Response assessment Response-adapted chemotherapy treatment of
In addition to CT restaging, metabolic imaging with HL might also be developed on the basis of interim
18
2-[18F]fluoro-2-deoxyglucose–positron emission FDG–PET scan results. Several phase II trials reported
tomography (18FFDG–PET) has gained an increasing a negative predictive value of over 90% for interim
18
relevance in response assessment of HL patients. A FDG–PET scans; additionally, a strong correlation
significant prognostic value of negative 18FDG–PET between result of the early interim 18FDG–PET scan
scans after completion of the chemotherapy courses and PFS could be identified. However, the negative
could be verified in the HD15 trial of the GHSG. HL predictive value might vary dependent on the chemo-
patients included in the HD15 trial were randomly therapy regimen. The ongoing HD18 trial is evaluating
assigned to receive six or eight courses of a bleomycin, a response-adapted, BEACOPP-escalated-based chemo-
etoposide, adriamycin, cyclophosphamide, vincristine, therapy approach in advanced HL patients dependent 69
procarbazine, prednisone (BEACOPP)-escalated- on the interim 18FDG–PET scan results.
Chapter 5: Hodgkin lymphoma
favorable HL. Treatment-associated toxicity might be clearly defined and there is an ongoing desire to opti-
further reduced by implementing a targeted drug such mize therapy in this risk group.
as the anti-CD30 immunoconjugate bentuximab vedo- Several trials have shown that the reduction of field
tin and by further reducing radiotherapy intensity. size to IF-RT does not compromise the efficacy of treat-
ment. The Milan trial, which compared subtotal nodal
Early-stage unfavorable disease irradiation (STNI) with IF-RT after four courses of
Early-stage unfavorable HL has been shown to be ABVD in patients with early-stage favorable and unfav-
associated with a significantly worse outcome than orable stages (see Table 5.4), reported a similar treat-
early-stage favorable HL and has therefore to be treated ment outcome in both arms. The H8U, conducted by
more intensively. So far, combined modality treatment EORTC/GELA, randomized patients to six cycles of
has to be regarded as standard in early-unfavorable MOPP/ABV + 36 Gy IF-RT, four cycles of MOPP/
HL. However, best chemotherapy, optimal number of ABV + 36 Gy IF-RT, and four cycles of MOPP/ABV 71
cycles, and the radiotherapy regime have not been + STNI. There was no difference between the arms in
Chapter 5: Hodgkin lymphoma
terms of response rates, EFS, or OS. The HD8 trial of as the number of cycles. Alternation or hybridization of a
the GHSG, the largest trial investigating radiotherapy MOPP-like regimen with ABVD did not produce better
reduction, confirmed these results: patients with early- outcomes when compared with data reported for ABVD
stage unfavorable disease were randomized to two alter- alone nor did the increase of chemotherapy courses.
nating cycles of COPP (cyclophosphamide, vincristine, Because of these disappointing outcome rates more
procarbazine, prednisone)/ABVD followed by 30 Gy intensive chemotherapy regimes, which were originally
radiotherapy in either EF (arm A) or IF (arm B) tech- developed for the treatment of advanced stages, have
nique. Final results at 5 years and the follow-up analysis been evaluated for the treatment of early unfavorable
at 10 years did not disclose significant differences stage (Table 5.5). The HD11 trial performed by GHSG
between the two arms in terms of FFTF and OS; how- and the H9U trial conducted by EORTC/GELA com-
ever, more toxicity was reported in the patients who pared four cycles of ABVD and four cycles of
were treated with EF-RT (Table 5.5). BEACOPP-baseline. The H9U trial additionally ana-
Despite the excellent initial remission rates obtained lysed whether six cycles of ABVD are more effective
with ABVD and radiotherapy, approximately 15% of than four cycles of ABVD. At a median follow-up of 4
patients in the early unfavorable stage relapse within years, the treatment arms of the H9U trial did not show
5 years and about another 5% have primary progressive any difference with respect to EFS and OS. The analysis
72 disease. Efforts have been made to improve the efficacy of the GHSG HD11 trial did not show any significant
of chemotherapy by altering drugs and schedules as well difference between ABVD and BEACOPP-baseline
Chapter 5: Hodgkin lymphoma
regimen was associated with more acute and long-term the eight courses of BEACOPP-escalated regimen and by
toxicity than ABVD such as a higher incidence of sec- evaluating the role of consolidative radiotherapy.
ondary leukemia and infertility. The HD12 trial of the GHSG compared eight
74 The subsequent GHSG trials, HD12 and HD15, courses of BEACOPP-escalated with four courses of
aimed to reduce therapy-associated toxicity by modifying BEACOPP-escalated followed by four courses of
Chapter 5: Hodgkin lymphoma
treatment of choice for HL patients who relapse or are ASCT, but their reported outcome was significantly
primary refractory to the first line chemotherapy. Two worse than the outcome of patients with relapsed HL.
prospective, randomized trials have shown that A further prognostic factor and a significant pre-
HDCT/ASCT results in an improved disease control dictive marker for the outcome after HDCT/ASCT is
compared to conventional chemotherapy. In the BNLI the chemosensitivity to second line induction therapy.
trial, patients with relapsed or refractory HL were Therefore the optimizing of the reinduction chemo-
treated with conventional-dose mini-BEAM (carmus- therapy is an important approach to improve the out-
tine, etoposide, cytosine arabinoside, melphalan) or come of relapsed HL.
high-dose BEAM and ASCT. The 3-year EFS was sig- There are several salvage chemotherapy regimens
nificantly better after high-dose chemotherapy (53% that have been reported to be effective in single-arm trials
versus 10%). The HDR1 multicenter trial conducted (Tables 5.7 and 5.8). Since the Dexa-BEAM regimen has
by the GHSG and the EBMT (European Group for
Blood and Bone Marrow Transplantation) compared
four cycles of Dexa-BEAM with two cycles of Dexa- Table 5.7 Second line polychemotherapy regimens.
BEAM followed by BEAM and ASCT. The freedom
from second failure (FF2F) at 3 years was significantly Salvage Drug combination References
better in the transplanted group than in the group regimen
receiving conventional salvage chemotherapy (55% ESHAP Etoposide, methylpredniso- Aparicio
versus 34%). The OS did not differ significantly lone, high-dose cytarabine, et al., 1999
cisplatin
between the two treatment arms in both trials.
The benefit from HDCT/ASCT is significantly ASHAP Doxorubicin, Solu-Medrol, Rodriguez
high-dose cytarabine, et al., 1999
influenced by prognostic factors, such as the duration cisplatin
of the first remission. As a result of a retrospective
ICE Ifosfamide, carboplatin, Moskowitz
analysis performed by the GHSG a prognostic score etoposide et al., 2001
for relapsed HL was developed; relapse within 12 DHAP Dexamethasone, high-dose Josting et al.,
months after initial therapy, stage III or IV disease, cytarabine, cisplatin 2005
and a low hemoglobin level (less than 12 g/dL for men GVD Gemcitabine, vinorelbine, Bertlett et al.,
or less than 10.5 g/dL for women) at relapse were pegylated liposomal 2007
identified as unfavorable prognostic factors. doxorubicin
Prospective phase II trials indicate that patients IGEV Ifosfamide, gemcitabine, Santoro
with primary refractory HL also benefit from HDCT/ vinorelbine et al., 2007
Table 5.8 Results of selected phase II trials with salvage chemotherapy regimens.
been shown to be too toxic and since the successful stem reduced with a maintenance treatment consisting of
cell collection rate with modified BEAM-regimens is the deacetylase inhibitor panobinostat.
reported to be rather low, new salvage chemotherapy For patients with primary refractory HL a tandem
regimens have been developed. Platin-based regimens ASCT might be an appropriate approach. In a prospec-
include ESHAP (etoposide, methylprednisolone, cytara- tive multicenter phase II trial of the French lymphoma
bine, cisplatin), ASHAP (doxorubicin, methylpredniso- study group (“Groupe d’Etude des Lymphomes de
lone, cytarabine, cisplatin), DHAP (dexamethasone, l’Adulte” (GELA)/SFGM), a tandem ASCT resulted in
high-dose cytarabine, cisplatin), and ICE (ifosfamide, a 5-year OS of 46% in primary refractory and poor-risk
carboplatin, etoposide). The response rates of these che- relapsed HL.
motherapy regimens range from 70% to 89%, including
complete remission (CR) rates from about 21% to 34%;
no severe acute treatment-related complications have Salvage radiotherapy
been described, with myelosuppression being the most Salvage radiotherapy alone offers a treatment option for
relevant toxicity. Gemcitabine-based regimens, such as a selected subset of patients with localized relapse in
IGEV (ifosfamide, gemcitabine, vinorelbine) and GVD previously non-irradiated areas. In a retrospective anal-
(gemcitabine, vinorelbine, pegylated liposomal doxoru- ysis of the GHSG published by Josting et al. (2005), the 5-
bicin), have been shown to induce similar overall year freedom from second failure (FF2F) and OS rates
response rates and a high rate of successful stem cell were 28% and 51% after salvage radiotherapy alone. The
mobilization. In the phase II trial evaluating the efficacy retrospective analysis of Campbell et al. (2005) reported
and safety of IGEV, a CR rate of 54% could be docu- 10-year FF2F and OS rates of 33% and 46% for those
mented; impressively, 60% of the primary refractory patients with salvage radiotherapy-treated recurrent HL
patients responded. in limited stage. So far, there are no data of controlled
So far, there are no randomized trials comparing the trials analyzing the role of salvage radiotherapy in a
efficacy of these different second line regimens. The selected group of patients with relapsed HL.
recently reported prospective, randomized HRD2 trial Summing up, HDCT, consisting of two courses of
performed by the GHSG, the EORTC, and other reinduction therapy such as DHAP, followed by ASCT
European groups confirmed the efficacy of DHAP, is the treatment of choice for patients with relapsed HL
and indicates that two courses of DHAP directly fol- after first line chemotherapy. For patients with primary
lowed by HDCT/ASCT are equally effective when com- refractory HL, a tandem ASCT might be a more appro-
pared to a more aggressive sequential induction priate treatment approach. A selected group of patients
treatment of two courses of DHAP followed by high- with localized relapse in previously non-irradiated areas
dose cyclophosphamide, methotrexate plus vincristine, can be effectively treated with salvage radiotherapy alone.
and etoposide. The documented 3-year-FFTF rates Those patients with relapsed disease after initial radio-
were 71% in the standard arm consisting of two courses therapy should receive an anthracycline-containing che-
of DHAP followed by HDCT/ASCT versus 65% in the motherapy regime.
experimental arm; the evaluated 3-year OS rates were
87% versus 80%. Thus, two courses of DHAP followed Relapse after high-dose
by HDCT/ASCT are considered current standard of
care for patients with relapsed HL. chemotherapy and autologous stem
A further increase in efficacy of the salvage induc- cell transplantation
tion treatment might be achieved by combining the
induction chemotherapy with a targeted therapy. In Allogeneic stem cell transplantation
the ongoing HDR3i trial the GHSG evaluates the Allogeneic stem cell transplantation is an experimental
combination of DHAP with the mTor inhibitor evero- treatment option resulting in long-term remissions in
limus (RAD001) compared to DHAP. The outcome of a subset of young HL patients with relapse after ASCT.
relapsed HL might be additionally improved by treat- However, the role of allogeneic stem cell transplanta-
ing those patients with an increased risk to relapse tion is controversial.
after HDCT with a maintenance treatment. In a The existence of a relevant “graft versus HL” (GVL)
placebo-controlled trial the GHSG is currently exam- effect has been one point of this discussion. There are no
ining if the relapse rate after ASCT can be significantly prospective randomized trials comparing the efficacy of 77
Chapter 5: Hodgkin lymphoma
allogeneic stem cell transplantation to a second ASCT or factors for poor OS and PFS. It is still a matter of dis-
to salvage radiochemotherapy. Small retrospective, non- cussion whether achievement of CR should be considered
randomized analyses reporting a lower relapse rate for as a precondition for allogeneic transplantation.
allogeneic stem cell transplantation than for ASCT sug- Allogeneic transplantation might be an appropriate
gested that allogeneic stem cell transplantation may be strategy in those patients with relapsed HL after ASCT
associated with a clinically significant GVL activity. who are in a good performance status and have chemo-
More convincing evidence for the existence of GVL sensitive disease. Retrospective analyses reported long-
activity comes from the use of donor lymphocytes term PFS rates of about 20% after reduced intensity
(DLIs) in HL patients who had relapsed after allogeneic conditioning allogeneic transplantation. However, pro-
transplantation. Single center retrospective analyses spective clinical studies are required to define clear
reported response rates to DLIs of up to 50%. indications for this treatment option.
Unfortunately, the majority of these HL patients did not
achieve long-term benefits so that further studies are
necessary to optimize the application of DLIs in HL.
Targeted therapy of HL
The retrospective analyses reported by Sureda et al. Based on the increasing knowledge of the pathogenesis
(2008) and by Sarina et al. (2010) supported the evidence of HL, new treatment approaches are currently being
of a GVL effect; those patients developing chronic “graft developed which target cell surface molecules, modulate
versus host disease” (GvHD) had a significantly lower risk the function of certain intracellular proteins, or modu-
of relapse compared to patients with no chronic GvHD. late the immune response of the microenvironment. An
There was no impact on treatment-related mortality. effective targeted therapy, given alone or in combina-
OS and EFS data reported for myeloablative tion with chemotherapy, might improve the long-term
allogeneic stem cell transplantation were rather disap- outcome of relapsed and primary refractory patients. As
pointing because of unacceptably high treatment-related a component of first line treatment a biologically based
mortality. To reduce treatment-related toxicity, different drug might eliminate residual lymphoma cells after
reduced-intensity conditioning (RIC) regimens have chemotherapy. Additionally, combination of chemo-
been developed. A retrospective analysis performed by therapy with a targeted therapy might help to reduce
the toxicity of the currently used regimens.
the European Group for Blood and Bone Marrow
In this chapter a selection of new drugs, which were
Transplantation (EBMT) and reported by Sureda et al.
either proved to be effective or which showed promis-
(2008) showed a significantly decreased non-relapse-
ing results in clinical trials are discussed.
related mortality (NRM) in relapsed HL patients receiv-
ing RIC compared to patients treated with myeloablative
conditioning (1-year incidences of NRM 23% versus Anti-CD30 monoclonal antibodies/immunotoxins
46%; P = 0.001). Despite a significantly higher relapse Because of its dense and highly restricted expression
rate (57.3% versus 30.4% after a median time of 6 on HRS cells, the surface antigen CD30 is an ideal
months), the 5-year OS rate was significantly higher in target for targeted therapy.
the RIC group (28% versus 22%; P = 0.003). In patients The clinical results of the first-generation anti-
receiving a reduced intensity conditioning, relapse rates CD30 monoclonal antibodies SGN-30 and MDX-060
seem to be higher with less-intensive non-myeloablative were rather disappointing. By coupling the anti-CD30
regimens than with more intensive regimens like the antibody cAC10 to the antitubulin agent monomethyl
combination of fludarabine and melphalan. A promis- auristatin E (SGN-35: brentuximab vedotin), a potent
ing new approach reported by Burroughs et al. (2007) is anti-CD30 targeted therapy was developed. In the first
the use of haploidentical donors in combination with phase I trial published by Younes et al. (2010), the
cyclophosphamide administration before and after maximum tolerated dose (MTD) of SGN-35 (brentux-
allogeneic transplantion. imab vedotin) administered every 3 weeks was 1.8 mg/
Multivariate analyses of several retrospective studies kg; of 12 patients who received the MTD, six patients
have identified the pre-transplant disease status as the had an objective response. In a subsequent phase II
most relevant prognostic factor for outcome after alloge- trial 102 patients suffering from HL relapse after autol-
neic transplantation. The largest retrospective analysis ogous stem cell transplantation were included. The
reported by Robinson et al. (2009) identified chemore- response rate was 73% with a median duration of
78 fractory disease and poor performance status as risk response of 6.7 months (Younes et al., 2012). Based
Chapter 5: Hodgkin lymphoma
on these results brentuximab vedotin received appro- The potential therapeutic value of mTOR inhibi-
val from the FDA and the EMA for treatment of HL tion in HL could be documented with the oral mTOR
patients who relapsed after autologous stem cell trans- inhibitor everolimus. In a phase II trial published by
plantation or received at least two prior systemic Johnston (2010), treatment of relapsed HL with daily
therapies and are not eligible for autologous stem doses of 10 mg everolimus resulted in a clinical
cell transplant. Subsequently, further clinical trials response in nine of 19 enrolled patients.
have been initiated or are being planned in order to
implement brentuximab in the first line or second
line treatment of HL. Targeting the microenvironment
Rituximab
Protein-specific “small molecules”
While the expression of the CD20 antigen on HRS cells
Histone deacetylase inhibitors (HDAC inhibitors) is rather rare, CD20 is highly expressed on reactive
Histone acetyltransferases (HATs) and histone deace- B-cells of the microenvironment. The interaction of
tylases mediate acetylation and deacetylation of spe- these reactive B-cells with HRS cells via CD40 or CD80
cific lysine residues on histone and non-histone is assumed to play an important role for survival of HRS
proteins and thereby regulate the expression and the cells. Thus, depletion of reactive B-cells with rituximab
functional status of a variety of proteins that are might improve treatment efficacy in classical HL. In a
involved in cell proliferation, differentiation, survival, pilot study performed by the M.D. Anderson Cancer
angiogenesis, and immunity. Overexpression of differ- Center, an overall response rate of 23% (5/22 enrolled
ent HDACs has been reported in a number of human patients) in patients with relapsed classical HL could be
cancers. HDAC inhibitors have been shown to induce achieved with rituximab monotherapy. The following
growth arrest and apoptosis of transformed cells, while phase II trial reported by Wedgwood et al. (2007) indi-
not-transformed cells are resistant to HDAC inhibi- cated that first line treatment results might be improved
tors in vitro. Several HDAC inhibitors have antiproli- by combining ABVD with rituximab. The ongoing
ferative activity in HL-derived cell lines. In addition to HD18 trial of the GHSG is evaluating if those patients
the antiproliferative and proapoptotic effect, HDAC with positive interim 18FFDG–PET scan benefit from
inhibitors might have additional modes of action such additional treatment with rituximab.
as an immunomodulatory effect.
Promising clinical results in HL could be docu-
Lenalidomide
mented with the oral pan-HDAC inhibitor panobino-
stat. In a phase I trial published by Prince et al. (2007), Lenalidomide, a thalidomide derivative, represents an
eight of 20 HL patients included in the trial achieved a attractive option for the treatment of HL as it targets
PR. The final analysis of a recently finished phase II trial several signal pathways that regulate survival of HRS
showed a response rate of 27% (Younes et al., 2012). cells. In addition to direct antiproliferative and pro-
A double-blind, placebo-controlled phase III trial eval- apoptotic effects, an immunomodulatory and antian-
uating the safety and efficacy of panobinostat as main- giogenetic activity of lenalidomide has been described.
tenance treatment after ASCT is currently ongoing. Two phase II trials and several named patient programs
have shown promising results with single-agent lenali-
mTOR inhibitors domide treatment in relapsed HL, so that a phase I/II
study of lenalidomide combined with conventional che-
The mammalian target of rapamycin kinase mTORC1
motherapy (adriamycin, vinblastine, dacarbazine) for
regulates protein synthesis by phosphorylation of two
elderly HL patients has been initiated by the GHSG.
translation initiation factors (4E-binding protein 1 and
In addition to the anti-CD30 immunoconjugate
p70S6K), which lead to an increased synthesis of pro-
brentuximab vedotin, which has substantially
teins involved in cycle progression, proliferation, angio-
changed the treatment options in HL patients,
genesis, and survival pathways. An increased or
additional promising targeted therapy approaches
constitutive activation of mTORC1, for example
include, CD20-specific monoclonal antibodies, his-
through the phosphatidylinositol 3-kinase (PI3K)/AKT
tone deacetylase inhibitors such as panobinostat,
pathway, can be documented in different cancers. In HL
mTOR inhibitors such as everolimus, and the
cell lines, the inhibition of mTORC1 induces cell cycle 79
thalidomide-derivative lenalidomide.
arrest and apoptosis.
Chapter 5: Hodgkin lymphoma
suitable parameter to guide treatment intensity. COPP-AVBD or with ABVD alone; the HD9 trial
Treatment outcome and toxicity of response-adapted additionally showed a higher infection rate for the
therapy is further evaluated in the EuroNet-PHL-C1 BEACOPP-escalated regimen. To reduce the infection
trial initiated by the European Network for Paediatric rate, an obligatory antibiotic prophylaxis and G-CSF
Hodgkin Lymphoma (EuroNet-PHL). application were incorporated into the HD18 protocol.
Published data focusing on the treatment of adoles- Bleomycin-based chemotherapy regimens, such as
cent HL patients (15–19 years old) are scarce, and it is ABVD or BEACOPP, as well as mediastinal irradia-
still a matter of discussion whether adolescent HL tion, may induce acute pneumonitis followed by lung
patients should be treated according to adult or pediatric fibrosis.
protocols. There are some retrospective analyses indicat- Acute cardiac toxicity as a result of mediastinal
ing a better outcome of adolescent than of adult HL radiotherapy or an anthracycline-containing chemo-
patients. Most of the prospective trials including adoles- therapy regimen is rather rare. It most frequently
cents have not performed age-specific survival analyses. presents with coronary heart disease or acute cardiac
Subgroup analysis of the DAL-HD-90 trial performed failure owing to therapy-induced cardiomyopathy.
by the German Paediatric Hodgkin Study Group
(GPOH-HD) and reported by Schellong et al. (1999)
indicated that treatment of adolescent HL patients with Long-term toxicity
a risk-adapted, combined modality approach including Similar to short-term treatment-related toxicity, long-
chemotherapy regimens such as OEPA or OPPA results term complications vary in frequency and severity
in EFS rates which do not differ significantly from those dependent on the chemotherapy regimen and the addi-
achieved for children. Therefore it seems to be reason- tional application of radiotherapy. Chemotherapy-
able to develop a risk- and response-adapted treatment associated long-term toxicity particularly depends on
approach for adolescent HL patients. Randomized trials the cumulative dose of certain chemotherapeutic agents.
are indispensable to define the optimal treatment regi- The most serious long-term treatment-associated
men for adolescent HL patients. complications are the development of secondary neo-
In summary, standard of care for childhood HL is a plasia and cardiovascular disease. Other late effects
combination chemotherapy combined with low-dose, contributing to morbidity and/or mortality of long-
limited-field radiation. A promising approach to realize term survivors of HL are pulmonary dysfunction,
individualized, risk-adapted treatment is early response- infectious complications, hypothyroidism, infertility,
adapted therapy. The optimal treatment regimen for and fatigue.
adolescents has to be determined in randomized trials.
Secondary malignancies
Treatment-associated toxicity The most frequent secondary malignancies are solid
As a consequence of the impressive long-term remis- tumors, particularly breast cancer and lung cancer.
sion rates in HL, the reduction of treatment-related Leukemia and NHL account for 20–25% of the
complications has become increasingly important. observed secondary cancers. While the risk of devel-
oping secondary leukemia is mainly confined to the
initial 5 years of follow-up, the incidence of secondary
Short-term toxicity solid tumors continuously increases up to a cumula-
Acute treatment-associated adverse effects include nau- tive risk of 23% at a follow-up of 25 years and exceeds
sea and vomiting, hematotoxicity, infections, mucositis, the cumulative risk of leukemia and NHL. However,
and neurological complications such as polyneuro- the relative risk of solid cancer varies significantly
pathy. Less frequent acute complications are pulmonary dependent on the age at diagnosis and the attained age.
dysfunction or acute cardiac failure. The severity of Treatment with alkylating chemotherapy has been
these therapy-associated complications depends on identified as the most important risk factor for the
the chemotherapy regimen and the additional applica- development of secondary leukemia, with the cumu-
tion of radiotherapy. Chemotherapy with BEACOPP- lative alkylator dose being the strongest determinant of
escalated results in a significantly higher incidence of risk. Additionally, treatment with topoisomerase
grade 3 and 4 leukopenia, thrombocytopenia, and ane- inhibitors has been associated with an increased risk 81
mia than treatment with BEACOPP-baseline, with of secondary acute myeloid leukemia (AML).
Chapter 5: Hodgkin lymphoma
Radiotherapy, chemotherapy with alkylating screening tests, the timing in relationship to initial
agents, and smoking are the most relevant risk factors HL treatment, and the frequency needs to be defined.
for the pathogenesis of secondary lung cancer.
Smoking can be considered as an amplifying risk fac-
tor for treatment-associated lung cancer. Fertility
According to the case-control study reported by Sterility is a long-term treatment-related complication
van Leeuwen et al. (2003), increased risk of breast with serious impact on the quality of life of young HL
cancer after HL is largely attributable to radiotherapy patients. A high incidence of primary ovarian insuffi-
of the chest, with a significant correlation of the ciency or oligospermia/azoospermia has been docu-
relative risk to the radiotherapy dose. A risk reduc- mented after treatment with alkylating agents such as
tion of breast cancer can be documented after chemo- cyclophosphamide or procarbazine or after radiother-
therapy with gonadotoxic substances such as apy to the pelvis or to the testes. Further risk factors for
alkylating agents. female infertility are advanced-stage HL and age over 30.
A strategy to reduce the incidence of secondary Pre-treatment semen analysis showed that HL itself
malignancies is the reduction of treatment intensity, is associated with a significant impairment of sperm
for example by implementing targeted drugs or by quality. Post-treatment dysspermia can be observed in
reducing radiation field size and radiation dose. up to 90% of HL patients depending on the intensity of
Further prevention strategies include smoking cessation the chemotherapy treatment. Therefore cryopreserva-
programs and counselling on sun-safety practice. Since tion prior to therapy is strongly recommended for
early detection through screening tests can reduce mor- male HL patients.
tality in several cancer types, several screening tests have So far, no standardized procedures for assessment
been evaluated in patients surviving HL. As a result of and for preservation of female fertility after HL
one of these screening trials, annual MRI or screening treatment have been established. Available data of
mammography is recommended by a UK screening prospective trials suggest that female HL patients
programme beginning 8 years after mediastinal radio- receiving six courses of BEACOPP-escalated do not
therapy and an age of at least 25 years. Screening pro- benefit from administration of oral contraceptives or
tocols for other secondary cancers have to be developed gonadotropin-releasing hormone agonists. Cryopre-
and established in treatment guidelines. servation prior to therapy is another approach to
preserve fertility, but it has still to be regarded as an
experimental procedure.
Cardiac toxicity In summary, the most relevant acute treatment-
Treatment-related long-term cardiac toxicity com- related adverse effects are mucositis, bone marrow depres-
prises a wide spectrum of cardiac diseases such as sion, infections, polyneuropathy, and, less frequently,
coronary artery disease, cardiomyopathy, pericardial acute pneumonitis and acute cardiac toxicity. Delayed
disease, valvular disease, arrhythmia, and autonomic treatment-associated complications include the develop-
dysfunction. Among these cardiac abnormalities, cor- ment of secondary malignancy or cardiac disease, pulmo-
onary artery disease accounts for 66% of all cases of nary dysfunction, infections, infertility, hypothyroidism,
fatal cardiac events in HL survivors. Mediastinal irra- and fatigue. In particular, second malignancies and car-
diation was reported as the most important risk factor diovascular disease represent the leading causes of excess
for developing cardiopulmonary complications. mortality of long-term HL survivors. To reduce
Additionally, anthracycline-containing chemotherapy treatment-related toxicity, ongoing trials evaluate the
regimens are associated with an increased risk for development of response-adapted, less intensive treatment
congestive heart failure and myocardial infarction protocols. Regular cancer and cardiac screening tests
and may further increase radiation-related cardiac might also contribute to reduce mortality of HL survivors.
toxicity. Strategies to reduce treatment-associated car- Additionally, clinical research should focus on the preser-
diac disease focus on improving radiation therapy vation of quality of life of long-term HL survivors.
techniques and further reduction of dose and field
size. To detect subclinical cardiac abnormalities, non-
invasive cardiac screening tests such as resting and Summary
82 stress echocardiogram have been evaluated; a stand- Classical HL is a B-cell lymphoma characterized by the
ardized screening procedure regarding the required presence of HRS cells which derive from germinal center
Chapter 5: Hodgkin lymphoma
CR/PR
ABVD X 2
observe
+IFRT
Favorable Consider CR/PR
clinical trial
relapse
Early stage
2 + 2 + IF-RT (up to NR/PD
CS−IIA the age of 60 years)
w/o bulk ABVD × 4 + RT in the
Unfavorable case of High dose
contraindications. therapy + ASCT
NR/PD
Clinical staging +IFRT
Consider
clinical trial
CS IIA with
bulk, CSII−IV
NR/PD
Advanced relapse
stage
Polychemotherapy
6 courses of BEACOPP
escalated + RT for
residual disease (up to
the age of 60 years).
observe
Consider CR/PR
clinical trial
Figure 5.13 Treatment algorithm for classic Hodgkin Lymphoma. CR, Complete remission; PR, partial remission; NR, no response; PD,
progressive disease.
B-cells, but have lost most of the B-cell typical genes. treatment consisting of two courses of ABVD followed
Pathognomonic for the less frequent NLP-HL are L&H by 20 Gy IF-RT based on the results of the HD10 trial.
cells. Both variants share common features, i.e. the neo- The current standard of care for HL patients with
plastic cells are surrounded by a large number of reactive early-stage unfavorable disease within GHSG consists
non-clonal hematopoietic cells. In addition to a dereg- of two courses of BEACOPP-escalated followed by two
ulation of multiple signaling pathways and transcription courses of ABVD and 30 Gy IF-RT. In the case of
factors, the interaction of HRS cells with these inflam- contraindications, four cycles of ABVD followed by
matory cells seems to play an important role in the 30 Gy IF-RT can be given.
pathogenesis of classical HL. Patients with advanced-stage disease are treated
Treatment strategies for patients with HL are based more aggressively with six to eight courses of che-
on an individual risk factor assessment (Figure 5.13). motherapy. In many countries ABVD is still
The major prognostic factors used for risk stratifica- considered the standard regime for treatment of
tion include stage of disease, bulky disease, and advanced HL; based on the data of the HD9 trial
B-symptoms. According to the risk stratification and the HD15 trial, the GHSG regards six courses
applied by most European groups, including the of BEACOPP-escalated as standard.
GHSG and the EORTC, patients are assigned to To date, over 80% of all patients achieve long-term
early-stage favorable, early-stage unfavorable or disease-free survival after first line therapy. The poor
advanced-stage risk groups. prognosis of relapsed HL was markedly improved by
The recommended treatment for patients with an the introduction of effective high-dose chemotherapy 83
early-stage favorable disease is a combined modality followed by ASCT.
Chapter 5: Hodgkin lymphoma
Ongoing trials aim at reducing treatment- III–IV high-risk Hodgkin lymphoma: First results of
associated toxicity while maintaining or even improv- EROTC 20012 intergroup randomized phase III clinical
ing outcome. In addition to risk factor assessment at trial. J Clin Oncol, 2012 ASCO Annual Meeting. Abstract
8002.
diagnosis, response-adapted, 18FFDG–PET-guided
treatment approaches might be an important step DeVita VT, Jr., Hubbard SM. Hodgkin’s disease. N Engl J
towards an individualized HL treatment and thereby Med, 1993;328(8):560–565.
might help to reduce treatment intensity in those HL Diehl V, Stein H, Hummel M, et al. Hodgkin’s lymphoma:
patients with good response to chemotherapy. biology and treatment strategies for primary, refractory,
and relapsed disease. Hematology Am Soc Hematol Educ
The combination of chemotherapy/radiotherapy
Program, 2003:225–47.
with an effective targeted therapy such as brentuximab
vedotin might also contribute to decreased therapy- Eich HT, Diehl V, Gorgen H, et al. Intensified chemotherapy
and dose-reduced involved-field radiotherapy in patients
associated toxicity. Furthermore, other targeted therapy with early unfavorable Hodgkin’s lymphoma: final
approaches, including the HDAC inhibitor panobino- analysis of the German Hodgkin Study Group HD11 trial.
stat, and mTOR inhibitors such as everolimus or lena- J Clin Oncol, 2010;28(27):4199–4206.
lidomide might be effective tools to circumvent Engert A, Schiller P, Josting A, et al. Involved-field
chemotherapy resistance in relapsed or refractory HL. radiotherapy is equally effective and less toxic compared
with extended-field radiotherapy after four cycles of
chemotherapy in patients with early-stage unfavorable
Further reading Hodgkin’s lymphoma: results of the HD8 trial of the
Andrè M, Reman U, Federico M, et al., Interim analysis of German Hodgkin’s Lymphoma Study Group. J Clin
the randomized EORTC/LYSA/FIL intergroup H10 trial Oncol, 2003;21(19):3601–3608.
on early PET-scan driven adaptation in stage I/II Engert A, Franklin J, Eich HT, et al. Two cycles of
Hodgkin Lymphoma. Ash, 2012, 623:549. doxorubicin, bleomycin, vinblastine, and dacarbazine
Bonadonna G, Zucali R, Monfardini S, et al. Combination plus extended-field radiotherapy is superior to
chemotherapy of Hodgkin’s disease with adriamycin, radiotherapy alone in early favorable Hodgkin’s
bleomycin, vinblastine, and imidazole carboxamide lymphoma: final results of the GHSG HD7 trial. J Clin
versus MOPP. Cancer, 1975;36(1):252–259. Oncol, 2007;25(23):3495–3502.
Bonadonna G, Bonfante V, Viviani S, et al. ABVD plus Engert A, Diehl V, Franklin J, et al. Escalated-dose
subtotal nodal versus involved-field radiotherapy in BEACOPP in the treatment of patients with
early-stage Hodgkin’s disease: long-term results. J Clin advanced-stage Hodgkin’s lymphoma: 10 years of
Oncol, 2004;22(14):2835–2841. follow-up of the GHSG HD9 study. J Clin Oncol,
2009;27(27):4548–4554.
Borchmann P, Haverkamp H, Diehi V, et al. Eight cycles of
escalated-dose BEACOPP compared with four cycles of Engert A, Borchmann P, Pluetschow A, et al. Dose-
escalated-dose BEACOPP followed by four cycles of escalation with BEACOPP escalated is superior to ABVD
basline-dose BEACOPP with or without radiotheraphy in in the combined-modality treatment of early unfavorable
patients with advanced-stage Hodgkin’s lymphoma: final Hodgkin lymphoma: final analysis of the German
analysis of the HD12 trial of the German Hodgkin Study Hodgkin Study Group (GHSG) HD14 trial. Blood,
Group. J Clin Oncol, 2011;29(32):4234–4242. 2010;116(21):336.
Brauninger A, Wacker HH, Rajewsky K, et al. Typing the Engert A, Kobe C, Markova J, et al. Assessment of residual
histogenetic origin of the tumour cells of lymphocyte- bulky tumor using FDG-PET in patients with advanced-
rich classical Hodgkin’s lymphoma in relation to tumour stage Hodgkin lymphoma after completion of
cells of classical and lymphocyte-predominance chemotherapy: final report of the GHSG HD15 trial.
Hodgkin’s lymphoma. Cancer Res, 2003;63:1644–1651. Blood, 2010;116(21):336.
Brink AA, Oudejans JJ, van den Brule AJ et al. Low p53 and Engert A, Plutschow A, Eich HT, et al. Reduced treatment
high bcl-2 expression in Reed-Sternberg cells predicts intensity in patients with early-stage Hodgkin’s
poor clinical outcome for Hodgkin’s disease: involvement lymphoma. N Engl J Med, 2010;363(7):640–652.
of apoptosis resistance? Mod Pathol, 1998;11:376–383. Fan G, Kotylo P, Neiman RS, Braziel RM. Comparison of
Buettner M, Greiner A, Avramidou A, et al. Evidence of fascin expression in anaplastic large cell lymphoma and
abortive plasma cell differentiation in Hodgkin and Reed- Hodgkin disease. Am J Clin Pathol, 2003;119:199–204.
Sternberg cells of classical Hodgkin lymphoma. Hematol Fan Z, Natkunam Y, Bair E, et al. Characterisation of variant
Oncol, 2005;23:127–132. patterns of nodular lymphocyte predominant Hodgkin
84 Carde PP, Karrasch M. Fortpied C, et al. ABVD (8 cycles) lymphoma with immunohistologic and clinical
versus BEACOPP (4 escalated cycles ≥ baseline) in stage correlation. Am J Surg Pathol, 2003;27:1346–1356.
Chapter 5: Hodgkin lymphoma
Federico M, Luminari S, Iannitto E, et al. ABVD compared MacLennan KA, Bennett MH, Vaughan HB, Vaughan HG.
with BEACOPP compared with CEC for the initial Diagnosis and grading of nodular sclerosing Hodgkin’s
treatment of patients with advanced Hodgkin’s disease: a study of 2190 patients. Int Rev Exp Pathol,
lymphoma: results from the HD2000 Gruppo Italiano per 1992;33:27–51.
lo studio dei linfomi Trial. J Clin Oncol, 2009;10;27 Mauz-Korholz C, Hasenclever D, Dorffel W, et al.
(5):805–811. Procarbazine-free OEPA-COPDAC chemotherapy in
Fermé C, Eghbali H, Meerwaldt JH, et al. Chemotherapy plus boys and standard OPPA-COPP in girls have comparable
involved-field radiation in early-stage Hodgkin’s disease. effectiveness in pediatric Hodgkin’s lymphoma: the
N Engl J Med, 2007;357(19):1916–1927. GPOH-HD-2002 study. J Clin Oncol, 2010;28
Garcia JF, Camacho FI, Garcia JF et al. Hodgkin and Reed- (23):3680–3686.
Sternberg cells harbour alterations in the major tumor Montalban C, Garcia JF, Abraira V et al. Influence of
suppressor pathways and cell-cycle checkpoints: analysis biologic markers on the outcome of Hodgkin’s
using tissue microarrays. Blood, 2003;101:681–689. lymphoma: a study by the Spanish Hodgkin’s Lymphoma
Gordon LI, Hong F, Fisher RI, et al. A randomized phase III Study Group. J Clin Oncol, 2004;22:1664–1673.
trial of ABVD vs. Stanford V +/- radiation therapy in Morschhauser F, Brice P, Ferme C, et al. Risk-adapted
locally extensive and advanced stage Hodgkin’s salvage treatment with single or tandem autologous stem-
lymphoma: an intergroup study coordinated by the cell transplantation for first relapse/refractory Hodgkin’s
Eastern Cooperative Oncology Group (E2496). Blood, lymphoma: results of the prospective multicenter H96
2010;116(21):185. trial by the GELA/SFGM study group. J Clin Oncol,
Hay A, Klimm B, Chen BE, et al., Treatment of Stage I–II A 2008;26(36):5980–5987.
Non-Bulky Hodgkin’s Lymphoma (HL): An Individual Ng AK, Mauch PM. Late complications of therapy of
Patient-Data Comparison of German Hodgkin Study Hodgkin’s disease: prevention and management. Curr
Group (GHSG) HD10 and HD11 Combined-Modality Hematol Rep, 2004;3(1):27–33.
Therapy (CMT) and NCIC Clinical Trials Group (NCIC Nogova L, Reineke T, Brillant C, et al. Lymphocyte-
CTG) HD.6 ABVD Alone. Blood (ASH Annual Meeting predominant and classical Hodgkin’s lymphoma: a
Abstracts), 2012;623:548. comprehensive analysis from the German Hodgkin Study
Jaffe ES, Stein H, Swerdlow SH, et al. B cell lymphoma, Group. J Clin Oncol, 2008;26(3):434–439.
unclassifiable, with features intermediate between diffuse Noordijk EM, Carde P, Dupouy N, et al. Combined-
large B cell lymphoma and classical Hodgkin lymphoma. modality therapy for clinical stage I or II Hodgkin’s
In Swerdlow SH, Campo E, Harris NL, et al. (eds). WHO lymphoma: long-term results of the European
Classification of Tumours of Haemopoietic and Lymphoid Organisation for Research and Treatment of Cancer H7
Tissues. Lyon: IARC, 2008:267–268. randomized controlled trials. J Clin Oncol,
Josting A, Rudolph C, Mapara M, et al. Cologne high-dose 2006;24(19):3128–3135.
sequential chemotherapy in relapsed and refractory Press OW, LeBlanc M, Lichter AS, et al. Phase III
Hodgkin lymphoma: results of a large multicenter study randomized intergroup trial of subtotal lymphoid
of the German Hodgkin Lymphoma Study Group irradiation versus doxorubicin, vinblastine, and subtotal
(GHSG). Ann Oncol, 2005;16(1):116–123. lymphoid irradiation for stage IA to IIA Hodgkin’s
Josting A, Muller H, Borchmann P, et al. Dose intensity of disease. J Clin Oncol, 2001;19(22):4238–4244.
chemotherapy in patients with relapsed Hodgkin’s Santoro A, Magagnoli M, Spina M, et al. Ifosfamide,
lymphoma. J Clin Oncol, 2010;28(34):5074–5080. gemcitabine, and vinorelbine: a new induction regimen
Klimm B, Goergen H, Fuchs M, et al. Risk factors in early- for refractory and relapsed Hodgkin’s lymphoma.
stage Hodgkin Lymphoma effectively define an Haematologica, 2007;92(1):35–41.
unfavorable subgroup with inferior outcome: an analysis Sasse S, Klimm B, Goergen H, et al. Comparing long-term
of international staging definitions (publication in toxicity and efficacy of combined modality treatment ng
progress). extended- or involved-field radiotherapy in early-stage
Linch DC, Winfield D, Goldstone AH, et al. Dose Hodgkin’s oma. Ann Oncol. 2012;23(11):2953–2959.
intensification with autologous bone-marrow Schmid C, Pan L, Diss T, Isaacson PG. Expression of B-cell
transplantation in relapsed and resistant Hodgkin’s antigens by Hodgkin’s and Reed-Sternberg cells. Am J
disease: results of a BNLI randomised trial. Lancet, Pathol, 1991;139:701–707.
1993;341(8852):1051–1054.
Schmitz N, Pfistner B, Sextro M, et al. Aggressive
Lorigan P, Radford J, Howell A, et al. Lung cancer after conventional chemotherapy compared with high-dose
treatment for Hodgkin’s lymphoma: a systematic review. chemotherapy with autologous haemopoietic stem-cell
Lancet Oncol, 2005;6(10):773–779. transplantation for relapsed chemosensitive Hodgkin’s 85
Chapter 5: Hodgkin lymphoma
86
Chapter
6
Follicular lymphoma
Kristian Bowles, Daniel Hodson, and Robert Marcus
Pathology: Andrew Wotherspoon Molecular cytogenetics: Andreas Rosenwald and German Ott
Lymphoma, Second Edition, ed. Robert Marcus, John W. Sweetenham, and Michael E. Williams. Published by 87
Cambridge University Press. © Cambridge University Press 2014.
Chapter 6: Follicular lymphoma
Immunophenotypically the cells of FL express the cervical), Waldeyer’s ring, testis, or other extranodal
pan-B-cell markers CD19, CD20, CD22, CD79a, and sites, as is usually at an early stage. These lymphomas
PAX5 and have surface immunoglobulin (sIg). In the often have large follicles, are more frequently grade 3
majority of cases they have a germinal center cell and lack BCL-2 expression. There is an appreciation
phenotype with expression of CD10 and BCL-6 pro- that lymphomas that arise in the intestine, predomi-
tein. A proportion of cases show loss of CD10. In nantly duodenum, behave differently from nodal FL.
many cases the neoplastic cells in the interfollicular They present frequently as polyps and may involve
compartment show downregulation of expression of multiple sites within the intestine but infrequently dis-
CD10 and BCL-6. Recent gene expression analysis has seminate (stage IE or IIE) and are associated with an
highlighted expression of germinal-center B-cell- excellent prognosis. Other extranodal FL may also rep-
expressed transcript (GCET) as a characteristic of fol- resent a variant with predominantly localized disease.
licle center cells. There is usually no staining for Primary cutaneous follicle center lymphoma is a
MUM1, although rare cases may express this antigen, distinct and separate entity from nodal and non-cuta-
usually in cases with higher numbers of centroblasts. neous extranodal FL.
In contrast to reactive germinal centers, the cells of FL It has recently been recognized that some nodes
express BCL-2 protein in the majority (about 85%) of that have preserved architecture can harbor neoplastic
cases. This is more frequently the case in grade 1–2 follicle center cells occupying some or all of the ger-
lymphoma (85–90%) but is found in only 50% of grade minal center in a variable proportion of the follicles.
3 FLs. A proportion of cases express CD23 but the cells While this pattern may be seen in nodes from patients
are negative for CD5 and cyclin D1. The neoplastic with concurrent or prior classical FL, it is also encoun-
follicles contain an FDC meshwork that is highlighted tered in a minority of patients with no history of
by staining for CD21 or CD23 and contains a variable lymphoma. In these cases the term intrafollicular neo-
component of reactive T-cells, including those of fol- plasia or in-situ FL has been used. The atypical cell
licular T-helper type and histiocytes (Figure 6.5). population is composed of centrocytes for the most
Proliferation is variable but is generally low in grade part and is highlighted by the presence of strong stain-
1 and higher in cases that are grade 3. The proliferation ing for BCL-2 protein. The significance of the finding
does not show the organized zoned pattern of reactive of in-situ FL is unknown.
follicles.
The 4th edition of the WHO classification recog-
nizes clinicopathological variants of classical FL deter- Molecular and cytogenetics
mined by age or site at presentation. Pediatric FL The cytogenetic hallmark of FL, the t(14;18)(q32;q21)
characteristically involves lymph nodes (particularly chromosome translocation, is present in roughly 85%
A B
Figure 6.5 Follicular lymphoma stained for BCL-2 protein (lymph node). The cells of follicular lymphoma show strong staining for BCL-2 89
protein (a) in distinction to the cells of reactive germinal centers, which are (with the exception of the intrafollicular T-cells) BCL-2-negative (b).
Chapter 6: Follicular lymphoma
of grade 1 and 2 tumors. By virtue of this transloca- the concept of varied cytogenetic pathways central to
tion, which is easily detectable by fluorescence in-situ pathogenesis has emerged from the analysis of secon-
hybridization (FISH) techniques, the BCL-2 proto- dary chromosomal alterations defined by the acquisi-
oncogene is juxtaposed to enhancer sequences of the tion of aberrations like 6q–, +7, or der(18)t(14;18). The
IgH gene promoter region leading to BCL-2 deregula- presence of a deletion in the long arm of chromosome 6
tion and overexpression in the neoplastic follicles. This (6q–) has been associated with inferior outcome in FL.
rearrangement occurs during the recombination of FL frequently progresses to DLBCL, and this event
variable (V), diversity (D), and joining (J) regions in seems to be associated with certain recurring chromo-
precursor B-cells. In reactive follicles, the BCL-2 pro- some aberrations. Transformed FLs frequently acquire
tein is not expressed, thus enabling the removal of deletions in the short arms of chromosomes 9 and 17,
B-cells with non-functional B-cell receptors via apop- the sites of the p16INK4A and TP53 tumor suppressor
tosis. The constitutive activation of the antiapoptotic genes. The additional introduction of a t(8;14)(q24;
BCL-2 via the t(14;18) bypasses this physiological q32)/MYC rearrangement in FL leads to a highly
mechanism and, hence, FL may be viewed as a disease aggressive “Burkitt-like” neoplasia, sometimes with a
of programmed cell death deregulation. A relatively precursor B-cell phenotype.
new, but highly controversial, finding is the presence FL grade 3A, in accordance with its morphological
of the translocation in FL samples not only in the and immunophenotypical similarities to FL grades 1
neoplastic lymphoid cells, but also in microvascular and 2, is also frequently t(14;18)-positive, although to
endothelial cells within the tumor. More recently, a slightly lesser extent than its grade 1 and 2 counter-
t(14;18) has also been demonstrated in histiocytic or parts. In contrast, FL3B is infrequently rearranged for
dendritic cell neoplasms from patients simultaneously BCL-2 and BCL-6, and 50% of DLBCL with an addi-
suffering from t(14;18)-positive FL, thus providing tional FL3B component harbor BCL-6 rearrange-
evidence for the transdifferentiation capacity of the ments. FLs occurring in the pediatric population, as a
neoplastic FL clone. However, according to data rule, are t(14;18)-negative and generally present as
obtained from mouse models the introduction of a localized disease and with higher histological grade.
t(14;18) in a lymphoid cell may not be sufficient to Recently, it was shown that t(14;18)-negative FLs of
cause a lymphoma on its own. Instead, this event may grades 1 and 2 do display a gene expression profile
rather constitute a prerequisite for the acquisition of that, although maintaining the typical global signature
secondary chromosomal aberrations by prolonging of FL, differ from their t(14;18) positive counterparts.
the lifespan of a germinal center B-cell. In this respect, Gene expression profiling emerged as a powerful
it is highly interesting that circulating t(14;18)-positive tool to identify robust prognostic subgroups in FL. The
B-cells can be found in the peripheral blood of analysis of a large cohort of FL resulted in the definition
up to 70% of healthy individuals, and that cases of of two gene expression signatures with highly predictive
“in-situ” FLs have been described in which clonal potential. One of these signatures, termed “immune
t(14;18)-positive “FL-like” cells partly colonize reac- response 1 signature,” was associated with a favorable
tive follicles in lymph nodes. The significance of both clinical course, while the “immune response 2 signature”
findings towards the development of overt FL is not conferred inferior survival times. As suggested by their
yet clear; however, in both conditions the clonal names, both signatures do not appear to be derived from
FL-like B-cells may carry the t(14;18) as the sole aber- the malignant B-cells, but instead from reactive tumor-
ration, not yet having acquired additional genetic infiltrating cells. In particular, the immune response 1
lesions. signature may primarily be derived from certain T-cell
Secondary chromosome aberrations are well subsets and the immune response 2 signature from
defined in FL. In particular, (partial) trisomies of chro- macrophages and dendritic cells. A mathematical
mosomes 1q, 7, 12, 18, and X as well as deletions in the model combining these two signatures predicts that the
short arm of chromosome 1 (1p) and the long arm of relative levels of subsets of these cells in the lymphoma
chromosome 6 (6q) have been reported. The acquisi- specimen are crucial for the survival times in FL, arguing
tion of these secondary chromosomal alterations for a prominent role of immunologic cross-talk between
appears to be non-random, and certain aberration cas- the malignant B-cells and infiltrating bystander cells.
cades may be determined by genetic alterations in the These findings were recently confirmed by a RT-PCR-
90 early steps of progression. In a large cohort of FL cases, based gene expression profiling attempt.
Chapter 6: Follicular lymphoma
Table 6.2 Overall survival according to risk group defined by the Follicular Lymphoma International Prognostic Index
(FLIPI).
the original field of radiation and usually do so within disease-specific survival (hazard ratio [HR] 0.65) and
the first 15 years. Beyond this time relapses are OS (HR 0.73). In the absence of data from a prospec-
extremely rare. tive randomized study these data suggest that upfront
In a series of 460 patients with stage I/II FL treated radiotherapy should remain the standard of care
with IF-RT, DFS rates were 56% at 5 years and 41% at for early stage FL. Historically a radiation dose of
10 years. Beyond 15 years only 2% of patients relapsed. 30–45 Gy has been used. However, a recent rando-
As expected, the majority of relapses occurred outside mised phase III study by Lowry et al. compared 24Gy
the original field of radiation. Similar results were seen with 40–45Gy in patients requiring local control of
in a series of 80 patients from M.D. Anderson treated their disease. This revealed no difference in response
with IF-RT alone. PFS at 15 years was 66% for stage I rate, in-field progression, PFS or OS making 24 Gy the
patients and 26% for stage II. Beyond 17 years no current recommended dose. Based upon reports of
relapses were seen, suggesting that this approach may high response rates in patients with relapsed FL treated
be curative in some patients. with just 4 Gy IF-RT a randomized UK trial compared
In an attempt to improve long-term PFS the role of 24 Gy with 4 Gy IF-RT. This trial has been closed early
adjuvant chemotherapy has been examined in single due to inferior local PFS in the 4Gy arm. Although
arm studies. Seymour reported a 10-year time to treat- lower doses may be useful in the context of palliation
ment failure (TTF) of 72% in stage I–III FL patients 24 Gy remains the standard of care for IF-RT delivered
treated with 10 cycles of CVP-bleo (cyclophospha- with curative intent.
mide, vincristine, prednisolone, and bleomycin) and Patients with fully resected stage I disease have a
IF-RT. Whilst these results are superior to historical good prognosis, with up to 80% achieving long-term
controls treated with IF-RT there was a significant remission. It is not known whether radiotherapy con-
increase in secondary malignancy. No randomized fers any additional advantage or whether PET–CT
trial has compared IF-RT with combined modality may identify those requiring further treatment.
treatment. The Trans Tasman Radiation Oncology
Group is currently conducting such a study to examine Summary
IF-RT ± six cycles of CVP plus Rituximab. Clearly any
At present 24Gy IF-RT is considered standard of care
improvement in disease control must be balanced
for small volume stage I and II FL. In around 40% of
against immediate and late toxicity of this approach
such patients, radiotherapy will be curative. In com-
and in particular the risk of secondary malignancy.
pletely excised stage I disease the toxicity of treatment
The efficacy of rituximab in advanced-stage FL, com-
may outweigh the potential reduction in risk of
bined with its low toxicity, suggests that immunother-
recurrence.
apy may ultimately prove beneficial in early-stage
disease, either alone or in combination with radio-
therapy. In retrospective analysis the addition of
Rituximab to IFRT has been reported to improve
Treatment of advanced disease
PFS. However, this approach is yet to be validated by in asymptomatic patients
a prospective or randomized trial. The majority of patients with FL present with advanced
Because of the indolent nature of the disease disease and many of these are asymptomatic at diagno-
and concerns over late treatment-related toxicity sis. In contrast to early-stage FL conventional treatment
other researchers have advocated the approach of “no is not curative and the aim of treatment is disease
initial therapy” in early-stage disease. No prospective, control. Three randomized trials, conducted by the
randomized trial compares the use of upfront IF-RT NCI, GELA, and BNLI, have confirmed that early
with watchful waiting. A retrospective analysis of the treatment with conventional chemotherapy offers no
Surveillance, Epidemiology and End Results Database survival advantage in asymptomatic patients with FL.
included 6568 patients with early-stage FL diagnosed Young and colleagues randomized 99 patients with
between 1973 and 2004. The authors compared the indolent NHL to watchful waiting versus aggressive
long-term outcomes of patients who received initial combined modality treatment with ProMACE–MOPP
radiotherapy with those who did not. Multivariate (prednisone, methotrexate, doxorubicin, cyclophos-
analysis showed a clear and independent association phamide, etoposide plus mechlorethamine, vincristine,
92 between upfront radiation and improvement in both procarbazine, and prednisone) followed by total nodal
Chapter 6: Follicular lymphoma
control arm but was not reached at 47 months in the were 46%, 62%, and 59% for the respective treatment
rituximab arm. Most notably the addition of rituximab groups with significantly higher toxicity in those receiv-
to MCP in patients with FL was associated with a sig- ing R-FM. Overall survival at three years was 95% for all
nificantly improved OS at 4 years (87% versus 74%; patients. A meta-analysis conducted by Itchaki and col-
P = 0.01). Finally, the FL-200 study conducted by leagues concluded that the addition of anthracycline to
GELA randomized 359 patients with previously either chemotherapy or immunochemotherapy was
untreated FL and high tumor burden to 12 cycles of associated with no improvement in OS. Rummel and
CHVP/interferon (cyclophosphamide, doxorubicin, eto- colleagues have compared R-CHOP and R-bendamus-
poside, prednisolone, and alpha-interferon) or six cycles tine in a study of 549 patients with grade 1/2 FL and
of CHVP/interferon with six doses of rituximab. mantle cell lymphoma. Patients were randomized to
Patients receiving rituximab had a significantly higher receive six cycles of R-CHOP or R-bendamustine (90
response rate at 6 months (ORR 94% versus 85%; com- mg/m2 day 1+2 every 28 days). Improvement in median
plete remission/complete remission – unconfirmed (CR/ PFS (69.5 months versus 31.2 months) was seen in the R-
Cru; 76% versus 49%) and prolonged EFS at 5 years bendamustine arm. In addition to its superior PFS, R-
(53% versus 37%). A prolongation of OS at 5 years was bendamustine was better tolerated and showed reduced
seen in patients with FLIPI risk score 3–5 but did not toxicity compared to R-CHOP. Response by FLIPI score
achieve statistical significance in low-risk patients. was not presented and it is not known if the benefit
The survival benefit of adding rituximab to chemo- extends to all prognostic subgroups. OS was not different
therapy for first line treatment of FL has been exam- at the time of analysis and longer follow-up is awaited. It
ined in a recent Cochrane meta-analysis. Of note, this is not clear from current evidence if bendamustine is
combined survival data from the above randomized optimally used upfront or better reserved for later relap-
trials was composed prior to the updated survival data ses where it is also associated with high response rates.
from the GELA and Marcus trials. Even without this All patients who require treatment for stage III and
data the analysis favored immunochemotherapy with IV FL should receive remission induction treatment of
a hazard ratio of 0.63 (95% CI: 0.51–0.79). rituximab combined with chemotherapy (e.g. CVP,
Whilst the benefit of adding rituximab to chemo- CHOP, bendamustine). In those unable to tolerate che-
therapy is established, the optimal chemotherapy regi- motherapy, single agent rituximab should be considered.
men to use remains unclear. A randomised trial
comparing R-CVP versus R-CHOP versus R-FM for Rituximab maintenance
the initial treatment of patients with advanced-stage The benefit of rituximab as maintenance therapy
follicular lymphoma appeared to favour R-CHOP. was originally shown following treatment for
After a median follow-up of 34 months, 3-year TTFs relapsed FL. It has also been shown following first 95
Chapter 6: Follicular lymphoma
line treatment in patients who received single agent favorable response to salvage treatment in the control
rituximab and in patients who received chemother- arm meant that no difference in OS was observed.
apy without rituximab. Until recently it was Importantly, all these studies predate the introduction
unclear whether patients receiving first line immu- of rituximab maintenance.
nochemotherapy would benefit from rituximab
maintenance. This question was addressed by the Radioimmunotherapy
GELA sponsored PRIMA study. The study Radioimmunotherapy (RIT) has shown impressive results
randomized 1217 patients who had achieved a min- in the first line treatment of FL. Kaminski treated 76
imum of partial remission (PR) following treat- previously untreated patients with stage III/IV FL with
ment with immunochemotherapy (either R-CVP, 131
I-tositumomab and reported a 95% response rate (75%
R-CHOP, or R-FCM [rituximab, fludarabine, CR) and PFS of 59% and 40% at 5 and 10 years, respec-
cyclophosphamide, and mitoxantrone]) to mainte- tively. For those patients achieving a CR the median PFS
nance therapy with rituximab (375 mg/m2 every 8 was 10.9 years. One case of secondary myelodysplastic
weeks for 2 years) or no further therapy. At 3 years syndrome (MDS) was reported. Although these results
maintenance rituximab was associated with pro- seem impressive it is important to note that the median
longed PFS (74.9% versus 57.6%). No difference age of the patients included in this study was only 49
was observed in OS and quality of life assessment years. No randomized trial has compared RIT with the
was similar in the maintenance and control arms. current standard of care – immunochemotherapy.
Little additional toxicity was seen in the mainte- RIT has also been examined in the context of post-
nance group. Longer follow-up data, particularly induction consolidation. In a SWOG study 90 patients
that addressing OS and the response to salvage were treated with six courses of CHOP chemotherapy
therapy, is currently awaited. A similar effect was followed by 131I-tositumomab consolidation and
seen in a retrospective population-based analysis of reported a response rate of 91% (69% CR) and 5-year
patients treated before and after the introduction of PFS of 67%. This compared favorably to their historical
maintenance rituximab as standard in British controls treated with CHOP alone. A single phase III
Columbia; at 3 years PFS was 62% and 83%, multicenter randomized controlled trial – the FIT
respectively, with no identifiable change in OS. study – compared consolidation with 90Y-ibritumomab
tiuxetan to no further treatment following first line
Autologous stem cell transplant chemotherapy in 414 patients with advanced-stage FL.
(ASCT) in first remission RIT consolidation resulted in high (77%) PR-to-CR
conversion, leading to a final CR rate of 87%. In those
Four prospective randomized trials have examined the randomized to receive RIT, median PFS was prolonged
benefit of ASCT in first remission. The GELA study by 2 years compared to the observation group. After 5
compared ASCT with IFN following conventional years of follow-up, PFS (RIT versus observation) was
induction chemotherapy and found no difference in 57% versus 43% in patients achieving a CR and 38%
either PFS or OS. The GLSG compared ASCT with versus 14% in patients achieving a PR following induc-
IFN maintenance following induction therapy with tion therapy. As only 15% of patients received rituximab
CHOP or MCP. Autografting was associated with as a component of their induction therapy it remains to
prolonged PFS at 5 years (64.7% versus 33.3%) be determined whether RIT consolidation improves PFS
although data for OS was not presented. The updated in patients treated with immunochemotherapy and
GOELAMS study also showed improved PFS in whether consolidation with RIT is superior to mainte-
patients randomized to ASCT (64% versus 39% at 9 nance therapy with rituximab. No significant difference
years). However, partly because of increased toxicity in in OS was observed and at 5 years a trend towards
the ASCT arm, this failed to translate into a survival increased risk of secondary MDS was seen (six cases in
advantage (80% versus 81% at 9 years). The GITMO the RIT arm versus one case in the observation arm).
study was the only one to examine the use of ASCT
following the use of upfront rituximab. This study
confirmed that the improvement in PFS associated
Summary: first line treatment
with ASCT is still seen after rituximab (68% versus of advanced lymphoma
96 31% at 4 years). However, as with previous studies the Observation alone remains an appropriate strategy
combination of increased toxicity in the ASCT and the for asymptomatic patients with advanced FL. Those
Chapter 6: Follicular lymphoma
patients who do require treatment should receive an induced by R-chemo, it seems reasonable to retreat
alkylator-based regimen in combination with rituxi- with rituximab.
mab. The inclusion of an anthracycline improves Bendamustine appears to be an increasingly prom-
response rate but does not improve OS. There is some ising drug in relapsed FL. In a phase III randomized
evidence that the efficacy of R-bendamustine may be at trial from Rummel and colleagues of relapsed indolent
least as good as, if not better than, R-CHOP, and the lymphoma therapy (46% of patients had FL), benda-
toxicity lower. The PRIMA study suggests that main- mustine–rituximab (BR) was compared with fludara-
tenance rituximab prolongs the duration of remission in bine–rituximab (FR). The BR combination was
patients induced with R-chemotherapy. ASCT is not associated with significantly improved ORR (83.5%
indicated in first remission. The use of RIT is associated versus 52.5%), CR rate (38.5% versus 16.2%), and
with prolonged remissions but its role in the era of median PFS (30 versus 11 months). Serious adverse
immunotherapy and rituximab maintenance has yet events were similar for the BR and FR groups (17.4%
to be defined. and 22.2%), and OS did not differ with a median
follow-up period of 33 months. Furthermore, in an
Management of relapsed unplanned subanalysis, rituximab maintenance ther-
apy significantly prolonged OS and PFS in a group of
follicular lymphoma 40 patients who received this treatment.
The management of patients with relapsed disease is Fludarabine is an effective agent in patients who
influenced by previous therapy, duration of remission, relapse following first line alkylator-based therapy. It
and performance status. Similar to the situation in showed improved PFS in comparison with CVP but did
patients with newly diagnosed FL, observation alone not significantly improve response rate or OS. In patients
may be appropriate for relapsed patients who are asymp- with refractory and relapsed FL response rates appear to
tomatic. The risk of transformation to high-grade lym- be improved by combining fludarabine with an anthracy-
phoma is approximately 20% at 5 years. Repeat biopsy is cline, an alkylating agent, or both. However, no direct
therefore important to exclude transformation at each comparison was made to alkylator therapy alone, and
recurrence. Restaging of the disease is also recommen- superiority in terms of OS has not been demonstrated.
ded at each relapse to establish the extent of the lym- Fludarabine treatment is also associated with significantly
phoma and to provide a baseline from which to define greater toxicity than alkylator- or anthracycline-based
response to therapy. There are a number of treatment regimens and its benefits as second line treatment must
options available to clinicians in the management of be weighed against the risks of infection, stem cell toxicity
symptomatic relapsed FL, but the heterogeneity of the (which may compromise future attempts at stem cell
patient population and specifically the nature of pre-
harvesting), and treatment-related MDS.
vious treatments and the duration of previous remis-
In relapsed patients unable to tolerate conventional
sions, combined with a lack of good quality randomized
chemotherapy, single agent rituximab induces a response
controlled trial data, means that treatment at relapse is
in approximately 50% of patients (who had not previously
far from standardized.
received antibody therapy), with a median TTF of 13
months. Single agent rituximab is also active in patients
Immunochemotherapy who have previously responded to rituximab therapy,
Several recent randomized trials confirm that the addi- with a response rate in this group of 40% and a median
tion of rituximab to chemotherapy in relapsed FL is time to progression estimated at 18 months.
associated with improvement in both response rate Although the benefit of rituximab seems clear the
and survival. Van Oers randomized 465 patients with choice of chemotherapy to use in combination will be
relapsed FL to R-CHOP or CHOP and saw improved influenced by, among other things, prior treatment (in
PFS (33.1 months versus 20.2 months) and OS (85% particular, anthracycline exposure), duration of first
versus 77%) at 3 years. Forstpointner randomized 65 remission, need for future stem cell harvest, and
patients with relapsed FL to R-FCM or FCM alone and patient fitness. Therapeutic options include alkylators,
also observed improved PFS and OS in the rituximab anthracyclines, bendamustine, and fludarabine, either
arm. Most of this data comes from patients not pre- alone or in combination. Of these, bendamustine
viously exposed to rituximab. However, where appears to combine efficacy with a favorable side effect 97
patients relapse following a long (>2 years) remission profile.
Chapter 6: Follicular lymphoma
Signaling through the B cell receptor is a requirement suggestion of a plateau in survival curves. A combined
for the survival of most normal B cells and many B cell series from St Bartholomews and Dana Faber reported a
lymphomas. A number of small molecule kinase plateau in OS at 48% at 12 years. Although most of these
inhibitors that target pathways downstream of the B studies recruited patients prior to the introduction of
cell receptor are currently in early stage clinical trials upfront rituximab, a recent case series has suggested
for FL. Of particular interest are Idelalisib (formerly that ASCT remains an effective treatment for relapsed
CAL-101) a selective inhibitor of the delta isoform of patients who have received prior rituximab.
PI3-Kinase and Ibrutinib, an irreversible inhibitor of The role of rituximab at the time of autograft for
Bruton's tyrosine Kinase. relapsed FL has been investigated in the Lym1 study.
This study shows that rituximab in vivo purging followed
by 2 years maintenance postautograft gives a superior PFS
High-dose therapy and autologous compared to controls, with a suggestion that both the
stem cell transplantation purging and the maintenance contributed to the overall
The European CUP (Chemotherapy Unpurged improvement. No benefit in OS has yet been reported.
Purged) trial is the only randomized control trial that Perhaps the most contentious issue is the optimum
has compared chemotherapy and ASCT in relapsed timing of ASCT for relapsed FL. The Rohatimer series
FL. Between 1993 and 1997, 89 patients with relapsed showed a clear association of improved survival with
or refractory FL who had achieved a CR or PR follow- patients receiving ASCT in second (as opposed to
ing induction therapy (the majority of patients treated later) remissions, suggesting that early ASCT is indi-
with CHOP), had a WHO performance status of 0–2, cated. However, this series (in line with others) showed
and had limited bone marrow infiltration (defined as a high rate of late treatment-related toxicity, with a
<20% B-cells in the marrow aspirate) were random- 12% mortality due to secondary MDS/AML. When
ized to receive three further chemotherapy courses, an taken in the context of the recent improvements in
unpurged ASCT or a purged ASCT. As the numbers in outcome for relapsed FL patients treated with ritux-
the three groups were small the two transplant groups imab and other newly introduced treatment modal-
(unpurged and purged) were combined and compared ities, this high toxicity means that ASCT may be best
with chemotherapy, and an OS benefit for ASCT was reserved for patients with early relapse, aggressive dis-
reported (Hazard ratio 0.40 [range 0.18–0.89], ease, and those who have failed antibody treatment.
P = 0.026), with an OS at 2 years of 71% in the ASCT The exact role for ASCT remains to be defined in
group and 46% in the chemotherapy group. randomized clinical trials.
Importantly, this trial was conducted prior to the
introduction of rituximab to first line therapy. Allogeneic stem cell transplantation
A retrospective analysis of 100 consecutive patients
with relapsed or refractory FL treated with HDCT/ in FL (allo-SCT)
ASCT from 1993 to 2008 in Calgary showed an EFS Allogeneic stem cell transplant (allo-SCT) is a poten-
and OS at 5 years of 57% and 71%; at 10 years these are tially curative therapy for relapsed FL and appears an
projected to be 55% and 63%, respectively. With a attractive proposition, particularly in a selected group
median follow-up of 63 months, severe toxicity of younger patients. However, in contrast to other hem-
included two treatment-related deaths and four deaths atologic malignancies where relapsed disease is often
from secondary acute myeloid leukemia (AML)/MDS. untreatable, relapsed FL may still be amenable to treat-
The Calgary group observed improvement in EFS for ment with many other less toxic lines of therapy, includ-
patients treated since 2000 in their cohort, which they ing novel agents. This fact must be weighed against the
hypothesized may be related to the more frequent use of significant morbidity and mortality attached to allo-
rituximab within 6 months of ASCT since 2000, either SCT, and the timing of transplant is a key clinical
with stem cell mobilization or preceding reinduction decision. While historically much of the decreased dis-
therapy, reflecting an “in vivo” purging of autografts. ease recurrence seen after allo-SCT appears to have
In addition, a number of phase II studies suggest an been offset by the toxicity of the treatment, the risk–
apparent survival benefit for relapsed FL treated with benefit ratio for selected patients may be becoming
ASCT when compared to historical controls. In some more favorable as outcomes of allo-SCT for FL appear
studies with prolonged follow-up there is also a to be showing improvement over the past decade. 99
Chapter 6: Follicular lymphoma
Unfortunately the absence of good trial evidence in about 50% of patients with a median duration of 6
makes it difficult to make specific recommendations in months. As with RIT following first line FL treatment,
the area of allo-SCT. There is limited data available to in relapsed disease RIT remission consolidation has not
guide patient selection and the optimum timing of been directly compared with rituximab maintenance
transplantation in the natural history of disease. Case strategies.
series and registry data of allo-SCT suggest that a The principal toxicities of RIT are neutropenia and
reduced intensity conditioning approach presently thrombocytopenia, and it is necessary that patients
appears to be preferable to myeloablative conditioning have <25% marrow tumor infiltration and good tri-
regimens. The data also suggest that, while a sibling lineage hematopoiesis with well-preserved platelet
donor for a reduced intensity conditioning (RIC)-allo- counts to have RIT. Treatment on the whole, however,
SCT is preferable, a suitable unrelated HLA-compat- is well tolerated and may be most appropriate in
ible donor can result in similar outcomes. patients considered unfit for standard immunochemo-
Chemorefractory disease at time of transplant consis- therapy regimens. Careful consideration should be
tently appears as a poor prognostic factor in a number given before treating patients who have received mul-
of publications. An overview of the increasing num- tiple prior therapies, particularly fludarabine. There
bers of studies reporting results from a number of also remain concerns about the long-term risks of
different conditioning regimens, in heterogeneous MDS/AML in patients treated with RIT, particularly
groups of what appear to be generally poor risk those who have had previous purine analog therapy.
patients under the age of 65, suggests that non-relapse In relapsed FL, RIT appears effective and generally
mortality rates vary between about 15% and 35% well tolerated. In selected patients, such as those not
depending on the study. considered fit for immunochemotherapy or those who
are refractory or relapse early following rituximab-con-
Summary of allogeneic transplantation taining regimens, RIT may be highly efficacious.
Allogeneic stem cell transplant should be discussed with
younger patients with aggressive disease who have
relapsed or progressed despite conventional treatments
High-grade transformation of FL
(chemotherapy, antibody therapies, and ASCT), and Transformation of FL to high-grade lymphoma
who are considered otherwise fit enough for the proce- appears to be an early event in the course of the disease
dure. The significant toxicities mean that allogeneic and is mainly observed in patients with known adverse
transplantation will be an option for only a minority prognostic factors or those who do not achieve CR
of patients. The optimum conditioning regimen is after initial treatment. Furthermore, transformed
unclear but reduced intensity strategies appear to be FL carries a poor prognosis despite treatment with
associated with improved outcomes compared with aggressive chemotherapy. In a retrospective analysis
myeloablative conditioning protocols, at least in part of 220 patients over 15 years, Bastion and colleagues
because of lower non-relapse mortality. A suitable unre- reported a probability of transformation of 22%
lated HLA-matched donor appears to be a reasonable at 5 years and 31% at 10 years, and observed a
alternative if an HLA-matched sibling donor is not plateau after 16 years. Predictive factors for transfor-
available. Allogeneic transplantation remains an exper- mation were non-achievement of CR after initial ther-
imental therapy and patients should be treated where apy, low serum albumin level (<35 g/L), and beta
possible within the context of a clinical trial. 2-microglobulin level greater than 3 mg/L at diagnosis.
Transformation accounted for 44% of deaths and was
associated with a poor outcome, with a median
Radioimmunotherapy survival time of 7 months.
In patients with relapsed and refractory FL RIT shows Conflicting data exists as to whether the timing or
significant clinical activity, with remission rates nature of first line therapy influences the risk of high-
reported to be similar to those seen with R-CHOP. In grade transformation. In contrast to other published
relapsed disease durable (>18-month) remissions have reports, a case series from British Columbia observed a
been achieved in about 50% of patients with iodine-131 significant correlation between initial treatment regi-
rituximab radioimmunotherapy. Furthermore, in men and risk of high-grade transformation. This series
100 patients with rituximab-refractory FL, treatment with included patients treated on two serial phase II studies;
Y90-ibritumomab has been shown to induce a response the first cohort received multiagent chemotherapy
Chapter 6: Follicular lymphoma
(bleomycin, cisplatin, etoposide, doxorubicin, cyclo- Patients with aggressive disease and those who relapse
phosphamide, vincristine, and prednisolone) followed early after immunochemotherapy should be considered
by 25 Gy IF-RT, while the second cohort received the for RIT or ASCT. Younger patients who relapse despite
combination of an alkylator and a purine analog. The the above should be considered for an allogeneic trans-
10-year risk of transformation in these two cohorts plant. Where possible, patients should be managed in
was 18% and 30%, respectively (P = 0.001). No differ- the context of a clinical trial.
ence in OS was seen. The authors hypothesized that
patients with advanced stage FL may, at diagnosis,
already harbor areas of occult transformation that Further reading
may be eradicated by a more aggressive initial treat- Al-Tourah AJ et al. Population-based analysis of incidence
ment regimen. and outcome of transformed non-Hodgkin’s lymphoma.
Data is limited as to the most appropriate course of J Clin Oncol, 2008;26:5165–5169.
treatment in transformed FL. Anthracycline-based Ardeshna KM et al. Long-term effect of a watch and wait
combinations (e.g. CHOP or FMD) should be used policy versus immediate systemic treatment for
asymptomatic advanced-stage non-Hodgkin lymphoma:
to induce remission induction. Intuitively, it seems
a randomised controlled trial. Lancet, 2003;362:516–522.
reasonable to add rituximab to the remission induc-
tion regimen, although data is lacking. In the event of Bastion Y et al. Incidence, predictive factors, and outcome of
lymphoma transformation in follicular lymphoma
an inadequate response, previous anthracycline expo-
patients. J Clin Oncol, 1997;15:1587–1594.
sure or progressive disease therapeutic options include
Buske C et al. The Follicular Lymphoma International
salvage chemotherapy (e.g. etoposide, methylpredni-
Prognostic Index (FLIPI) separates high-risk from
sone, cytarabine, cisplatin [ESHAP]; ifosfamide, car- intermediate- or low-risk patients with advanced-stage
boplatin, etoposide [ICE], IVE ± rituximab) to follicular lymphoma treated front-line with rituximab
establish disease response and subsequent ASCT in and the combination of cyclophosphamide, doxorubicin,
those with responsive disease. In patients treated this vincristine, and prednisone (R-CHOP) with respect to
way, 5-year survival is approximately 30%. RIT and treatment outcome. Blood, 2006;108:1504–1508.
allo-SCT are more experimental strategies, but may Federico M et al. Follicular lymphoma international
also be effective in selected individuals. Patients pre- prognostic index 2: a new prognostic index for follicular
senting in transformation are managed differently to lymphoma developed by the international follicular
those who transform following previous therapies. lymphoma prognostic factor project. J Clin Oncol,
2009;27:4555–4562.
Patients whose first presentation is with high-grade
transformation should generally be managed as high- Federico M et al. R-CVP versus R-CHOP versus R-FM for
the initial treatment of patients with advanced-stage
grade lymphoma, and the management of high-grade
follicular lymphoma: results of the FOLL05 trial
lymphoma is described in more detail in Chapter 12. conducted by the Fondazione ltaliana Linfomi. J Clin
Oncol 2013;31:1506–1513.
Summary of relapsed follicular lymphoma Fisher RI et al. New treatment options have changed the
survival of patients with follicular lymphoma. J Clin
At present relapsed FL remains an incurable disease for Oncol, 2005;23:8447–8452.
the majority of patients. Despite the increasing number
Forstpointner R et al. The addition of rituximab to a
of treatment options, the optimal timing of the various
combination of fludarabine, cyclophosphamide,
therapies has yet to be defined. Intervention is generally mitoxantrone (FCM) significantly increases the response
recommended only in the context of symptomatic or rate and prolongs survival as compared with FCM alone
progressive disease. Rebiopsy and restaging of the lym- in patients with relapsed and refractory follicular and
phoma is recommended at each relapse. The choice of mantle cell lymphomas: results of a prospective
therapy will be principally dictated by the tempo of randomized study of the German Low-Grade Lymphoma
disease, response to previous therapy, and the patient’s Study Group. Blood, 2004;104:3064–3071.
performance status. Chemotherapy or antibody combi- Gyan E et al. High-dose therapy followed by autologous
nations that have provided a previous remission of 2 purged stem cell transplantation and doxorubicin-based
chemotherapy in patients with advanced follicular
years or more should be considered again. Rituximab-
lymphoma: a randomized multicenter study by the
naïve patients should receive rituximab-based immu- GOELAMS with final results after a median follow-up of
nochemotherapy. Rituximab should also be given as 9 years. Blood, 2009;113:995–1001. 101
maintenance in patients with chemosensitive disease.
Chapter 6: Follicular lymphoma
Haas RL et al. High response rates and lasting remissions Marcus R et al. Phase III study of R-CVP compared with
after low-dose involved field radiotherapy in indolent cyclophosphamide, vincristine, and prednisone alone in
lymphomas. J Clin Oncol, 2003;21:2474–2480. patients with previously untreated advanced follicular
lymphoma. J Clin Oncol, 2008;26:4579–4586.
Harris NL, et al. Follicular lymphoma. In Swerdlow SH,
Campo E, Harris NL, et al. (eds) WHO Classification of Mann RB, Berard CW. Criteria for the cytologic
Tumours of Haemopoietic and Lymphoid Tissues. Lyon: subclassification of follicular lymphomas: a proposed
IARC, 2008:220–226. alternative method. Hematol Oncol, 1983;1:187–192.
Herold M et al. Rituximab added to first-line mitoxantrone, Morris E et al. Outcomes after alemtuzumab-containing
chlorambucil, and prednisolone chemotherapy followed reduced-intensity allogeneic transplantation regimen for
by interferon maintenance prolongs survival in patients relapsed and refractory non-Hodgkin lymphoma. Blood,
with advanced follicular lymphoma: an East German 2004;104:3865–3871.
Study Group Hematology and Oncology Study. J Clin Morschhauser F et al. Phase III trial of consolidation therapy
Oncol, 2007;25:1986–1992. with yttrium-90-ibritumomab tiuxetan compared with
Hiddemann W et al. Frontline therapy with rituximab added no additional therapy after first remission in advanced
to the combination of cyclophosphamide, doxorubicin, follicular lymphoma. J Clin Oncol, 2008;26:5156–5164.
vincristine, and prednisone (CHOP) significantly Pettengell R et al. Rituximab purging and/or maintenance
improves the outcome for patients with advanced-stage in patients undergoing autologous transplantation for
follicular lymphoma compared with therapy with CHOP relapsed follicular lymphoma: a prospective randomized
alone: results of a prospective randomized study of the trial from the lymphoma working party of the European
German Low-Grade Lymphoma Study Group. Blood, group for blood and marrow transplantation. J Clin
2005;106:3725–3732. Oncol 2013;31:1624–1630.
Jegalian AG, et al. Follicular lymphoma in situ: clinical Pugh TJ, Ballonoff A, Newman F, Rabinovitch R. Improved
implications and comparisons with partial survival in patients with early stage low-grade follicular
involvement by follicular lymphoma. Blood, lymphoma treated with radiation: a Surveillance,
2011;118:2976–2984. Epidemiology, and End Results database analysis. Cancer,
Kaminski MS et al. 131I-tositumomab therapy as initial 2010;116:3843–3851.
treatment for follicular lymphoma. N Engl J Med, Rohatiner AZ et al. Myeloablative therapy with autologous
2005;352:441–449. bone marrow transplantation for follicular lymphoma at
Ladetto M et al. Prospective, multicenter randomized the time of second or subsequent remission: long-term
GITMO/IIL trial comparing intensive (R-HDS) versus follow-up. J Clin Oncol, 2007;25:2554–2559.
conventional (CHOP-R) chemoimmunotherapy in Rummel MJ et al. Bendamustine plus rituximab versus
high-risk follicular lymphoma at diagnosis: the CHOP plus rituximab as first-line treatment for
superior disease control of R-HDS does not translate into patients with indolent and mantle-cell lymphomas:
an overall survival advantage. Blood, an open-label, multicentre, randomised, phase 3
2008;111:4004–4013. non-inferiority trial. Lancet 2013;381:1203–1210.
Leahy MF, Turner JH. Radioimmunotherapy of relapsed Salles G et al. Rituximab combined with chemotherapy and
indolent non-Hodgkin lymphoma with 131I-rituximab interferon in follicular lymphoma patients: results of the
in routine clinical practice: 10-year single-institution GELA-GOELAMS FL2000 study. Blood,
experience of 142 consecutive patients. Blood, 2008;112:4824–4831.
2011;117:45–52.
Salles G et al. Rituximab maintenance for 2 years in patients
Lenz G et al. Myeloablative radiochemotherapy followed by with high tumour burden follicular lymphoma
autologous stem cell transplantation in first remission responding to rituximab plus chemotherapy (PRIMA):
prolongs progression-free survival in follicular a phase 3, randomised controlled trial. Lancet,
lymphoma: results of a prospective, randomized trial of 2011;377:42–51.
the German Low-Grade Lymphoma Study Group. Blood,
2004;104:2667–2674. Schouten HC et al. High-dose therapy improves
progression-free survival and survival in relapsed
Lowry L et al. Reduced dose radiotherapy for local control in follicular non-Hodgkin’s lymphoma: results from the
non-Hodgkin lymphoma: a randomised phase III trial. randomized European CUP trial. J Clin Oncol,
Radiother Oncol 2011;100:86–92. 2003;21:3918–3927.
Marcus R et al. CVP chemotherapy plus rituximab Schulz H et al. Chemotherapy plus rituximab versus
compared with CVP as first-line treatment for advanced chemotherapy alone for B-cell non-Hodgkin’s
follicular lymphoma. Blood, 2005;105:1417–1423.
102
Chapter 6: Follicular lymphoma
103
Chapter
7
MALT and other marginal zone lymphomas
Emanuele Zucca and Francesco Bertoni
Pathology: Andrew Wotherspoon Molecular cytogenetics: Andreas Rosenwald and German Ott
Extranodal marginal zone B-cell lymphomas, in which it has been reported in up to one-
quarter of the patients. It has been postulated that dis-
lymphoma (MALT lymphoma) semination of MALT lymphoma may be a result of
specific expression of special homing receptors or adhe-
Clinical features sion molecules on the surface of most B-cells of MALT,
Lymphoma of the mucosa-associated lymphoid tissue whether normal or transformed.
(MALT lymphoma) comprises about 7–8% of all non- MALT lymphoma usually has a favorable outcome,
Hodgkin lymphomas (NHLs). It is a neoplasm of adults with overall survival (OS) at 5 years higher than 85% in
with a median age at presentation of about 60 years and most series. Patients at high risk according to the
with a slightly higher proportion of females than males. International Prognostic Index (IPI) were found to
The presenting symptoms are essentially related to the carry a worse prognosis in retrospective series and in
primary location. Few patients present with elevated one prospective study. Patients with lymph node or
lactate dehydrogenase (LDH) or β2 microglobulin lev- bone marrow involvement at presentation, but not
els. Constitutional B-symptoms are extremely uncom- those with involvement of multiple mucosal sites, may
mon. MALT lymphoma usually remains localized for a have a worse prognosis. If initially localized, the disease
prolonged period within the tissue of origin, but dis- is generally slow to disseminate. Recurrences may
semination to multiple sites is not uncommon and has involve either extranodal or nodal sites. The median
been reported in up to one-quarter of cases, with either time to progression has been reported to be better for
synchronous or metachronous involvement of multiple the GI than for the non-GI lymphomas, but with no
mucosal sites or non-mucosal sites such as the bone significant differences in OS between the groups.
marrow, the spleen, or the liver. Regional lymph nodes Localization may have prognostic relevance because of
can also be involved. Bone marrow involvement is organ-specific clinical problems that determine partic-
reported in up to 20% of cases. ular management strategies, but also because of specific
The stomach is the commonest localization, repre- local pathogenic mechanisms, as suggested by the
senting about one-third of the cases. Other typical reports of different frequency of different chromosomal
presentation sites include the salivary glands, the translocations at distinct anatomic locations.
orbit, the thyroid, and the lung; the frequency of In a radiotherapy study from Toronto, gastric and
occurrence in different organs is shown in Table 7.1. thyroid MALT lymphomas had the best outcome,
Within the stomach, MALT lymphoma is often whereas distant failures were more common for
multifocal, possibly explaining the reports of relapses other sites. In a multicenter series from the
in the gastric stump after surgical excision. Gastric International Extranodal Lymphoma Study Group
MALT lymphoma can often disseminate to the small (IELSG), patients with the disease initially presenting
intestine and to the splenic marginal zone. Concomitant in the upper airway appeared to have a slightly poorer
gastrointestinal (GI) and non-GI involvement can be outcome (Table 9.1), but their small number pre-
detected in approximately 10% of cases. Disseminated vented any definitive conclusion. In general, despite
disease appears to be more common in non-GI MALT frequent relapses, MALT lymphomas most often
104 Lymphoma, Second Edition, ed. Robert Marcus, John W. Sweetenham, and Michael E. Williams. Published by
Cambridge University Press. © Cambridge University Press 2014.
Chapter 7: MALT and other MZLs
Table 7.1 Main clinical features and survival of MALT lymphomas at different presentation sites.
Figure 7.2 Extranodal marginal B-cell (MALT) lymphoma (stomach). Figure 7.3 Extranodal marginal B-cell (MALT) lymphoma (salivary
Some cases show prominent plasma cell differentiation. In this case gland). MALT lymphomas at all sites share a common morphological
the plasma cells are distended by accumulated immunoglobulin. appearance. The cells infiltrate associated epithelial structures to form
lymphoepithelial lesions.
neoplasm (Figure 7.2). Scattered large cells with abun- lymphoma (e.g. Hashimoto’s thyroiditis in the thyroid
dant cytoplasm, vesicular nuclei, and prominent and Sjögren’s syndrome in the salivary glands) is
nucleoli are present within the infiltrate. essential.
In cells associated with epithelial structures the
infiltrate lies between the follicle and the epithelium,
and neoplastic cells infiltrate and destroy the epithe-
The molecular biology of MALT
lial structures to form lymphoepithelial lesions lymphomas
(Figure 7.3). The disruption of the epithelial struc- The molecular genetic and cytogenetic features of
tures is associated with morphological changes in the extranodal lymphomas are in keeping with their
epithelial cells that impart an eosinophilic appear- marked heterogeneity with respect to morphology,
ance to their cytoplasm. immunophenotype, clinical presentation, and clinical
Involved lymph nodes show either a perifollicular outcome.
infiltrate or, in more advanced cases, a nodular pro- Extranodal marginal zone B-cell lymphomas of
liferation of similar cells. These cases may mimic nodal MALT-type represent mature B-cell tumors with pro-
marginal zone lymphoma (MZL). ductively rearranged immunoglobulin heavy-chain
Immunophenotypically MALT lymphomas express (IgH) genes. Interestingly, the VH4, VH3, and VH1
surface immunoglobulin (usually IgM, less often IgA or family genes known to be involved in autoantibody
IgG). The lymphocytes express CD19, CD20, CD22, formation are preferentially used, suggesting a possi-
CD79a, and PAX5. Approximately half of cases express ble derivation from autoreactive B-cell clones. The
CD43. They are usually negative for CD5, CD23, CD10, rearranged IgVH genes are usually mutated and often
and cyclin D1. There is expression of BCL-2 protein. A show a mutation pattern indicative of selective antigen
small percentage of cases show expression of CD5 and pressure. Recent studies provide evidence that MALT-
these may be associated with a more aggressive presen- type lymphomas from special localizations (gastric
tation including a leukemic phase. and salivary gland), in contrast to any other lym-
MALT lymphomas need to be distinguished from phoma subtype, frequently express B-cell receptors
other low-grade lymphomas, including follicular lym- with strong homology to the rheumatoid factor (RF).
phoma (FL), mantle cell lymphoma, and small lym- This finding underscores the strong link between
phocytic lymphoma/B-cell chronic lymphocytic chronic antigenic stimulation, autoimmunity, and
leukemia (B-CLL). This can usually be done by mor- development of MALT-type lymphomas.
phology and immunophenotypic studies. Distinction Recurrent genetic aberrations in MALT-type lym-
106 from underlying reactive conditions that are known to phomas include trisomies of chromosomes 3 and 18,
be precursor lesions to the development of MALT DNA gains at 6p and 8q, and losses at 6q23 (TNFAIP3),
Chapter 7: MALT and other MZLs
Malt lymphoma staging disease seem to have the same long-term outcome as
those with localized disease.
Since patients presenting with lymphoma dissemi-
Outside clinical trials, aggressive staging proce-
nated to multiple mucosal sites may have a favorable
dures should be tailored to the individual patient
outcome similar to the patients with localized disease,
according to the clinical conditions (presenting local-
the traditional Ann Arbor staging system, which is
ization, age, intended treatment, performance status,
mainly based on the extension of nodal areas, is not
and symptoms). Staging procedures should always
optimal. Alternative staging systems (Table 7.2) for
comprise a complete clinical history and physical
extranodal lymphomas were proposed in the 1970s,
examination with a careful evaluation of all lymph
but a general consensus has not yet been achieved.
node regions, inspection of the upper airway and ton-
The main problem in staging patients with MALT
sils, thyroid examination, and clinical evaluation of the
lymphoma is the number of extranodal sites to be
size of liver and spleen. Standard chest radiographs
explored at diagnosis. The finding of asymptomatic
and a CT scan of thorax, abdomen, and pelvis should
dissemination in patients with apparently localized
be performed. Bone marrow biopsy must be per-
disease, and the relatively high proportion of patients
formed at diagnosis. Laboratory tests should include
with early dissemination, suggests the need for exten-
complete blood counts with cytological examination,
sive staging procedures in all patients with MALT
LDH, and β2 microglobulin levels, evaluation of renal
lymphoma. However, patients with disseminated
Ann Arbor Musshoff Lugano (Rohatiner, TNM modifications in the Lymphoma extension
(Carbone, (Musshoff, 1994) Paris system
1971) 1977) (Ruskone-Fourmestraux,
2003)
IE IE1 I = confined to GI tract T1m N0 M0 Mucosa
(single primary or
multiple, non-
contiguous)
IE IE2 T1sm N0 M0 Submucosa
IE T2 N0 M0 Muscularis propria
T3 N0 M0 Serosa
IIE IIE1 II = extending into Tl–3 Nl M0 Perigastric lymph nodes
abdomen
IIE IIE2 II1 = local nodal Tl–3 N2 M0 More distant regional nodes
involvement
II2 = distant nodal Tl–3 N3 M0 Extra-abdominal lymph nodes
involvement
IIE IIE IIE = penetration of T4 N0 M0 Invasion of adjacent structures
serosa to involve
adjacent organs or
tissues
IIIE IIIE IV = disseminated T1–4 N3 M0 Lymph nodes on both sides of the
extranodal diaphragm, and/or additional
involvement or extranodal sites with non-
concomitant continuous involvement of
supradiaphragmatic separate GI site
nodal involvement
IVE IVE T1–4 N0–3 M1 Or non-continuous involvement
T1–4 N0–3 M2 of non-GI sites
Tl–4 N0–3 M0–2 BX Bone marrow not assessed
Tl–4 N0–3 M0–2 B0 Bone marrow not involved
108 Tl–4 N0–3 M2 B1 Bone marrow involvement
Chapter 7: MALT and other MZLs
Table 7.3 Site-specific investigations that may be useful for the staging of extranodal
lymphomas of MALT type.
Stomach Gastroduodenal endoscopy with multiple gastric biopsies from all the visible
lesions and the non-involved areas with a complete mapping
Histological and histochemical examination for H. pylori (Genta stain or
Warthin–Starry stain of antral biopsy specimen) and serology studies if
histology results are negative
FISH for the t(ll;18) translocation
Endoscopic ultrasound
Small intestine Endoscopy
(IPSID) Small bowel double-contrast X-ray
Campylobacter jejuni search in the tumor biopsy by PCR,
immunohistochemistry or in-situ hybridization
Large intestine Colonoscopy
Endoscopic ultrasound
Ocular adnexa MRI (or CT scan) and ophthalmologic examination
Chlamydophila psittaci in the tumor biopsy and blood
Mononuclear cells by PCR
Lung Bronchoscopy with bronchoalveolar lavage
Salivary gland, ENT examination and ultrasound
tonsils, parotid
Thyroid Ultrasound ±CT scan of the neck and thyroid function tests
Breast Mammography and MRI
Skin Borrelia burgdorferi in the tumor biopsy by PCR
CT, Computed tomography; ENT, ear, nose, and throat; FISH, fluorescent in situ hybridization; IPSID,
immunoproliferative small intestinal disease; MRI, magnetic resonance imaging; PCR, polymerase chain
reaction.
and liver function, and hepatitis C virus (HCV) and was developed for Hodgkin’s disease and is not
HIV serology. Utility of positron emission tomogra- adequate for the specific problems posed by GI tract
phy (PET) scanning remains unclear, with conflicting lymphomas. A variety of alternative systems have
reports on 18FDG avidity of extranodal MZLs, and, therefore been proposed, as summarized in
thus, the exam is not currently recommended. Then, Table 7.2. We have largely used the modification of
depending upon the particular clinical presentation, the Blackledge system known as the “Lugano” staging
staging investigations should focus on the specific system. However, this system was proposed before
organs suspected of being involved (Table 7.3). the wide use of endoscopic sonography and does not
accurately describe the depth of infiltration in the
Diagnosis and staging of gastric MALT lymphoma gastric wall, a parameter that is highly predictive for
The stomach is the most common and best-known the MALT lymphoma response to anti-Helicobacter
MALT lymphoma site. The most frequent presenting therapy.
symptoms of gastric MALT are non-specific upper There is a general consensus that initial staging
GI complaints (dyspepsia, epigastric pain, nausea, procedures should include a gastroduodenal endos-
and chronic manifestations of GI bleeding such as copy, with multiple biopsies from each region of the
anemia), often leading to an endoscopy that usually stomach, duodenum, and gastroesophageal junction,
reveals non-specific gastritis or peptic ulcer, with as well as from any abnormal-appearing site. Fresh
mass lesions being unusual. Diagnosis is based on biopsy and washings material should be available for
the histopathologic evaluation of the gastric cytogenetic studies in addition to routine histology
biopsies. and immunohistochemistry. Fluorescent in situ
The best staging system is still controversial. The hybridization (FISH) analysis or a molecular assay
Ann Arbor staging system, routinely used for NHL, for the detection of t(11;18) is recommended for 109
Chapter 7: MALT and other MZLs
Table 7.4 Complete remission and relapse rates after antibiotic therapy for localized (stage I–II) gastric MALT lymphomas in recent series
with at least 3 years of median follow-up.
Several effective anti-H. pylori programs are avail- Anti-Helicobacter therapy in gastric diffuse
able. The choice should be based on the epidemiology large B-cell lymphoma
of the infection in the different countries, taking into Anti-Helicobacter therapy may also be of benefit in
account the locally expected antibiotic resistance. The some cases of DLBCL of the stomach since, in the
most commonly used regimen is triple therapy with a subset of cases that may have been derived from a
proton pump inhibitor (e.g. omeprazole, lansoprazole, MALT lymphoma, antibiotics may eliminate a resid-
pantoprazole, or esomeprazole) in association with ual or relapsed low-grade component that can be
amoxicillin and clarithromycin. Metronidazole can responsible for tumor recurrence following antigen
substitute for amoxicillin in penicillin-allergic indivi- stimulation. Cases of regression of high-grade lesions
duals. Other regimens that include bismuth or H2- after anti-H. pylori therapy have been reported,
receptor antagonists with antibiotics are also effective. suggesting that high-grade transformation is not ne-
The role of additional chemotherapy after H. pylori cessarily associated with the loss of H. pylori depend-
eradication was investigated in a randomized study ence. However, relying solely on antibiotic therapy
(whose power, however, has been limited from not for gastric large-cell lymphomas cannot be advised
having reached the planned accrual). In this study, outside clinical trials until large-scale prospective
chlorambucil conferred no benefit, with progression- studies have validated its use as first line therapy,
free survival (PFS) and OS rates similar for observed- and at present we recommend treating them as local- 111
only and chlorambucil-treated patients. ized DLBCL.
Chapter 7: MALT and other MZLs
H. pylori eradication if
infection present
Table 7.5 The Wotherspoon and GELA systems for the histological evaluation of gastric MALT lymphoma endoscopic biopsies.
eradication antibiotic therapy, although not necessar- and does not imply a frank clinical progression. The
ily to chemotherapy or immunotherapy. Moreover, clinical relevance of the detection of monoclonal B-cells
the cases carrying this translocation rarely undergo by molecular methods remains unclear. In the long-
high-grade transformation and have been reported to term follow-up of some cases with minimal residual
have a significantly longer relapse-free survival irre- disease neither lymphoma clinical growth nor histolog-
spective of treatment modality. There are no large- ical transformation was documented despite a persistent
scale studies to suggest the utility of any specific treat- clonal population, suggesting that a watch-and-wait
ment strategy in this subgroup. H. pylori eradication policy could be feasible and safe and that these patients
may have some benefit on symptoms, but is usually do not necessarily require additional treatment. On the
unable to induce a lymphoma regression. The treat- other hand, cases of lymphoma recurrence following
ment of antibiotic-resistant gastric MALT lymphoma, H. pylori reinfection have been reported, suggesting
as discussed later, is a controversial issue, with no that residual dormant tumor cells can be present despite
evidence to suggest that one therapeutic approach is histological remission.
better than another. More recent long-term follow-up studies suggest
that even minimal residual histological disease detected
Clinical and molecular follow-up after macroscopic regression may remain clinically dor-
A number of molecular follow-up studies have shown mant, and that a “wait-and-see” policy seems safe and
that post-antibiotic histological and endoscopic remis- could be taken, at least for patients agreeable to frequent
sion does not necessarily mean a cure. The long-term endoscopies. Relapses have also been documented in
persistence of monoclonal B-cells after histologic the absence of H. pylori reinfection, indicating the pres-
regression of the lymphoma has been reported in ence of lymphoma clones that may have escaped the
about half of the cases, suggesting that H. pylori erad- antigenic drive. The frequency of histological transfor-
ication suppresses but does not eradicate the lymphoma mation into DLBCL is unclear, but its long-term risk
clones. Transient histological and molecular relapses seems lower than in other indolent lymphomas at least
have been reported during long-term follow-up of for the primary gastric presentations. Several cases of
antibiotic-treated patients, but without the stimulus synchronous or metachronous gastric adenocarcino- 113
from H. pylori this usually remains a self-limited event mas in patients with gastric MALT lymphomas have
Chapter 7: MALT and other MZLs
been documented. Indeed, a Dutch nationwide tumor fludarabine and cladribine, which might, however, be
registry study has shown a sixfold increased risk of associated with an increased risk of secondary myelo-
gastric adenocarcinoma in patients with gastric MALT dysplastic syndrome (MDS), and of a combination
lymphomas in comparison with the general population. regimen of chlorambucil/mitoxantrone/prednisone.
A strict follow-up is therefore strongly advisable, and Aggressive anthracycline-containing chemotherapy
histological evaluation of repeated biopsies continues to should be reserved for patients with high tumor bur-
be the fundamental follow-up procedure. We perform a den (bulky masses, high IPI score).
breath test 2 months after treatment to document H. The activity of the anti-CD20 monoclonal anti-
pylori eradication and repeat post-treatment endoscop- body rituximab has been demonstrated in a phase II
ies with multiple biopsies every 6 months for 2 years, study, with a response rate of about 70%, and this may
then yearly to monitor the histological regression of the represent an additional option for the treatment of
lymphoma. systemic disease. The efficacy of the combination of
rituximab with chlorambucil has been explored in a
Management of H. pylori-negative or randomized study of the International Extranodal
Lymphoma Study Group (IELSG) in gastric (failing
antibiotic-resistant cases
antibiotics) or non-gastric MALT lymphomas
No definitive guidelines exist for the management of (IELSG19, NCT00210353). In comparison with chlor-
the subset of H. pylori-negative cases and for the ambucil alone, chlorambucil plus rituximab results in
patients who fail antibiotic therapy. In two retrospec- increased complete remission and event-free survival
tive series of patients with gastric low-grade MALT (EFS) rates but 5-year OS was identical in both groups.
lymphoma, no statistically significant difference was Similar approaches can be considered for gastric
apparent in survival between patients who received lymphoma patients experiencing a frank relapse after
different initial treatments (including chemotherapy an initial lymphoma response to antibiotics. Local
alone, surgery alone, surgery with additional chemo- relapses may benefit from IF-RT, while systemic relap-
therapy or radiation therapy, or antibiotics against H. ses may require systemic treatment.
pylori). Surgery has been widely and successfully used
in the past. It leads to excellent long-term local control,
but its role should be redefined in view of the promis- Management of non-gastric localizations
ing results of the conservative approach. Whether MALT lymphomas have been described in various non-
stage I patients treated with radical gastrectomy need GI sites, such as salivary glands, thyroid, skin, conjunc-
further treatment is unclear, but a wait-and-see policy tiva, orbit, larynx, lung, breast, kidney, liver, prostate,
is most often appropriate and for the majority of and even the intracranial dura. In general, MALT lym-
patients other treatment approaches are preferred. A phoma arises in mucosal sites where lymphocytes are
modest dose of involved-field radiotherapy (IF-RT; not normally present and where the MALT is acquired
30 Gy given in 4 weeks to the stomach and perigastric in response either to autoimmune processes such as
nodes) gives excellent disease control and it might be Hashimoto’s thyroiditis and Sjögren’s syndrome or to
the treatment of choice for patients with stage I–II chronic infectious conditions. H. pylori gastritis is the
MALT lymphoma of the stomach without evidence best studied condition, but other infectious agents have
of H. pylori infection, or with persistent lymphoma more recently been implicated in the pathogenesis of
after antibiotics. MALT lymphomas arising in the skin (Borrelia burg-
Patients with disseminated disease should be con- dorferi), in the ocular adnexa (Chlamydophila psittaci)
sidered for systemic treatment (i.e. chemotherapy and/ and in the small intestine (Campylobacter jejuni).
or immunotherapy with an anti-CD20 monoclonal Among viruses, the involvement of HCV infection in
antibody). In the presence of advanced-stage disease, the development of some MZLs (especially the splenic
chemotherapy is an obvious choice, but only a few type) has recently been proposed.
compounds and regimens have been tested specifically Non-gastric MALT lymphomas have been diffi-
in MALT lymphomas. Oral alkylating agents (either cult to characterize because these tumors are distrib-
cyclophosphamide or chlorambucil, with median uted so widely throughout the body that it is difficult
treatment duration of 1 year), can result in a high to assemble adequate series for any given site. One-
rate of disease control. Phase II studies have demon- quarter of non-GI MALT lymphomas have been
114
strated some antitumor activity of the purine analogs reported to present with involvement of multiple
Chapter 7: MALT and other MZLs
mucosal sites or non-mucosal sites such as bone lymphomas. Similarly, the finding that C. psittaci is
marrow. Non-GI MALT lymphomas, despite pre- associated with MALT lymphoma of the ocular adnexa
senting with stage IV disease more often than the may provide the rationale for the antibiotic treatment
gastric variant, frequently have a quite indolent with doxycycline of localized lesions, and preliminary
course regardless of treatment type, but they are encouraging results have been reported, but this
significantly more prone to relapse than the primary approach remains investigational and will need to be
gastric cases (most often at other mucosal sites). A confirmed by larger clinical studies.
multicentric retrospective survey of 180 non-gastric
cases showed no evidence of a clear advantage for any
type of therapy and, despite the high proportion of
Immunoproliferative small
cases with disseminated disease, which should intestinal disease (IPSID)
require a systemic approach, no clear advantage was This condition, known in the past as alpha heavy chain
associated with any chemotherapy program. disease or Mediterranean lymphoma, is nowadays
In general, the treatment of non-gastric MALT considered a special subtype of MALT lymphoma.
lymphoma can be approached in the same way as The distinguishing feature of IPSID is the synthesis
that of the H. pylori-negative cases described above. of alpha heavy chain that is secreted and detectable in
The sole use of moderate-dose radiotherapy for the serum, urine, saliva, and duodenal fluid in approx-
patients with localized disease results in a high rate of imately two-thirds of cases; in the remainder, the
local control and often cure. Effective radiotherapy is heavy chain protein is demonstrable by tumor cell
frequently possible for both common and rare presen- immunohistochemistry but not secreted. Most of the
tations of the disease. Side effects of RT are mild and cases have been described in the Middle East, espe-
reversible. For patients with localized non-gastric cially in the Mediterranean area, where the disease is
MALT lymphoma, given the unique biologic behavior endemic and affects young adults of both sexes, but
with a tendency to relapse in MALT tissues and an predominantly males. A few cases have been reported
indolent course, RT is often the first line treatment of from industrialized western countries, usually among
choice. The emerging literature on localized MALT immigrants from the endemic area.
lymphomas confirms a high rate of local control in The pathologic features of IPSID recognize a specific
MALT lymphoma, with a high proportion of patients type of MALT lymphoma that is characterized by thick-
likely to be cured of the disease. The moderate doses of ening of the wall of the proximal small bowel. The
radiation required for cure (25–35 Gy) are generally morphological appearances are similar to typical
associated with a minimal risk of long-term toxicity, MALT lymphoma but there is profound plasma cell
although special considerations are needed for partic- differentiation. There are scant lymphoid cells that
ular localizations such as the eye or the lung. may only become obvious following staining for
For some patients with stage III or IV disease, CD20. These are predominantly around the crypts, and
radiotherapy can also be an effective therapy in pro- lymphoepithelial lesions are seen. Immunophenotypical
viding local disease control, but the optimal manage- characteristics are the same as for other types of MALT
ment of disseminated MALT lymphomas is less clearly lymphoma but these cells show synthesis of IgA without
defined. Because no curative treatment exists, expect- either light chain.
ant observation can be an adequate initial policy in The natural course of IPSID is usually prolonged,
most patients. In general, the treatment should be often over many years, including a potentially rever-
“patient-tailored,” taking into account the site, the sible early phase, with the disease usually confined to
stage, and the clinical characteristics of the individual the abdomen. If untreated it degenerates, with high-
patient. When systemic treatment is needed, enroll- grade transformation, into large B-cell lymphoma.
ment in controlled clinical trials is advisable. As men- Since the histology of mucosal lesions and mesen-
tioned above, only one randomized study has thus far teric lymph nodes can be discordant, with a higher
been conducted (IELSG19) that showed a clinical ben- grade in the latter, staging laparotomy can be useful
efit from combining chlorambucil and rituximab, but in the evaluation of the mesenteric nodal involve-
with no impact on OS. ment, but surgery has no therapeutic role because
Systemic antibiotic treatment should be tried first of a generally diffuse intestinal involvement. It
in patients with B. burgdorferi-associated cutaneous has been known since the 1970s that early IPSID 115
Chapter 7: MALT and other MZLs
phases could achieve durable remissions with sus- Pathology of splenic MZL
tained antibiotic treatment (such as tetracycline or
The splenic pathology shows that the disease is pre-
metronidazole and ampicillin for at least 6 months).
dominantly within the white pulp, with a less pro-
Recently, the presence of C. jejuni has been demon-
nounced nodular and interstitial infiltration of
strated in IPSID tumor sections. At an early stage,
the red pulp. The white pulp is partially or com-
antibiotic treatment directed against C. jejuni may
pletely effaced by the neoplastic cells. These areas
lead to lymphoma regression, but there is no clear
show an inner zone of small lymphocytes that
evidence that antibiotics alone are of benefit in
resemble mantle cells that either surround residual
the advanced phases. Although the early studies
germinal centers or, more commonly, replace them.
reported that aggressive chemotherapy is not well
Around this small cell layer there is a further zone
tolerated by patients with advanced disease and
composed of slightly larger cells that have more
severe malabsorption, anthracycline-containing reg-
abundant cytoplasm. Large transformed cells with
imens, combined with nutritional support plus anti-
more abundant cytoplasm and vesicular nuclei that
biotics to control diarrhea and malabsorption, may
contain nucleoli are scattered within the outer layer.
offer the best chance of cure, with 5-year survival
Plasma cell differentiation may occur. The red pulp
rates up to 70%.
nodules are frequently composed of cells resem-
bling those of the inner zone of the white pulp
(Figure 7.5).
Primary splenic marginal zone Spread to the lymph nodes of the spleen hilum is
lymphoma frequent. These nodes may show a nodular pattern
based on pre-existing follicles or may show a zoned
Clinical features appearance similar to that seen in the splenic white
Splenic MZL is a very rare disorder, comprising less pulp. The bone marrow shows interstitial infiltration
than 1% of all lymphomas. Up to two-thirds of by small lymphocytes usually with a nodular compo-
patients with splenic MZL have circulating villous nent. In the interstitium cells can frequently be seen
lymphocytes with characteristic fine short cytoplasm within sinusoids, although this pattern is not specific
polar projections. When these are more than 20% of to splenic MZL.
the lymphocyte count, the term “splenic lymphoma Immunophenotypically the cells show surface
with villous lymphocytes” is commonly used. Despite immunoglobulin, usually IgM and IgD. They express
relevant geographical variations, HCV seems to be
involved in lymphomagenesis of a portion of cases.
An association with malaria and with Epstein–Barr
virus (EBV) infection has been shown in tropical
Africa, and the tropical cases are characterized by a
high percentage of circulating villous lymphocytes.
Most patients with primary splenic MZL are over 50,
with a similar incidence in males and females. The
disease usually presents with massive splenomegaly,
which produces abdominal discomfort and pain.
Lymphadenopathy is usually minimal or absent. The
diagnosis may be made at splenectomy, performed to
establish the cause of unexplained spleen enlargement.
B-symptoms are not uncommon: anemia, thrombocy-
topenia, or leukocytosis are reported in one-quarter of
cases. Autoimmune hemolytic anemia can be found in
Figure 7.5 Splenic marginal zone B-cell lymphoma (spleen). The
up to 15% of patients. Involvement of the splenic hilar white pulp shows infiltration by a population of lymphoid cells that
lymph nodes and/or the liver is reported in 25–30% of surround residual reactive germinal centers with a biphasic pattern.
cases. Frequently, a small clonal B-cell population may There is an inner rim of cells with scanty cytoplasm and an outer rim
of cells with larger nuclei and more abundant cytoplasm. Scattered
116 be detected by peripheral blood flow cytometry even in larger cells are seen in the outer layer. The red pulp contains small
the absence of overt lymphocytosis. nodular aggregates of lymphoma cells.
Chapter 7: MALT and other MZLs
CD20 and CD79a and are usually negative for CD5, Clinical course and management
CD23, CD10, and cyclin D1. There is expression of
The clinical diagnosis is based on a combination of
BCL-2 protein. Proliferation shows a characteristic
features, including lymphocyte morphology, immu-
pattern with infrequent cells staining for KI67/MiB1
nophenotype, cytogenetic abnormalities, bone mar-
in the inner, small cell zone and scattered cells positive
in the outer zone. row histology, and, when available, spleen histology.
Splenic MZLs have to be distinguished from other A set of minimum diagnostic criteria for splenic MZL
small B-cell lymphomas, almost all of which can adopt have been proposed and should be followed. Diagnosis
an organizational pattern that mimics that of splenic can be based either on spleen histology with immuno-
MZL. This can usually be accomplished by the immu- phenotypic features not diagnostic for chronic lym-
nocytochemical studies noted here. phocytic leukemia or on the presence of typical blood
and bone marrow morphology, immunophenotype
not consistent with the diagnosis of chronic lympho-
Molecular pathogenesis of splenic MZL cytic leukemia and intrasinusoidal infiltration by
As this is a relatively newly recognized lymphoma CD20-positive cells (if spleen histology is unavailable).
entity, its precise molecular pathogenesis is largely Nowadays, an increasing number of cases is diagnosed
unknown. Recent studies indicate that these tumors based on the examination of bone marrow and periph-
may harbor unmutated or mutated IgVH genes, thus eral blood specimens only.
being derived from naïve or antigen-experienced According to the WHO classification, peripheral
B-cells. The pattern of IgVH gene usage (biased lymph node involvement is typically absent. However,
usage of VH1–2 gene family) and the finding of patients with disseminated MZLs can be observed and a
somatic hypermutations in some cases suggest an precise diagnosis can be very difficult in cases presenting
antigen-driven lymphomagenesis. In line with these with concomitant splenic, extranodal, and nodal
findings, a possible relationship between HCV infec- involvement. In a retrospective French series of 124
tion and splenic MZL has recently been established. patients with non-MALT-type MZL, four clinical sub-
Cytogenetic and molecular genetic studies have types were observed: splenic (48% of cases), nodal (30%),
demonstrated that approximately 45% of splenic disseminated (splenic and nodal, 16%), and leukemic
MZL harbor allelic losses in the 7q32–q33 chromo- (neither splenic nor nodal, 6%). Even when the disease
some region and these cases are possibly more is restricted to the cases presenting with splenomegaly,
aggressive clinically. Other recurrent cytogenetic nearly all patients have bone marrow involvement, often
alterations include trisomies 3 and 18, +12q and accompanied by involvement of peripheral blood
translocations t(6;14)(p12;q32), and t(2;7)(p12;q22) (defined as the presence of absolute lymphocytosis of
with deregulation of Cyclin D3 and CDK6. more than 5%). Serum paraproteinemia is observed in
Gene expression profiling studies confirmed the about 10–25% of cases and is most frequently of IgM
homogeneity of this disease entity on the molecular type, posing the problem of differential diagnosis with
level and point to different pathways that may be lymphoplasmacytic lymphoma/Waldenström’s macro-
involved in lymphomagenesis. The molecular signa- globulinemia. The two diseases often present with sim-
ture of splenic MZL includes upregulation of several ilar clinical features (splenomegaly, bone marrow
genes involved in NFκB activation, B-cell receptor lymphoplasmacytic infiltration, anemia), but marked
and tumor necrosis factor (TNF) signaling, and of hyperviscosity and hypergammaglobulinemia are
genes associated with the splenic microenvironment uncommon in splenic MZL.
(SELL, LPXN). Genes located in the 7q31–7q32 dele- The clinical course is most usually indolent, with
tion region have been reported to be downregulated. 5-year OS ranging from 65% to 80%. Histological trans-
Two large studies – one cytogenetic study and a formation is rare, often associated with B-symptoms,
comprehensive genome-wide analysis of DNA copy disease dissemination, and poorer outcome. A prognos-
number alterations – confirmed that DNA copy tic model for clinical use was recently proposed, based
number imbalances in splenic MZLs are indeed spe- on the results of a large cooperative retrospective study
cific for the entity and differ from other lymphomas, of 309 patients from Italy. Using three readily available
including MALT-type MZL. More recent studies parameters, hemoglobin less than 120 g/L, LDH higher
have identified recurrent somatic mutations affect- than normal, and serum albumin less than 35 g/L, three 117
ing the NFκB and NOTCH pathways. prognostic groups can be identified: low-risk patients
Chapter 7: MALT and other MZLs
(5-year cause-specific survival 88%) with no adverse therapy should be considered in the positive cases
factors, intermediate-risk patients (5-year cause-specific before any decision is reached about more aggressive
survival 73%) with one adverse factor, and high-risk therapeutic approaches.
patients (5-year cause-specific survival 50%) with two
or three adverse factors.
The reported largest series shows that most patients
Nodal marginal zone lymphoma
can initially be managed with a wait-and-see policy, and
they do not seem to have a worse outcome. When
Clinical features
treatment is needed this is usually because of large In contrast with mucosa-based extranodal MALT lym-
symptomatic splenomegaly or cytopenia. Splenectomy phoma, nodal MZL (previously known as monocytoid
has been considered the treatment of choice: it allows a B-cell lymphoma) is typically lymph node-based. This
reduction/disappearance of circulating tumor lympho- type of lymphoma is exceedingly rare, accounting for
cytes and recovery of the lymphoma-associated cytope- less than 10% of all MZLs and less than 1% of all NHLs.
nia. Responses to splenectomy occur in approximately It has also been associated with HCV infection in some
90% of patients and the benefit often persists for several epidemiologic studies. The clinical data are sparse and
years; time to next treatment can be longer than 5 years. have been largely drawn from pathologic series rather
Adjuvant chemotherapy after splenectomy may result than clinical ones. Nodal MZL is a disease of older
in a higher rate of complete response, but there is no people, with the median age at presentation in the
evidence of a survival benefit. sixth decade, and affects both sexes, with a slight
Chemotherapy alone may be considered for female predominance. Whether this lymphoma car-
patients who require treatment but have contraindica- ries a worse prognosis in comparison with other MZLs
tions to splenectomy, and for the patient with clinical remains controversial.
progression after spleen removal. Alkylating agents The commonest presenting feature is localized –
(chlorambucil or cyclophosphamide) have been most often in the neck – or sometimes a generalized
reported to be active and can be used as single agents adenopathy. Bone marrow is involved at presentation
or in combination (as in the CVP and CHOP regi- in less than half of the cases. Transformation to high-
mens). Among the purine analogs, fludarabine has grade lymphoma has been described in some cases.
been shown to be effective but, curiously, cladribine
seems not to be active. Rituximab, alone or in combi- Pathology and genetics
nation with chemotherapy, induces good responses in The histologic pattern of lymph node involvement by
cases refractory to standard chemotherapy. In retro- primary nodal MZL is often indistinguishable from
spective series of rituximab monotherapy, sustained that of extranodal MZL of MALT type, and this lym-
responses are reported in up to 90% of patients, with phoma has previously been thought to be a nodal
rapid regression of splenomegaly and improvement of variety of MALT lymphoma. The tumor cell morphol-
cytopenias. OS seems comparable to that reported ogy is heterogeneous and is similar to the lymph node
following splenectomy. Since splenic MZL patients involvement of extranodal MZL and splenic MZL.
are often elderly and at high risk surgically, frontline Sometimes the tumor cells resemble monocytes;
treatment with rituximab alone or in combination indeed, the term monocytoid lymphoma was used in
with chemotherapy has become more widely used. the past to indicate this type of lymphoma.
Very interesting is the observation from one study Analysis of the IgH genes suggests a prevalence of
that some patients with splenic lymphoma with villous cases with mutated IgH genes, but, similarly to splenic
lymphocytes and HCV infection obtained a complete MZL, unmutated cases do exist. No specific genomic
response after treatment with interferon alpha, alone aberration is known to occur in nodal MZL. The most
or in combination with ribavirine. Interferon treat- common alterations, such as gain of 3q, are also
ment had no antitumor effect on HCV-negative present in extranodal MZL and splenic MZL.
splenic lymphomas. Analogous to H. pylori infection
in gastric MZL, it appears that HCV may be respon-
sible for an antigen-driven stimulation of the lym- Management
phoma clone. This report suggests that all patients Because no curative treatment exists, most patients
118 should be tested for HCV infection, and antiviral can initially be managed with expectant observation.
Chapter 7: MALT and other MZLs
There is at present no consensus about the best treat- Farinha P, Gascoyne RD. Molecular pathogenesis of
ment, individual cases being managed differently mucosa-associated lymphoid tissue lymphoma. J Clin
according to site and stage. Treatment options may Oncol, 2005;23:6370–6378
include single-agent chlorambucil or fludarabine or Ferreri AJ, Dolcetti R, Du MQ, et al. Ocular adnexal MALT
combination chemotherapy regimens (such as CVP lymphoma: an intriguing model for antigen-driven
or CHOP). Rituximab has single-agent efficacy and lymphomagenesis and microbial-targeted therapy. Ann
Oncol, 2008;19:835–846.
can be combined with chemotherapy. Anti-HCV
treatment may induce lymphoma regression in Hancock BW, Qian W, Linch D, et al. Chlorambucil versus
some HCV-infected patients. Autologous stem cell observation after anti-Helicobacter therapy in gastric
MALT lymphomas: results of the international
transplantation has been used in younger patients
randomised LY03 trial. Br J Haematol, 2009;144:367–375.
with adverse prognostic factors or with large-cell
transformation. Iannitto E, Tripodo C. How I diagnose and treat splenic
lymphomas. Blood, 2011;117:2585–2595.
Kiel MJ, Velusamy T, Betz BL, et al. Whole-genome
Further reading sequencing identifies recurrent somatic NOTCH2
mutations in splenic marginal zone lymphoma. J Exp
Arcaini L, Lazzarino M, Colombo N, et al. Splenic marginal
Med 2012;209:1553–65.
zone lymphoma: a prognostic model for clinical use.
Blood, 2006;107:4643–4649. Lecuit M, Abachin E, Martin A. et al. Immunoproliferative
small intestinal disease associated with Campylobacter
Berger F, Felman P, Thieblemont C, et al. Non-MALT
jejuni. N Engl J Med 2004;350:239–248.
marginal zone B-cell lymphomas: a description of clinical
presentation and outcome in 124 patients. Blood, Martinelli G, Laszlo D, Ferreri AJ, et al. Clinical activity of
2000;95:1950–1956. rituximab in gastric marginal zone non-Hodgkin’s
lymphoma resistant to or not eligible for anti-
Bennett M, Schechter GP. Treatment of splenic marginal
Helicobacter pylori therapy. J Clin Oncol
zone lymphoma: splenectomy versus rituximab. Semin
2005;23:1979–1983.
Hematol, 2010;47:143–147.
Matutes E, Oscier D, Montalban C, et al. Splenic marginal
Bertoni F, Zucca E. State-of-the-art therapeutics: marginal
zone lymphoma proposals for a revision of diagnostic,
zone lymphoma. J Clin Oncol, 2005;23:6415–6420.
staging and therapeutic criteria. Leukemia,
Bertoni F, Conconi A, Capella C, et al. Molecular follow-up 2008;22:487–495.
in gastric mucosa-associated lymphoid tissue
Mazloom A, Medeiros LJ, Mclaughlin PW, et al. Marginal
lymphomas: early analysis of the LY03 cooperative trial.
zone lymphomas: factors that affect the final outcome.
Blood, 2002;99:2541–2544.
Cancer, 2010;116:4291–4298.
Chen LT, Lin JT, Tai JJ, et al. Long-term results of anti-
Mollejo M, Camacho FI, Algara P, et al. Nodal and splenic
Helicobacter pylori therapy in early-stage gastric high-
marginal zone B cell lymphomas. Hematol Oncol,
grade transformed MALT lymphoma. J Natl Cancer Inst,
2005;23:108–118.
2005;97:1345–1353.
Montalban C, Abraira V, Arcaini L, et al. Simplification of risk
Copie-Bergman C, Gaulard P, Lavergne-Slove A, et al.
stratification for splenic marginal zone lymphoma: a point-
Proposal for a new histological grading system for post-
based score for practical use. Leuk Lymphoma, 2013.
treatment evaluation of gastric MALT lymphoma. Gut,
2003;52:1656. Rinaldi A, Mian M, Chigrinova E, et al. Genome wide DNA-
profiling of marginal zone lymphomas identifies subtype-
Copie-Bergman C, Wotherspoon AC, Capella C, et al. Gela
specific lesions with an impact on the clinical outcome.
histological scoring system for post-treatment biopsies of
Blood, 2011;117:1595–1604.
patients with gastric MALT lymphoma is feasible and
reliable in rountine practice. Br J Haematol, Rossi D, Trifonov V, Fangazio M, et al. The coding genome
2013;160:47–52. of splenic marginal zone lymphoma: activation of
NOTCH2 and other pathways regulating marginal zone
de Boer JP, Hiddink RF, Raderer M, et al. Dissemination
development. J Exp Med 2012;209:1537–1551.
patterns in non-gastric MALT lymphoma.
Haematologica, 2008;93:201–206. Ruskone-Fourmestraux A, Dragosics B, Morgner A,
Wotherspoon A, De Jong D. Paris staging system for
Dreyling M, Thieblemont C, Gallamini A, et al. ESMO
primary gastrointestinal lymphomas. Gut,
Consensus conferences: guideliness on malignant
2003;52:912–913.
lymphoma. part 2: marginal zone lymphoma, mantle
cell lymphoma, T-cell lymphoma. Ann Oncol, Salido M, Baro C, Oscier D, et al. Cytogenetic aberrations
2013;24:857–877. and their prognostic value in a series of 330 splenic 119
Chapter 7: MALT and other MZLs
marginal zone B-cell lymphomas: a multicenter study of aggressive clinical behaviour. Leuk Lymphoma,
the Splenic B-Cell Lymphoma Group. Blood, 2001;42:823–829.
2010;116:1479–1488. Williams ME, Hong F, Kahl B, et al. Results of E4402
Stathis A, Chini C, Bertoni F, et al. Long-term outcome (RESORT): a randomized phase III study comparing two
following Helicobacter pylori eradication in a different rituximab dosing strategies for low tumor
retrospective study of 105 patients with localized gastric burden indolent B-cell lymphoma. Proc ASCO, 2012;
marginal zone B-cell lymphoma of MALT type. Ann abstract 8007.
Oncol 2009;20:1086–1093. Zucca E, Dreyling M. Gastric marginal zone lymphoma of
Stathis A, Bertoni F, Zucca E. Treatment of gastric marginal MALT type: ESMO Clinical Practice Guidelines for
zone lymphoma of MALT type. Expert Opin diagnosis, treatment and follow-up. Ann Oncol, 2010;21
Pharmacother, 2010;11:2141–2152. Suppl 5:v175–176.
Thieblemont C, Berger F, Dumontet C, et al. Mucosa- Zucca E, Bertoni F, Roggero E, Cavalli F. The gastric
associated lymphoid tissue lymphoma is a disseminated marginal zone B-cell lymphoma of MALT type. Blood,
disease in one third of 158 patients analyzed. Blood, 2000;96:410–419.
2000;95:802–806. Zucca E, Conconi A, Pedrinis E, et al.
Thieblemont C, Felman P, Callet-Bauchu E, et al. Splenic Nongastric marginal zone B-cell lymphoma of
marginal-zone lymphoma: a distinct clinical and mucosa-associated lymphoid tissue. Blood,
pathological entity. Lancet Oncol, 2003;4:95–103. 2003;101:2489–2495.
Tsang RW, Gospodarowicz MK. Radiation therapy for Zucca E, Conconi A, Laszlo D, et al. Addition of
localized low-grade non-Hodgkin’s lymphomas. Hematol rituximab to chlorambucil produces superior
Oncol, 2005;23:10–17. event-free survival in the treatment of patients with
Wenzel C, Dieckmann K, Fiebiger W, et al. CD5 expression extranodal marginal-zone B-cell lymphoma: 5-year
in a lymphoma of the mucosa-associated lymphoid tissue analysis of the IELSG-19 Randomized Study. J Clin
(MALT)-type as a marker for early dissemination and Oncol 2013;31:565–572.
120
Chapter
8
Small lymphocytic lymphoma/chronic
lymphocytic leukemia
Peter Hillmen
Pathology: Andrew Wotherspoon Molecular cytogenetics: Andreas Rosenwald and German Ott
Lymphoma, Second Edition, ed. Robert Marcus, John W. Sweetenham, and Michael E. Williams. Published by 121
Cambridge University Press. © Cambridge University Press 2014.
Chapter 8: Small lymphocytic lymphoma/chronic lymphocytic leukemia
Figure 8.1 B-cell small lymphocytic lymphoma/chronic Figure 8.2 B-cell small lymphocytic lymphoma/chronic
lymphocytic leukemia (lymph node). The nodal architecture is lymphocytic leukemia (bone marrow). There is an infiltrate of small
effaced by a diffuse proliferation of small cells with scanty cytoplasm lymphocytes in the bone marrow with a nodular and interstitial
and round nuclei. Occasional larger cells are present singly and in growth pattern.
small groups (proliferation centers/pseudofollicles).
translocation in rare cases. While translocations stratifying patients to specific therapies. In contrast,
appear to be uncommon in B-CLL, chromosomal the more recently described prognostic factors can
imbalances at several loci characterize this leukemia more accurately predict the outcome for individual
on the genetic level. In particular, frequent aberrations patients, and these have now been validated by large
include deletions in chromosomal band 13q14 that prospective studies. These newer biological variables,
lead to the loss of microRNA miR-15a/miR-16–1 clus- which are now being used to stratify treatment in
ter, detected in up to 40% of cases, trisomy 12, and current clinical studies, fall into four principal types.
deletions in 17p13 and 11q22–23. Deletions in 17p
1. Those that are inherent to the individual patient’s
frequently target the TP53 gene, and 11q deletions
disease and which will not alter during the course
commonly result in the loss of one copy of the ATM
of their illness – such as somatic mutation of the
(ataxia telangiectasia-mutated) gene as well as the miR-
immunoglobulin gene.
34b/miR-34c cluster. The del(11q) has also been asso-
2. Genetic abnormalities that develop during the
ciated with mutation of the spliceosome gene SF3B1.
disease course and which are indicative of genetic
Both aberrations have been shown to be of prognostic
evolution, and in some cases therapeutic
significance in defining patient subgroups with infe-
resistance, of disease – such as deletion of the short
rior prognosis and are commonly found in the IgVH-
arm of chromosome 17, studied by fluorescent in
unmutated B-CLL variant. Interaction of the miR and
situ hybridization (FISH).
TP53 aberrations has been associated both with CLL
pathogenesis and patient prognosis.
Despite the presence of IgVH-mutated and IgVH-
unmutated B-CLL subsets, there is evidence from Table 8.1 Rai staging system for CLL.
DNA microarray studies that the two B-CLL sub- Stage Risk Clinical Median
groups share a homogeneous gene expression profile. features survival
B-CLL, therefore, can be viewed as a single entity with
0 Low Lymphocytes >5 >10 years
a distinct transcriptional profile, distinguishing it from × 109/L and
other low-grade lymphomas. >40% lympho-
In contrast to both B-CLL and mantle cell lym- cytes in the
marrow
phoma, IgVH genes in B-PLL are generally mutated.
According to older reports, the most prevalent cyto- I Intermediate Stage 0 plus 7 years
enlarged lymph
genetic marker is a rearranged chromosome 14 with nodes
14q32 breaks. Other genetic aberrations in B-PLL II Stages 0 or I with
include trisomy 12 and structural rearrangements of an enlarged liver
1p32 and 6q21. Previously, a t(11;14)(q13;q23) trans- or spleen
location was considered to be a recurring aberration in III High Stages 0, I, or II 1.5–4
this tumor, but a critical review of these cases points to with hemoglo- years
bin <100 g/L
a significant overlap with a more refined classification
of these tumors as mantle cell lymphoma. IV Stages 0, I, II, or III
with platelets
<100 × 109/L
Therapy
Table 8.2 Binet staging system for CLL.
Treatment of CLL by risk stratification
Conventional prognostic variables, such as clinical Stage Clinical features Median
stage (either Rai or Binet staging; Tables 8.1 and 8.2) survival
or lymphocyte doubling time, have been used to pre- A <3 areas of lymphadenopathy and 12 years
dict the proportion of patients who will progress to no anemia or thrombocytopenia
treatment or who will respond well to therapy. These B 3 or more areas of lymphadeno- 7 years
variables are useful for predicting the outcomes of pathy and no anemia or
thrombocytopenia
groups of patients (for example, when assessing the
merits of different trials), but are not useful for pre- C Hemoglobin <100 g/L and/or 2–4 years
platelets <100 × 109/L 123
dicting the outcome of individual patients or for
Chapter 8: Small lymphocytic lymphoma/chronic lymphocytic leukemia
Table 8.3 Criteria for NCI Working Group complete response. Table 8.4 Comparison of methods of residual disease
monitoring in CLL.
B-symptoms Absent
MRD flow Allele-specific
Lymph nodes Not palpablea
cytometry oligonucleotide
Liver/spleen Not palpablea PCR
Peripheral blood lymphocytes ::: 4 × 109/L Applicable >95% 85–95%
Peripheral blood neutrophils ::: 1.5 × 109/L patients
(2) Evidence of progressive bone marrow failure with either progressive anemia (unless there is another cause) or thrombocytopenia
(platelets <100 × 109/L). If the cytopenia is stable, then treatment may not be immediately required.
(3) Autoimmune anemia and/or thrombocytopenia poorly responsive to corticosteroids or other standard therapy.
(4) Massive (i.e. >6 cm below the costal margin) or progressive or symptomatic splenomegaly.
(5) Massive lymphadenopathy (i.e. >10 cm in longest diameter) or progressive or symptomatic lymphadenopathy.
(6) Progressive lymphocytosis with an increase of >50% over a 2-month period, or anticipated doubling time of <6 months. (Doubling
time is not used as a single treatment parameter if initial lymphocyte count <30 000/mcl.)
(7) Marked hypogammaglobulinemia or the development of a monoclonal pattern, in the absence of any of the above criteria for active
disease, is not sufficient for protocol therapy.
This broadly captures the patient population that the combination of fludarabine, cyclophosphamide,
should be considered for alternatives to FCR, although and rituximab (FCR). Even in patients considered fit
the view of the patient and their social circumstances enough for this therapy, only three-quarters complete
will also influence their choices. the full six cycles. The median age at diagnosis of CLL is
It is also important to consider the biological 71 years and, therefore, with current therapies, the aim
characteristics of the patient’s CLL when selecting of treatment in most is disease control rather than
the most appropriate therapy. Currently, chromoso- eradication of detectable disease and prolonged remis-
mal analysis by FISH influences the choice of treat- sion. However, studies assessing the quality of life in
ment. FISH is used to identify patients with deletion patients following therapy have demonstrated that
of the short arm of chromosome 17 (17p deletion), better remissions are associated with a better quality of
which is the location of the TP53 gene. This is the life for longer. Therefore if a patient is fit enough for
only abnormality at present that is routinely being FCR-like therapy then they should be offered this.
used to modify therapeutic choices. It is clear that an
intact TP53 gene is necessary for the cell to react
appropriately to genotoxic therapies, such as chemo- Alkylating agents in CLL
therapy. Therefore it is not surprising that patients Randomized controlled trials in the 1990s demonstra-
with 17p deletion respond poorly to chemotherapy. ted that alkylating agents used as monotherapy were
Hallek et al. demonstrated in the German CLL8 Trial equally effective when compared to combinations of
that 90% of previously untreated patients without alkylating agents with steroids and vinca alkaloids
17p deletion will respond to FCR, with almost 50% (CVP) or with the inclusion of an anthracycline
achieving a complete remission (CR), whereas only (CHOP). Randomized trials comparing purine analogs
65% of patients with a 17p deletion will respond, with with alkylating agents have all demonstrated a signifi-
only 5% achieving a CR. Therefore, patients with 17p cantly higher response rate (in the region of 15% com-
deletion should be considered for alternative thera- plete response for fludarabine monotherapy, compared
pies that do not depend on an intact p53 pathway for to <5% for chlorambucil) and improved PFS for purine
their efficacy. At present monoclonal antibodies, analogs (mainly fludarabine), but no improvement in
such as alemtuzumab, and corticosteroids are the OS (median PFS for fludarabine monotherapy in the
routinely available therapies that work in a p53-inde- range of 18–24 months). It appears that this lack of
pendent manner. FISH is widely available but, improvement in OS is largely due to the crossover
increasingly, p53 mutational analysis is also being from chlorambucil to fludarabine. Therefore, until
used in clinical trials and is likely to enter routine recently chlorambucil remained the therapy most com-
practice in the near future. A proportion of patients monly used as the initial therapy in CLL, and it remains
have mutated p53 with normal FISH. It appears that an option for elderly and medically unfit patients.
these patients have a similarly poor outcome to those
with 17p-deleted disease and in future are likely to be
included in the 17p-deleted/p53 dysfunctional cate- Choice of first line therapy in CLL
gory of CLL. Previously untreated patients without 17p deletion
At present neither the other poor biological who are fit
markers of CLL, such as deletion of the long arm of Improved proportions of patients achieving complete
chromosome 11 (11q del) or unmutated immunoglo- remissions following the combination of FCR were first
bulin genes, nor the good biological markers, such as reported approximately 10 years ago by Keating and
isolated deletion of the long arm of chromosome 13 colleagues. This has subsequently been shown to result
(13q del) or mutated immunoglobulin genes, influ- in significantly better PFS and OS when compared to
ence the choice of treatment. However, there are the previous gold standard fludarabine plus cyclophos-
ongoing trials specifically using these criteria to influ- phamide in the German CLL Study Group CLL8 Trial.
ence therapy. The latest updates from the German CLLSG CLL8 Trial
The aim of therapy will depend on the fitness of the indicate that patients treated with FCR as their initial
patient as well as the prognosis in each individual’s case. therapy have a median time to progression in excess of 6
At present the more effective therapies in terms of depth years. Therefore FCR is now the gold standard for
and duration of remission are more intensive, such as patients who have no significant comorbidities. 127
Chapter 8: Small lymphocytic lymphoma/chronic lymphocytic leukemia
Table 8.6 Randomized controlled trials comparing fludarabine monotherapy with fludarabine plus cyclophosphamide.
for 17p deleted patients was less than 5% and all these response rates, patients are continuing to relapse
patients had died within 4 years of their initial therapy. early and at present allogeneic stem cell transplant
This has led to the search for therapies that do not in the first remission is advised for the minority
depend on an intact p53 to be effective. The current of patients where this is an option. We are also
therapies that are p53-independent are either mono- looking at improving the effectiveness of therapy
clonal antibodies or high doses of corticosteroids as by adding other potentially non-p53-dependent
well as allogeneic stem cell transplantation (SCT). therapies, such as lenalidomide (the UK
The current trials of therapy for 17p-deleted NCRI CLL210 trial has just opened which looks
CLL have combined alemtuzumab (Campath) with at alemtuzumab [Campath], dexamethasone, and
high-dose steroids (i.e. CamPred in the UK NCRI lenalidomide [CamDexRev]), with either the
CLL206 trial [Pettitt et al., 2012] and CamDex in patients going onto allogeneic SCT if applicable
the German CLLSG CLL2O Trial [Stilgenbauer or being randomized to lenalidomide maintenance
et al., 2011]). Both of these studies have resulted or observation.
in higher response rates and complete remission Of considerable interest at present is the activity of
rates when compared to historical experience with the novel agents in development for CLL in 17p-
conventional chemotherapy, even including FCR. deleted CLL. It seems that the B-cell receptor pathway
With CamPred 17 patients were treated in the antagonists, either the Bruton’s tyrosine kinase inhib-
frontline setting and all 17 patients responded, itor (PCI-32765 or ibrutinib) or the PI3 kinase d
with 65% achieving a complete response. With (CAL-101/GS-1101 or idelalisib) may well be active
CamDex in the frontline setting 97% of patients in 17p-deleted CLL and could alter the treatment of
responded, whereas only 20% have thus far been these poor-risk patients.
reported to have experienced a CR although the
follow-up in this trial is less mature. It is also
clear that the depth of remission and tolerability Monoclonal antibody therapy in CLL
of such therapies is better if patients have them as Until recently the only monoclonal antibody that was
their first therapy rather than in second or subse- approved for use in CLL as a monotherapy was alem-
quent line. It is therefore important to test for 17p tuzumab (Campath or MabCampath), which was
deletion prior to any line of therapy so that the licensed for fludarabine refractory CLL. The response
patient is not exposed to ineffective but potentially rates to single agent alemtuzumab in relapsed, refrac-
toxic therapy. It is also clear that, despite higher tory CLL range between 33% and 50%, with up to 25%
129
Chapter 8: Small lymphocytic lymphoma/chronic lymphocytic leukemia
of patients achieving a complete response. The most Choice of therapy in relapsed CLL
important predictor of response to alemtuzumab is the Although in some patients the choice of the initial
presence or absence of significant lymphadenopathy. therapy is difficult, the choice of therapy in relapsed
Patients with massive lymphadenopathy (single lymph and refractory disease is often even more challenging.
nodes greater than 5 cm in diameter) have a very low It will depend on a variety of factors, including the
response rate. In these patients a more effective strat- following:
egy is to control the lymphadenopathy with an alter-
native therapy, such as high-dose methylprednisolone, 1. Previous treatment received: it is often difficult to
prior to alemtuzumab therapy. Two recently reported deliver intensive therapies, such as FCR, to patients
phase II trials of subcutaneous alemtuzumab in flu- who have previously received six cycles of FCR,
darabine-refractory CLL suggest a similar efficacy especially if this was relatively recently. In addition,
compared to when the drug is given intravenously there are increasing concerns over the potential
but with a much improved toxicity profile. In August long-term bone marrow consequences of multiple
2012, Genzyme surrendered the license for alemtuzu- episodes of purine analogs. Unless the remission
mab, pending regulatory approval to re-introduce it as after such therapy is prolonged (at least 2 years and
a treatment for multiple sclerosis. Alemtuzumab is still perhaps even 3 years), it is probably better to
available for CLL through an Access Programme but it consider alternative therapies if available.
is likely that it’s use will fall despite a long history of 2. Quality of the latest remission: the depth and
effectiveness. duration of the previous remission will define
Ofatumumab (Arzerra), a second generation fully whether a similar type of therapy as was previously
human anti-CD20 antibody, was approved by FDA in used should be re-used. For example, patients who
2009 and the European Medicines Agency in 2010 for have a remission lasting less than 2 years after FCR
CLL that is refractory to fludarabine and alemtuzu- are considered to have relapsed early and should
mab. Ofatumumab is well tolerated with infusion have a change of therapy where possible. In this
reactions being the major complication with approx- situation alemtuzumab-based therapy may be an
imately 50% of patients responding with a median option, either alone or in combination with
progression free survival of approximately 6 months. corticosteroids. In patients who are considered
Ofatumumab provides a useful option for patients unsuitable for such therapy, by way of either their
who have failed conventional chemotherapies partic- age and/or fitness, the aim of therapy will probably
ularly those with cytopenias preventing further mye- be different and, if it is palliative, steroids alone or
lotoxic therapy. therapies such as single agent ofatumumab might
Rituximab (Rituxan or MabThera) has also been be considered.
used in large numbers of patients with CLL, both alone 3. Physical condition of the patient: not only age
and in combination. Rituximab as a single agent used and frailty but also the history of infections
at the conventional dose of 375 mg/m2 weekly for 4 (patients with recurrent lower chest infections are
weeks has little efficacy in relapsed or refractory CLL, a particular concern) and the degree of cytopenias
with only partial remissions observed in a minority of will influence the choice of therapy. Whether a
patients, and these remissions only persist for a few patient can be treated with intensive therapies such
months. The partial response rate increases with as FCR, might be considered for an allogeneic stem
higher doses of rituximab but complete responses cell transplant, or even less intensive therapies such
were not achieved and the doses used were extremely as bendamustine-based treatments is very much an
high (up to 2250 mg/m2). The use of “conventional” individual clinical decision.
doses of rituximab has also been reported in untreated 4. Degree of marrow recovery: after FCR-like
CLL, with higher response rates (up to 50% of therapies the degree of bone marrow recovery is
patients), but still very few complete responses, and variable. Many patients might have a mild to
these are not durable. Rituximab therefore appears to moderate degree of thrombocytopenia or,
have no role as monotherapy in CLL, but has an occasionally, neutropenia and anemia, after
important place in combination with chemotherapy completion of their prior therapy. A quarter of
as the initial therapy for CLL, at least in a proportion of patients will not have completed the six cycles of
130 patients (see above). FCR previously, often because of persistent
Chapter 8: Small lymphocytic lymphoma/chronic lymphocytic leukemia
cytopenias. In these patients an alternative to on a number of factors, including the patient’s age and
fludarabine-based therapies is preferred, and comorbidities as well as the depth of remission at the
treatments such as bendamustine with rituximab time of the transplant. It is now possible to define a
might be considered. group of patients who are unlikely to respond with
5. Philosophy of the patient: the central role of the durable remissions to conventional therapy and whose
patient and his or her family in deciding which expected survival is poor. Since allogeneic stem cell
therapy to choose cannot be underestimated. For transplant is potentially curative it should be consid-
example, some patients will be searching for a ered at an early stage as soon as a patient is known to
prolonged durable remission or even cure almost have a poor prognosis. The rationale for considering
regardless of the effort and risk involved, and such transplantation earlier rather than later is that the
patients are likely to want very intensive therapies, chances of a successful outcome are related to a large
such as allogeneic stem cell transplant, or might be degree to the quality of remission going into the trans-
prepared to travel distances to enter clinical trials plant and to the physical condition of the patient at
of novel therapies (of which there are currently that time. There are clear European Guidelines for
many [see below]). On the other hand, some who should be selected for allogeneic SCT as follows:
patients may prefer a less intensive approach with
– the presence of deletion of chromosome 17p
less disruption, may only want oral therapies
– relapse within 2 years of FCR-like therapy
rather than parenteral treatment, or even a purely
– relapse within 12 months of fludarabine
palliative approach.
monotherapy (or similar).
6. p53 pathway abnormalities: it is important to
repeat the cytogenetic examination by FISH and Novel therapeutic agents
TP53 mutation analysis (if available) prior to any
change or reinitiation of therapy. The frequency of There has been an explosion in the number of novel
p53 pathway abnormalities increases with therapeutic targets in CLL and an even greater num-
increasing numbers of prior therapies and the ber of potential agents. This has been driven to a large
degree of resistance to treatment, and when they extent by our changing and increasing understanding
occur in relapsed/refractory disease their of the pathophysiology of CLL. It is clear now that
implications are similar to the frontline. CLL is not, as previously often thought, simply the
accumulation of a population of mature lympho-
7. The availability of licensed and trial treatments:
cytes. It is now apparent that CLL is actually a very
there are now a number of approved therapies for
proliferative disorder, demonstrated by the finding
both frontline and relapsed CLL, including
that in some patients up to 2% of the CLL population
chlorambucil, steroids (available but not approved
is being “born” every day. This proliferation is
for CLL), fludarabine-based therapy, cladribine,
driven, at least in part, by the interaction between
deoxycoformycin, bendamustine, rituximab in
the CLL cell and its microenvironment, and is medi-
combination, ofatumumab, alemtuzumab, and a
ated by stimulation, presumably due to antigen,
number of novel therapies in early- and late-stage
through the B-cell receptor (the surface immunoglo-
clinical trials as well as the possibility of allogeneic
bulin on the CLL cell). This proliferation is counter-
SCT. The choice of therapy will obviously depend
balanced by a high rate of apoptosis within the clone.
on whether this treatment is approved locally and
The rate of growth (simply the lymphocyte doubling
whether there is funding.
time) is dependent on the balance between prolifer-
ation and apoptosis. This gives us several potential
Role of allogeneic stem cell transplantation in CLL novel therapeutic targets that are proving to be
CLL is currently incurable and most patients, even extremely promising and may well redefine the way
those with a good response to treatment, will eventu- we treat CLL. The novel therapeutic approaches are
ally relapse and die as a result of their CLL. The current as follows.
exception is allogeneic stem cell transplant, which
offers the opportunity of cure to the patient. 1. Interfering with the microenvironment and its
However, this comes with a relatively high price both interaction with the CLL cell
in terms of the risk of mortality and of significant Drugs such as lenalidomide have activity in CLL and it 131
toxicity. The success of allogeneic SCT in CLL depends appears likely that this is mediated through its influence
Chapter 8: Small lymphocytic lymphoma/chronic lymphocytic leukemia
over the interaction between CLL cells and their micro- usually mild and reversible abnormalities in liver func-
environment. In addition, other therapies, such as tion tests. Grade 3 or 4 serious adverse events are rare.
targeting CXCR4 with plerixafor, are being tested. Combination with other agents. The combination
of these agents with either chemotherapy, such as
bendamustine with or without rituximab, or with
2. Targeting downstream signaling from the B-cell receptor anti-CD20 antibodies (rituximab or ofatumumab)
It is becoming clear that signaling through the B-cell seeks to ameliorate or resolve the lymphocytosis and
receptor is of central importance in the pathophysiol- is currently being studied in a number of randomized
ogy of CLL, and it appears that engagement of the phase III trials.
B-cell receptor leads to a proliferative signal being
passed through a series of tyrosine kinases. This pro-
vides a series of novel targets in CLL in that interfering 3. Apoptosis
with the transmission of this stimulus offers potential The other side of the CLL pathophysiological see-saw
novel therapeutic targets. There are several different is cell death or apoptosis. CLL cells invariably have a
potential targets in this signaling pathway, including high expression of BCL-2 that in part leads to the
Src kinase, Bruton’s tyrosine kinase (Btk) and protection of the cell against apoptosis. It appears
phosphatidyl-inositol-3-kinase δ (PI3kd). The most that this overexpression of BCL-2 is caused by the
advanced agents in this class are two orally active loss of two microRNAs, miR15a and miR16–1, that
drugs, namely ibrutinib (formerly known as PCI- negatively regulate the transcription of BCL-2.
32765), which inhibits Btk, and idelalisib (GS-1101, Therefore BCL-2 is a potential target in CLL, but
formerly known as CAL-101), which inhibits PI3kd. unfortunately previous attempts to target BCL-2
Both of these agents have shown a similar pattern of with either antisense technology, namely oblimersen,
response when used as a single agent in relapsed and or small molecules, for example with obatoclax, have
refractory CLL. failed to show sustained responses and the subse-
Pattern of response. Patients experience an almost quent trials failed to demonstrate significant enough
immediate improvement in their constitutional activity for approval. These attempts to inhibit the
symptoms associated with CLL and within days or a activity of BCL-2 were proof of principle that this
couple of weeks notice an improvement in, or in approach might be an effective therapeutic strategy if
some cases, resolution of, their lymphadenopathy. a selective and high-affinity inhibitor could be
At the same time there is an increase in the lympho- identified.
cytosis to sometimes two or three times the starting The recent report of the BH3 mimetic, navitoclax
value, and this lymphocytosis peaks at approximately or ABT-263, is an alternative strategy to inhibiting
2 months after which it then begins to fall. In some the BCL-2 family of proteins. Navitoclax binds to
cases the lymphocytosis falls to below the starting BCL-2, BCL-xL and BCL-w, releasing the proapop-
point and may normalize, whereas in others it totic molecules, BAX and BAK, which then oligo-
plateaus. Therefore patients experience a “nodal merize on the mitochondrial outer membrane,
response” with lymphocytosis and, over time, this triggering apoptosis. Roberts et al. reported a phase
usually converts to a partial remission or even com- I trial of navitoclax in relapsed and refractory CLL,
plete remission. The pattern of response is intriguing demonstrating clinical activity, with 19 of 21 patients
and is the subject of further studies. It appears that having at least a 50% reduction in their lymphocyto-
these agents initially redistribute the CLL cells from sis and with nine (35%) of 26 patients achieving a
the tissue compartment into the peripheral blood and partial remission. Also, the activity of navitoclax
then perhaps, because of the inhibition of prolifera- appears to be seen in patients with abnormalities of
tion, there is a natural decay of the remaining lym- the p53 pathway (i.e. those with deletion of chromo-
phocytes. Effectively they appear to tip the balance of some 17p). However, the main dose-limiting side
proliferation:apoptosis in favor of cell death. effect of navitoclax is thrombocytopenia, almost cer-
Side-effect profile. These drugs are generally asso- tainly because of its activity against BCL-XL, which is
ciated with relatively mild side effects that are some- expressed by platelets. The next generation of BH3
what variable. For example, inhibition of Btk appears mimetics, ABT-199 or GDC-0199, is specific for
132 to result in a self-limiting and usually relatively mild BCL-2 and has relatively little activity against BCL-
diarrhea. GS-1101 (inhibition of PI3kd) can lead to XL. These agents promise to be effective without the
Chapter 8: Small lymphocytic lymphoma/chronic lymphocytic leukemia
Initial therapy
Symptomatic Asymptomatic
No comorbidity
Watchful
Fludarabine/cyclophosphamide (FC)
waiting, with
With comorbidity
treatment at
Chlorambucil monotherapy
progression
Under investigation
Rituximab + fludarabine ± cyclophosphamide
Alemtuzumab consolidation
Chemoimmunotherapy Investigational
As above Stem cell transplantation
Pentostatin/cyclophosphamide Various monoclonal antibodies
±Rituximab (PCR)
Alemtuzumab Lenalidomide
Bendamustine ± Rituximab BCL-2 antisense therapy
Pulse corticosteroids B-cell receptor pathway inhibitors
Figure 8.5 Therapeutic approaches in small lymphocytic lymphoma/CLL. As standard therapies for front-line and for relapsed or refractory
disease are not yet established, treatment of these patients on a clinical trial is encouraged.
localization into lipid rafts after binding of the mono- Recommended therapeutic
clonal antibody to CD20, but increased antibody-
dependent cellular cytotoxicity (ADCC) related to an approaches in CLL
improved binding of GA-101 to the different allotypes As a general rule, patients should be offered entry into
of FcgRIIIa expressed by natural killer (NK) cells and well-designed clinical trials if available and if the
monocytes. Compared with rituximab, the increased patient is eligible. A treatment algorithm is shown in
direct cell death induction related to an elbow hinge Figure 8.5.
amino exchange of the Fab region and type II binding
of the CD20 epitope, GA-101 has shown promising
activity in small phase I trials and is now being studied Conclusion
in a randomized phase III trial in combination with Whilst the therapy of CLL has improved over the past
chlorambucil, compared to both chlorambucil alone decade or so the current treatments have many prob-
134 and chlorambucil plus rituximab (the German CLL11 lems. The most effective treatments are associated with
Trial). significant short- and long-term toxicities, which means
Chapter 8: Small lymphocytic lymphoma/chronic lymphocytic leukemia
that they are only an option for a proportion of younger Caligaris-Cappio F. Role of the microenvironment in
fitter patients. In these patients they are not curative and chronic lymphocytic leukaemia. Br J Haematol,
potentially lead to long-term toxicities that may com- 2003;123(3):380–388.
promise subsequent therapeutic interventions. The CLL Trialists’ Collaborative Group. Chemotherapeutic
treatment of more elderly patients or those with comor- options in chronic lymphocytic leukemia: a meta-analysis
bidities, who form the majority of patients with CLL, of the randomized trials. JNCI J Natl Cancer Inst,
1999;91:861–868.
has hardly advanced over the past few decades. The
increased understanding of the pathophysiology of Davids MS, Lannutti BJ, Brown JR, and Letai AG. BH3
profiling demonstrates that restoration of apoptotic
CLL and the rapid development of a variety of novel
priming contributes to increased sensitivity to PI3K
targeted therapies has the potential to change the treat- inhibition in stroma-exposed chronic lymphocytic
ment landscape in CLL completely. The challenge that leukemia cells. Blood (ASH Annual Meeting Abstracts),
we now face is how best to utilize these new agents to 2011;118:974.
confer the most rapid and effective change in the treat- Döhner H, Stilgenbauer S, Benner A, et al. Genomic
ment paradigm of CLL to optimize outcomes for all aberrations and survival in chronic lymphocytic
patients with the disease. leukemia. N Engl J Med, 2000;343(26):1910–1916.
Dreger P, Corradini P, Kimby E, et al. Chronic Leukemia
Working Party of the EBMT. Indications for allogeneic
Further reading stem cell transplantation in chronic lymphocytic
leukemia: the EBMT transplant consensus. Leukemia,
Andritsos LA, Byrd JC, Hewes B, et al. Preliminary results 2007;21(1):12–17.
from a phase I/II dose escalation study to determine the
maximum tolerated dose of plerixafor in combination Dreger P, Dohner H, Ritgen M, et al. Allogeneic stem cell
with rituximab in patients with relapsed chronic transplantation provides durable disease control in poor-
lymphocytic leukemia. Haematologica, 2010;95 risk chronic lymphocytic leukemia: long-term clinical
(Suppl.2): Abstract 0772. and MRD results of the German CLL Study Group
CLL3X trial. Blood, 2010;116:2438–2447.
Binet JL, Auquier A, Dighiero G, et al. A new prognostic
classification of chronic lymphocytic leukaemia derived Eichhorst BF, Busch R, Hopfinger G, et al. The German CLL
from a multivariate survival analysis. Cancer, Study Group. Fludarabine plus cyclophosphamide versus
1981;48:198–206. fludarabine alone in first-line therapy of younger patients
with chronic lymphocytic leukemia. Blood,
Bologna L, Gotti E, Manganini M, et al. Mechanism of
2006;107:885–891.
action of type II, glycoengineered, anti-CD20
monoclonal antibody GA101 in B-chronic lymphocytic Eichhorst BF, Busch R, Stilgenbauer S, et al. German CLL
leukemia: whole blood assays in comparison with Study Group (GCLLSG). First-line therapy with
rituximab and alemtuzumab. J Immunol, fludarabine compared with chlorambucil does not result
2011;186(6):3762–3769. in a major benefit for elderly patients with advanced
Brentjens RJ, Rivière I, Park JH, et al. Safety and persistence chronic lymphocytic leukemia. Blood,
of adoptively transferred autologous CD19-targeted T 2009;114(16):3382–3391.
cells in patients with relapsed or chemotherapy refractory Fabbri M, Bottoni A, Shimizu M, et al. Association of a
B-cell leukemias. Blood, 2011;118:4817–4828. microRNA/TP53 feedback circuitry with pathogenesis
Burger JA, Ghia P, Rosenwald A, Caligaris-Cappio F. The and outcome of B-cell chronic lymphocytic leukemia.
microenvironment in mature B-cell malignancies: a JAMA, 2011;305:59–67.
target for new treatment strategies. Blood, Fischer K, Cramer P, Busch R, et al. Bendamustine
2009;114(16):3367–3375. combined with rituximab in patients with relapsed and/
Byrd JC, Lin TS, Dalton JT, et al. Flavopiridol or refractory chronic lymphocytic leukemia: a
administered using a pharmacologically derived multicenter phase II trial of the German Chronic
schedule is associated with marked clinical efficacy in Lymphocytic Leukemia Study Group. J Clin Oncol,
refractory, genetically high-risk chronic lymphocytic 2011;29(26):3559–3566.
leukemia. Blood, 2007;109:399–404. Flynn JM, Jones JA, Andritsos L, et al. Update on the phase I
Byrd JC, Shinn C, Waselenko JK, et al. Flavopiridol induces study of the cyclin dependent kinase inhibitor dinaciclib
apoptosis in chronic lymphocytic leukemia cells via (SCH 727965) in patients with relapsed or refractory
activation of caspase-3 without evidence of bcl-2 chronic lymphocytic leukemia (CLL): confirmation of
modulation or dependence on functional p53. Blood, clinical activity and feasibility of long-term
1998;92:3804–3816. administration. Blood, 2010;116:1396. 135
Chapter 8: Small lymphocytic lymphoma/chronic lymphocytic leukemia
Furman RR, Byrd JC, Brown JR, et al. CAL-101, an isoform- chronic lymphocytic leukemia B cells. J Clin Invest,
selective inhibitor of phosphatidylinositol 3-kinase P110, 2005;115(3):755–764.
demonstrates clinical activity and pharmacodynamic O’Brien S, Burger JA, Blum KA, et al. The Bruton’s tyrosine
effects in patients with relapsed or refractory chronic kinase (BTK) inhibitor PCI-32765 induces durable
lymphocytic leukemia. Blood (ASH Annual Meeting responses in relapsed or refractory (R/R) chronic
Abstracts), 2010;116:55. lymphocytic leukemia/small lymphocytic lymphoma
Hallek M, Cheson BD, Catovsky D, et al. Guidelines for the (CLL/SLL): follow-up of a phase Ib/II study. Blood, (ASH
diagnosis and treatment of chronic lymphocytic Annual Meeting Abstracts), 2011;118:983.
leukemia: a report from the International Workshop on O’Brien SM, Cunningham CC, Golenkov AK, et al. Phase I
Chronic Lymphocytic Leukemia updating the National to II multicenter study of oblimersen sodium, a Bcl-2
Cancer Institute-Working Group 1996 guidelines. Blood, antisense oligonucleotide, in patients with advanced
2008;111:5446–5456. chronic lymphocytic leukemia. J Clin Oncol,
Hallek M, Fischer K, Fingerle-Rowson G, et al. Addition of 2005;23:7697–7702.
rituximab to fludarabine and cyclophosphamide in O’Brien SM, Claxton DF, Crump M, et al. Phase I study of
patients with chronic lymphocytic leukaemia: a obatoclax mesylate (GX15–070), a small molecule pan-
randomised, open-label, phase 3 trial. Lancet, Bcl-2 family antagonist, in patients with advanced
2010;376(9747):1164–1174. chronic lymphocytic leukemia. Blood, 2009;113:299–305.
Herman SE, Gordon AL, Wagner AJ, et al. Oltersdorf T, Elmore SW, Shoemaker AR, et al. An inhibitor
Phosphatidylinositol 3-kinase-δ inhibitor CAL-101 of Bcl-2 family proteins induces regression of solid
shows promising preclinical activity in chronic tumours. Nature, 2005;435:677–681.
lymphocytic leukemia by antagonizing intrinsic and
extrinsic cellular survival signals. Blood, Oscier D, Wade R, Davis Z, et al. Prognostic factors
2010;116(12):2078–2088. identified three risk groups in the LRF CLL4 trial,
independent of treatment allocation. Haematologica,
Herman SE, Gordon AL, Hertlein E, et al. Bruton tyrosine 2010;95(10):1705–1712.
kinase represents a promising therapeutic target for
treatment of chronic lymphocytic leukemia and is Pettitt AR, Jackson R, Carruthers S, et al. Alemtuzumab in
effectively targeted by PCI-32765. Blood, combination with methylprednisolone is a highly
2011;117(23):6287–6296. effective induction regimen for patients with chronic
lymphocytic leukemia and deletion of TP53: final results
Jaglowski SM, Flynn JM, Jones JA, et al. Flavopiridol is an of the National Cancer Research Institute CLL206 trial. J
effective therapy to bridge patients with chronic Clin Oncol, 2012;30(14):1647–1655.
lymphocytic leukemia (CLL) to reduced-intensity
conditioning allogeneic stem cell transplant. Blood, Porter DL, Levine BL, Kalos M, Bagg A, June CH. Chimeric
2010;116:2383. antigen receptor-modified T cells in chronic lymphoid
leukemia. N Engl J Med, 2011;365(8):725–733.
Keating MJ, O’Brien S, Albitar M, et al. Early results of a
chemoimmunotherapy regimen of fludarabine, Rai KR, Sawitsky A, Cronkite EP, et al. Clinical staging of
cyclophosphamide, and rituximab as initial therapy for chronic lymphocytic leukemia. Blood, 1975;46:219–234.
chronic lymphocytic leukemia. J Clin Oncol, Rai KR, Peterson BL, Appelbaum FR, et al. Fludarabine
2005;23(18):4079–4088. compared with chlorambucil as primary therapy for
Knauf WU, Lissichkov T, Aldaoud A, et al. Phase III chronic lymphocytic leukemia. N Engl J Med,
randomized study of bendamustine compared with 2000;343:1750–1757.
chlorambucil in previously untreated patients with Roberts AW, Seymour JF, Brown JR, et al. Substantial
chronic lymphocytic leukemia. J Clin Oncol, susceptibility of chronic lymphocytic leukemia to BCL2
2009;27(26):4378–4384. inhibition: results of phase 1 study of navitoclax (ABT-
Lanasa MC, Andritsos L, Brown JR, et al. Interim analysis of 263) in patients with relapsed or refractory disease. J Clin
EFC6663, a multicenter phase 2 study of Alvocidib Oncol, 2012;30(5):488–496.
(flavopiridol), demonstrates clinical responses among Schetelig J, van BA, Brand R, et al. Allogeneic hematopoietic
patients with fludarabine refractory CLL. Blood, stem-cell transplantation for chronic lymphocytic
2010;116:58. leukemia with 17p deletion: a retrospective European
Lin TS, Ruppert AS, Johnson AJ, et al. Phase II study of Group for Blood and Marrow Transplantation analysis. J
flavopiridol in relapsed chronic lymphocytic leukemia Clin Oncol, 2008;26:5094–5100.
demonstrating high response rates in genetically high- Sharman J, de Vos S, Leonard JP, et al. A phase 1 study of the
risk disease. J Clin Oncol, 2009;27:6012–6018. selective phosphatidylinositol 3-kinase-delta (PI3K)
Messmer BT, Messmer D, Allen SL, et al. In vivo inhibitor, CAL-101 (GS-1101), in combination with
136 measurements document the dynamic cellular kinetics of rituximab and/or bendamustine in patients with relapsed
Chapter 8: Small lymphocytic lymphoma/chronic lymphocytic leukemia
or refractory chronic lymphocytic leukemia (CLL). Blood Wang L, Lawrence MS, Wan Y, et al. SF3B1 and other novel
(ASH Annual Meeting Abstracts), 2011;118:1787. cancer genes in chronic lymphocytic leukemia. N Engl J
Stilgenbauer S, Cymbalista F, Leblond V, et al. Alemtuzumab Med, 2011;365:2497–2506.
plus oral dexamethasone, followed by alemtuzumab Wierda WG, Padmanabhan S, Chan GW, et al.
maintenance or allogeneic transplantation in ultra high- Ofatumumab is active in patients with fludarabine-
risk CLL: interim analysis of a phase ii study of the refractory CLL irrespective of prior rituximab:
GCLLSG and fcgcll/MW. Blood (ASH Annual Meeting results from the phase 2 international study. Blood,
Abstracts), 2011;118:2854. 2011;118(19):5126–5129.
Tam CS, O’Brien S, Wierda W, et al. Long-term results of the Zenz T, Häbe S, Denzel T, et al. Detailed analysis of p53
fludarabine, cyclophosphamide, and rituximab regimen pathway defects in fludarabine-refractory chronic
as initial therapy of chronic lymphocytic leukemia. Blood, lymphocytic leukemia (CLL): dissecting the contribution
2008;112:975–980. of 17p deletion, TP53 mutation, p53-p21 dysfunction,
Taylor RP, Lindorfer MA. Antigenic modulation and and miR34a in a prospective clinical trial. Blood,
rituximab resistance. Semin Hematol, 2009;114(13):2589–2597.
2010;47(2):124–132. Zenz T, Eichhorst B, Busch R, et al. TP53 mutation and
Tse C, Shoemaker AR, Adickes J, et al. ABT-263: a potent survival in chronic lymphocytic leukemia. J Clin Oncol,
and orally bioavailable Bcl-2 family inhibitor. Cancer Res, 2010;28(29):4473–4479.
2008;68:3421–3428. Zenz T, Vollmer D, Trbusek M, et al. TP53 mutation
van Delft MF, Wei AH, Mason KD, et al. The BH3 mimetic profile in chronic lymphocytic leukemia: evidence
ABT-737 targets selective Bcl-2 proteins and efficiently for a disease specific profile from a comprehensive
induces apoptosis via Bak/Bax if Mcl-1 is neutralized. analysis of 268 mutations. Leukemia,
Cancer Cell, 2006;10:389–399. 2010;24(12):2072–2079.
137
Chapter
9
Waldenström’s macroglobulinemia/
lymphoplasmacytic lymphoma
Steven P. Treon and Giampaolo Merlini
Introduction
Waldenström’s macroglobulinemia (WM) is a distinct
clinicopathological entity resulting from the accumula-
tion, predominantly in the bone marrow, of clonally
related lymphocytes, lymphoplasmacytic cells, and
plasma cells that secrete a monoclonal IgM protein
(Figure 9.1). This condition is considered to correspond
to lymphoplasmacytic lymphoma (LPL) as defined by
the World Health Organization classification system.
Most cases of LPL are WM, with less than 5% of cases
made up of IgA, IgG, and non-secreting LPL.
Epidemiology and etiology Figure 9.1 Aspirate from a patient with Waldenstrom’s
macroglobulinemia demonstrating excess mature lymphocytes,
WM is an uncommon disease, with a geometrical lymphoplasmacytic cells, and plasma cells (courtesy of Marvin
increase with age. The incidence rate for WM is higher Stone M.D.).
among Caucasians, with African descendants repre-
senting only 5% of all patients. Genetic factors appear
molecular diagnostic studies for HCV infection in
to be important to the pathogenesis of WM. A com-
100 consecutive WM patients.
mon predisposition for WM with other malignancies
has been raised, and there have been numerous reports
of familial predisposition, including clustering of fam- Biology
ily members with WM and other B-cell lymphoproli-
ferative diseases. In a recent study, 28% of 924 serial Cytogenetics
WM patients presenting to a tertiary referral had a first Chromosome 6q deletions encompassing 6q21–25 have
or second degree relative with either WM or another been observed in up to half of WM patients, and at a
B-cell disorder. Frequent familial associations with comparable frequency amongst patients with and with-
other immunological disorders in healthy relatives out a familial history. The presence of 6q deletions has
have also been reported. The role of environmental been suggested to discern patients with WM from those
factors in WM remains to be clarified, but chronic with IgM monoclonal gammopathy of unknown signifi-
antigenic stimulation from infections, certain drug cance (MGUS), and to have potential prognostic signifi-
and agent orange exposures remain suspect. An etio- cance, including impact on progression-free survival
logical role for hepatitis C virus (HCV) infection has (PFS) following treatment response, though others
been suggested, though in one study no association have reported no prognostic significance to the presence
could be established using both serological and of 6q deletions in WM. Other abnormalities by
138 Lymphoma, Second Edition, ed. Robert Marcus, John W. Sweetenham, and Michael E. Williams. Published by
Cambridge University Press. © Cambridge University Press 2014.
Chapter 9: Waldenström’s macroglobulinemia/lymphoplasmacytic lymphoma
cytogenetic or fluorescent in situ hybridization (FISH) admixed with tumor aggregates. The role of mast
analyses include deletions in 13q14, TP53, and ATM, cells in WM has been investigated in one study
trisomy 4, 12, and 18. Recently, the somatic point muta- wherein coculture of primary autologous or mast cell
tion MYD 88 L265P (Leucine → Valine) has been iden- lines with WM LPC resulted in dose-dependent WM
tified in 90% of WM samples but this is only rarely cell proliferation and/or tumor colony formation, pri-
present in related B-cell lymphoproliferative disorders. marily through CD40 ligand (CD40L) signaling.
As such, it promises to become a useful diagnostic Furthermore, WM cells, through elaboration of solu-
marker and to provide new pathogenetic insights. ble CD27 (sCD27), induced the upregulation of
CD40L on mast cells derived from WM patients and
mast cell lines, suggesting a microenvironmental sup-
Nature of the clonal cell port system. High levels of CXCR4 and VLA-4 have
The WM bone marrow B-cell clone shows intraclonal also been observed in WM cells. In blocking experi-
differentiation from small lymphocytes with large mental studies, CXCR4 was shown to support migra-
focal deposits of surface immunoglobulins, to lym- tion of WM cells, while VLA-4 contributed to
phoplasmacytic cells, to mature plasma cells that adhesion of WM cells to bone marrow stromal cells.
contain intracytoplasmic immunoglobulins. Clonal
B-cells are detectable among blood B-lymphocytes,
and their number increases in patients who fail to Clinical features
respond to therapy or who progress. These clonal The clinical and laboratory findings at time of diagnosis
blood cells present the peculiar capacity to differentiate of WM in one large institutional study are presented in
spontaneously, in in vitro culture, to plasma cells. This Table 9.1. Unlike most indolent lymphomas, splenome-
is through an interleukin-6 (IL-6)-dependent process in galy and lymphadenopathy are prominent in only a
IgM MGUS and mostly an IL-6-independent process in minority of patients (≤15%). Purpura is frequently asso-
WM patients. All these cells express the monoclonal ciated with cryoglobulinemia and more rarely with AL
IgM present in the blood and a variable percentage of amyloidosis, while hemorrhagic manifestations and
them also express surface IgD. The characteristic neuropathies are multifactorial (see later). The morbid-
immunophenotypic profile of the lymphoplasmacytic ity associated with WM is caused by the concurrence of
cells in WM includes the expression of the pan-B-cell two main components: tissue infiltration by neoplastic
markers CD19, CD20, CD22, CD79, and FMC7.2. cells and, more importantly, the physicochemical and
The phenotype of lymphoplasmacytic cells in WM immunological properties of the monoclonal IgM. As
cells suggests that the clone is a post-germinal center shown in Table 9.2, the monoclonal IgM can produce
B-cell. This indication is further strengthened by the clinical manifestations through several different mech-
results of the analysis of the nature (silent or amino anisms related to its physicochemical properties,
acid replacing) and distribution (in framework or CDR non-specific interactions with other proteins, antibody
regions) of somatic mutations in immunoglobulin activity, and tendency to deposit in tissues.
heavy- and light-chain variable regions performed in
patients with WM. This analysis showed a high rate of
replacement mutations, compared with the closest Morbidity mediated by the effects
germline genes, clustering in the CDR regions and with-
out intraclonal variation. Subsequent studies showed a of IgM
strong preferential usage of VH3/JH4 gene families, no
intraclonal variation, and no evidence for any isotype- Hyperviscosity syndrome
switched transcripts. These data indicate that WM may Blood hyperviscosity is affected by increased serum IgM
originate from an IgM+ and/or IgM+ IgD+ memory levels, leading to hyperviscosity-related complications.
B-cell. Normal IgM+ memory B-cells localize in bone The main determinants are: (1) a high concentration of
marrow, where they mature to IgM-secreting cells. monoclonal IgMs, which may form aggregates and may
bind water through their carbohydrate component; and
(2) their interaction with blood cells. Monoclonal IgMs
Bone marrow microenvironment increase red cell aggregation (rouleaux formation) and
Increased numbers of mast cells are found in the bone red cell internal viscosity while also reducing deform- 139
marrow of WM patients, wherein they are usually ability. The possible presence of cryoglobulins can
Chapter 9: Waldenström’s macroglobulinemia/lymphoplasmacytic lymphoma
Table 9.1 Clinical and laboratory findings for 149 consecutive newly diagnosed patients with the consensus panel diagnosis of WM
presenting to the Dana Farber Cancer Institute.
Table 9.2 Physicochemical and immunological properties of the monoclonal IgM protein in Waldenstrom’s macroglobulinemia.
140
Chapter 9: Waldenström’s macroglobulinemia/lymphoplasmacytic lymphoma
contribute to increasing blood viscosity as well as to the usually occur when the monoclonal IgM concentration
tendency to induce erythrocyte aggregation. Serum exceeds 50 g/L or when serum viscosity is >4.0 centi-
viscosity is proportional to IgM concentration up to poises (cp), but there is a great individual variability,
30 g/L, then increases sharply at higher levels. Plasma with some patients showing no evidence of hypervis-
viscosity and hematocrit are directly regulated by the cosity even at 10 cp. The most common symptoms are
body. Increased plasma viscosity may also contribute to oronasal bleeding, visual disturbances resulting from
inappropriately low erythropoietin production, which is retinal bleeding, and dizziness that may rarely lead to
the major reason for anemia in these patients. Clinical coma. Heart failure can be aggravated, particularly in
manifestations are related to circulatory disturbances the elderly, owing to increased blood viscosity,
that can be best appreciated by ophthalmoscopy, expanded plasma volume, and anemia. Inappropriate
which shows distended and tortuous retinal veins, hem- transfusion can exacerbate hyperviscosity and may pre-
orrhages, and papilledema (Figure 9.2). Symptoms cipitate cardiac failure.
Cryoglobulinemia
In up to 20% of WM patients, the monoclonal IgM can
behave as a cryoglobulin (type I), but it is symptomatic
in 5% or less of the cases. Cryoprecipitation is mainly
dependent on the concentration of monoclonal IgM;
for this reason plasmapheresis or plasma exchange are
commonly effective in this condition. Symptoms
result from impaired blood flow in small vessels and
include Raynaud’s phenomenon, acrocyanosis, and
necrosis of the regions most exposed to cold such as
the tips of the nose, ears, fingers, and toes (Figure 9.3),
malleolar ulcers, purpura, and cold urticaria. Renal
manifestations may occur but are infrequent.
Autoantibody activity
Monoclonal IgM may exert its pathogenic effects
through specific recognition of autologous antigens,
Figure 9.2 Fundoscopic examination of a patient with the most notable being nerve constituents, immunoglo-
Waldenstrom’s macroglobulinemia demonstrating hyperviscosity- bulin determinants, and red blood cell (RBC) antigens.
related changes, including dilated retinal vessels, peripheral
hemorrhages, and “venous sausaging” (courtesy of Marvin Stone M.D.).
141
Chapter 9: Waldenström’s macroglobulinemia/lymphoplasmacytic lymphoma
lymphoplasmacytic cell population constituted by small involvement and the serum level of the IgM monoclo-
lymphocytes with evidence of plasmacytoid/plasma cell nal protein (owing to the impact of IgM on intra-
differentiation (Figure 9.1). The pattern of bone mar- vascular fluid retention), has emerged as a strong
row infiltration may be diffuse, interstitial, or nodular, adverse prognostic factor, with hemoglobin levels of
showing usually an intertrabecular pattern of infiltra- <9–12 g/dL associated with decreased survival in sev-
tion. A solely paratrabecular pattern of infiltration is eral series. Cytopenias have also been regularly identi-
unusual and should raise the possibility of follicular fied as a significant predictor of survival. However, the
lymphoma. The bone marrow infiltration should rou- precise level of cytopenias with prognostic significance
tinely be confirmed by immunophenotypic studies (flow remains to be determined. High beta-2 microglobulin
cytometry and/or immunohistochemistry) showing the levels (>3–3.5 g/dL), a high serum IgM M-protein
following profile: sIgM+CD19+CD20+CD22+CD79+. (>7 g/dL), as well as a low serum IgM M-protein
Up to 20% of cases may express either CD5, CD10, or (<4 g/dL), and the presence of cryoglobulins were
CD23. In these cases, care should be taken to exclude shown in several studies to be adverse factors. A few
chronic lymphocytic leukemia and mantle cell lym- scoring systems have been proposed based on these
phoma satisfactorily. “Intranuclear” periodic acid– analyses (Table 9.3).
Schiff (PAS)-positive inclusions (Dutcher–Fahey
bodies) consisting of IgM deposits in the perinuclear
space, and sometimes in intranuclear vacuoles, may be
Treatment of Waldenström’s
seen occasionally in lymphoid cells in WM. An macroglobulinemia
increased number of mast cells, usually in association
with the lymphoid aggregates, is commonly found in Treatment indications
WM, and their presence may help in differentiating Consensus guidelines on indications for treat-
WM from other B-cell lymphomas. ment initiation were formulated as part of the
Second International Workshop on Waldenström’s
Macroglobulinemia. Initiation of therapy should not
Other investigations be based on the IgM levels since this may not correlate
Magnetic resonance imaging (MRI) of the spine in with either disease burden or symptomatic status.
conjunction with computed tomography (CT) of the Initiation of therapy is appropriate for patients with
abdomen and pelvis are useful in evaluating the disease constitutional symptoms, such as recurrent fever,
status in WM. Bone marrow involvement can be docu- night sweats, fatigue due to anemia, or weight loss.
mented by MRI studies of the spine in over 90% of The presence of progressive, symptomatic lymph-
patients, while CT of the abdomen and pelvis demon- adenopathy or splenomegaly provides additional rea-
strated enlarged nodes in 43% of WM patients. Lymph sons to begin therapy. The presence of anemia with a
node biopsy may show preserved architecture or hemoglobin value of ≤10 g/dL or a platelet count
replacement by infiltration of neoplastic cells with ≤100 × 10/L on this basis of disease is also a reasonable
lymphoplasmacytoid, lymphoplasmacytic, or poly- indication for treatment initiation. Certain complica-
morphous cytological patterns. tions of WM, such as hyperviscosity syndrome, symp-
tomatic sensorimotor peripheral neuropathy, systemic
amyloidosis, renal insufficiency, or symptomatic cryo-
Prognosis globulinemia are also indications for therapy.
WM typically presents as an indolent disease, although
considerable variability in prognosis can be seen. The
median survival reported in several large series has Treatment options
ranged from 5 to 10 years, though in a recent study A precise therapeutic algorithm for therapy of WM
of 436 consecutive patients with WM, the median remains to be defined given the paucity of random-
overall survival (OS) from time of diagnosis was in ized clinical trials. Active agents include alkyla-
excess of 10 years. The presence of 6q deletions as a tors (chlorambucil, cyclophosphamide), nucleoside
prognostic marker remains controversial. Age is con- analogs (cladribine, fludarabine), monoclonal
sistently an important prognostic factor (>60–70 antibodies (rituximab, ofatumumab, alemtuzumab),
144 years), but this factor is often impacted by unrelated bortezomib, thalidomide, everolimus, and bendamus-
morbidities. Anemia, which reflects both marrow tine. Combination therapy, particularly with rituximab,
Chapter 9: Waldenström’s macroglobulinemia/lymphoplasmacytic lymphoma
has been associated with improved clinical outcomes. whilst on the intermittent schedule patients will typi-
Individual patient considerations, including the pres- cally receive 0.3 mg/kg for 7 days, every 6 weeks. In a
ence of cytopenias, need for more rapid disease control, prospective randomized study, Kyle et al. reported no
age, and candidacy for autologous transplant therapy, significant difference in the overall response rate
should be taken into account in making the choice of a between these schedules, although, interestingly, the
first line agent. For patients who are candidates for median response duration was greater for patients
autologous transplant therapy, exposure to continuous receiving intermittent versus continuously dosed chlor-
chlorambucil or nucleoside analog therapy should be ambucil (46 versus 26 months). Despite the favorable
limited given the potential for stem cell damage. The use median response duration in this study for use of the
of nucleoside analogs may also increase risk for histo- intermittent schedule, no difference in the median OS
logical transformation to diffuse large B-cell lymphoma was observed. Moreover, an increased incidence for
(DLBCL), as well as myelodysplasia and acute myelo- development of myelodysplasia and acute myelogenous
genous leukemia. leukemia with the intermittent (3 of 22 patients) versus
the continuous (0 of 24 patients) chlorambucil schedule
Chlorambucil prompted the authors of this study to express prefer-
Oral alkylating drugs, alone and in combination ther- ence for use of continuous chlorambucil dosing. The
apy with steroids, have been extensively evaluated in the use of steroids in combination with alkylator therapy
upfront treatment of WM. The greatest experience with has also been explored. Dimopoulos and Alexanian
oral alkylator therapy has been with chlorambucil, evaluated chlorambucil (8 mg/m2) along with predni-
which has been administered on both a continuous sone (40 mg/m2) given orally for 10 days, every 6 weeks,
(i.e. daily dose schedule) as well as an intermittent and reported a major response (i.e. reduction of IgM by
schedule. Patients receiving chlorambucil on a contin- greater than 50%) in 72% of patients. Additional factors 145
uous schedule typically receive 0.1 mg/kg per day, to be taken into account in considering alkylator
Chapter 9: Waldenström’s macroglobulinemia/lymphoplasmacytic lymphoma
therapy for patients with WM include necessity for Autologous stem cell collection may be impaired in
more rapid disease control given the slow nature of patients who received a nucleoside analog. The long-
response to alkylator therapy, as well as consideration term safety of nucleoside analogs in WM was recently
for preserving stem cells in patients who are candidates examined by Leleu et al. in a large series of WM
for autologous transplant therapy. patients. A sevenfold increase in transformation to
an aggressive lymphoma, and a threefold increase in
Nucleoside analogs the development of acute myelogenous leukemia/
myelodysplasia were observed amongst patients
Both cladribine and fludarabine have been extensively
who received a nucleoside analog versus other thera-
evaluated in untreated as well as previously treated WM
pies for their WM. A recent meta-analysis by Leleu
patients. Cladribine administered as a single agent by
et al. of several trials utilizing nucleoside analogs in
continuous intravenous infusion, by 2-hour daily infu-
WM patients, which included patients who had pre-
sion, or by subcutaneous bolus injections for 5–7 days
viously received an alkylator agent, showed a crude
has resulted in major responses in 40–90% of patients
incidence of 6.6–10% for development of disease
who received primary therapy, whilst in the salvage
transformation, and 1.4–8.9% for development of
setting responses have ranged from 38% to 54%.
myelodysplasia or acute myelogenous leukemia.
Median time to achievement of response in responding
patients following cladribine ranged from 1.2 to 5
months. The overall response rate with daily infusional Monoclonal antibodies
fludarabine therapy administered mainly on 5-day Rituximab is a chimeric monoclonal antibody that
schedules in previously untreated and treated WM targets CD20, a widely expressed antigen on lympho-
patients has ranged from 38% to 100% and 30% to plasmacytic cells in WM. Several retrospective and
40%, respectively, which are on par with the response prospective studies have indicated that rituximab,
data for cladribine. Median time to achievement of when used at standard dosimetry (i.e. 4 weekly infu-
response for fludarabine was also on par with cladribine sions at 375 mg/m2) induced major responses in
at 3–6 months. In general, response rates and durations approximately 27–35% of previously treated and
of responses have been greater for patients receiving untreated patients. Furthermore, it was shown in
nucleoside analogs as first line agents, although in sev- some of these studies that patients who achieved
eral of the above studies wherein both untreated and minor responses or even stable disease benefited
previously treated patients were enrolled, no substantial from rituximab, as evidenced by improved hemoglo-
difference in the overall response rate was reported. bin and platelet counts and reduction of lymphaden-
Myelosuppression commonly occurred following pro- opathy and/or splenomegaly. The median time to
longed exposure to either of the nucleoside analogs, as treatment failure in these studies was found to range
did lymphopenia with sustained depletion of both CD4+ from 8 to 27+ months. Studies evaluating an extended
and CD8+ T-lymphocytes observed in WM patients rituximab schedule consisting of four weekly courses
1 year after initiation of therapy. Treatment-related at 375 mg/m2/week, repeated 3 months later by
mortality due to myelosuppression and/or opportunistic another 4-week course, have demonstrated major
infections attributable to immunosuppression occurred response rates of 44–48%, with time to progression
in up to 5% of all treated patients in some series with (TTP) estimates of 16+ to 29+ months.
either nucleoside analog. Factors predicting for response In many WM patients, a transient increase of serum
to nucleoside analogs in WM included age at start of IgM (IgM flare) may be noted immediately following
treatment (<70 years), pre-treatment hemoglobin initiation of rituximab treatment. The IgM flare may be
>95 g/L, platelets >75 000/mm3, disease relapsing off related to release of IL-6 by bystander immune cells in
therapy, patients with resistant disease within the first response to binding of rituximab to FcγRIIA receptors,
year of diagnosis, and a long interval between first line and also occurs in response to intravenous immuno-
therapy and initiation of a nucleoside analog in relapsing globulin administration in WM patients. The IgM flare
patients. The combination of nucleoside analogs with in response to rituximab does not herald treatment
cyclophosphamide and/or rituximab has been investi- failure, and, while most patients will return to their
gated and is discussed below. baseline serum IgM level by 12 weeks, some patients
The safety of nucleoside analogs has been the may flare for months despite having tumor responses in
146 subject of investigation in several recent studies. their bone marrow. Patients with baseline serum IgM
Chapter 9: Waldenström’s macroglobulinemia/lymphoplasmacytic lymphoma
investigated in combination with rituximab, and these comparison to patients treated with CHOP alone.
studies are discussed below. Dimopoulos et al. investigated the combination of rit-
uximab, dexamethasone, and oral cyclophosphamide
Bendamustine (RCD) as primary therapy in 72 patients with WM. At
Bendamustine is a recently approved agent for the least a major response was observed in 74% of patients
treatment of relapsed/refractory indolent non- in this study, and the 2-year PFS was 67%. Therapy was
Hodgkin’s lymphoma (NHL). Bendamustine has well tolerated, though one patient died of interstitial
structural similarities to both alkylating agents and pneumonia. In the salvage setting, the use of CHOP-R
purine analogs. Bendamustine in combination with has been investigated in relapsed/refractory WM
rituximab has been investigated in both previously patients. Among 13 evaluable patients, 10 patients
untreated and relapsed/refractory WM patients and achieved a major response (77%), including three com-
is discussed below. plete response (CR) and seven partial response (PR),
and two patients achieved a minor response. In a retro-
Everolimus spective study, Ioakimidis et al. examined the outcomes
Everolimus is an oral inhibitor of the mTOR pathway, of symptomatic WM patients who received CHOP-R,
which is approved for the treatment of renal cell car- cyclophosphamide, vincristine, prednisone, rituximab
cinoma. The Akt-mTOR-p70 pathway is active in (CVP-R), or cyclophosphamide, prednisone, rituximab
WM, and inhibition of this pathway leads to apoptosis (CP-R). Baseline characteristics for all three cohorts
of primary WM cells, and WM cell lines. Fifty patients were similar for age, prior therapies, bone marrow
with a median of three prior therapies were treated involvement, hematocrit, platelet count, and serum
with everolimus in a joint Dana Farber/Mayo Clinic beta-2 microglobulin, though serum IgM levels were
study. The overall response rate was 70%, with 42% of higher in patients treated with CHOP-R. The overall
patients attaining a major response. The PFS at 12 response rates to therapy were comparable among all
months was estimated to be 62%. Grade 3 or higher three treatment groups: CHOP-R (96%); CVP-R (88%)
related toxicities were observed in 56% of patients, and CP-R (95%), though more CR were observed
with cytopenias constituting the most common tox- among patients treated with either CVP-R or CHOP-
icity. Pulmonary toxicity occurred in 10% of patients. R. Comparison of adverse events for these regimens
Dose reductions due to toxicity occurred in 52% of showed a higher incidence for neutropenic fever as
patients. well as treatment-related neuropathy in patients receiv-
A clinical trial examining the activity of everoli- ing CHOP-R and CVP-R versus CP-R. These results
mus in previously untreated patients with WM has suggest that, in WM, the use of doxorubicin and vin-
been initiated by the WMCTG. IgM discordance to cristine may be omitted to minimize treatment-related
bone marrow tumor burden was common in this complications.
upfront study, and therefore serial bone marrow Combination therapy with nucleoside analogs has
biopsies are important in assessing disease response been investigated as both first line and salvage ther-
to everolimus. apy in WM. Laszlo et al. evaluated the combination of
subcutaneous cladribine with rituximab in 29 WM
patients with either untreated or previously treated
Combination strategies disease. Intended therapy consisted of rituximab on
Because rituximab is an active and a non- day 1 followed by subcutaneous cladribine 0.1 mg/kg
myelosuppressive agent, its combination with various for 5 consecutive days, administered monthly for
chemotherapeutic agents has been extensively explored four cycles. With a median follow-up of 43 months,
in WM. The combination of CHOP (cyclophospha- the overall response rate observed was 89.6%, with
mide, doxorubicin, vincristine, prednisone) with ritux- seven CR, 16 PR, and three minor responses.
imab (CHOP-R) was investigated in a randomized Response activity was similar between untreated
frontline study by the German Low Grade Lymphoma and previously treated patients. No major infec-
Study Group (GLSG) involving 69 patients, most of tions were observed despite the lack of antimicrobial
whom had WM. The addition of rituximab to CHOP prophylaxis.
resulted in a higher overall response rate (94% versus In a study by the WMCTG, the combination of
148 67%) and median TTP (63 versus 22 months) in rituximab and fludarabine was administered to 43
Chapter 9: Waldenström’s macroglobulinemia/lymphoplasmacytic lymphoma
WM patients, 32 (75%) of whom were previously which appears to ameliorate toxicity while maintain-
untreated. The overall response rate was 95.3%, and ing responses similar to intravenous administration.
83% of patients achieved a major response. The The combination of immunomodulator agents (tha-
median TTP was 51.2 months in this series, and was lidomide, lenalidomide) with rituximab was investi-
longer for those patients who were previously gated by the WMCTG. Thalidomide was administered
untreated and for those achieving at least a very good at 200 mg daily for 2 weeks, followed by 400 mg daily
PR. Hematological toxicity was common, particularly and thereafter for 1 year. Patients received four weekly
neutropenia and thrombocytopenia. Two deaths infusions of rituximab at 375 mg/m2, beginning 1 week
occurred in this study from non-Pneumocystis carinii after initiation of thalidomide, followed by four addi-
pneumonia. Secondary malignancies, including trans- tional weekly infusions of rituximab at 375 mg/m2
formation to aggressive lymphoma and development beginning at week 13. The overall and major response
of myelodysplasia or AML, were observed in six rate was 72% and 64%, respectively, and the median
patients in this series. TTP was 38 months in this series. Dose reduction and/
Tedeschi et al. reported a multicenter study with or discontinuation of thalidomide was common, and
fludarabine, cyclophosphamide, and rituximab (FCR) mainly attributed to treatment-related neuropathy. The
in symptomatic WM patients with untreated or investigators concluded in this study that lower doses of
relapsed/refractory disease to one line of chemother- thalidomide (i.e. 50–100 mg/day) should be considered
apy. Treatment consisted of rituximab at 375 mg/m2 in this patient population. The combination of lenali-
on day 1, fludarabine at 25 mg/m2, and cyclophospha- domide with rituximab was investigated by the
mide at 250 mg/m2 by intravenous administration on WMCTG using lenalidomide at 25 mg daily on a syn-
days 2–4 every 4 weeks. Forty-three patients were copated schedule wherein therapy was administered for
accrued to this study. The overall response rate was 3 weeks, followed by a 1-week pause for an intended
89%, with 83% of patients attaining a major remission, duration of 48 weeks. Patients received 1 week of ther-
and 14% a CR. Prolonged neutropenia was observed in apy with lenalidomide, after which rituximab (375 mg/
up to a third of patients. With a median follow-up of m2) was administered weekly on weeks 2–5, then 13–16.
15 months, the median PFS for this study has not been The overall and major response rates in this study were
reached. 50% and 25%, respectively, and a median TTP for
The combination of bortezomib, dexamethasone, responders was 18.9 months. However, an acute
and rituximab (BDR) has been investigated as primary decrease in hematocrit was observed during the first 2
therapy in patients with WM by the WMCTG. An weeks of lenalidomide therapy in 13/16 (81%) patients
overall response rate of 96%, major response rate of with a median absolute decrease in hematocrit of 4.8%,
83%, and CR in 22% was observed with BDR. The resulting in anemia-related complications and hospital-
updated median PFS in this study was >56.1 months. izations. Despite dose reduction, most patients in this
The incidence of grade 3 neuropathy was 30% in this study continued to demonstrate aggravated anemia with
study, which utilized a twice-a-week schedule for bor- lenalidomide. The mechanism for lenalidomide-related
tezomib administration at 1.3 mg/m2. Peripheral neu- anemia in WM remains to be determined, and the use of
ropathy from bortezomib was reversible in most this agent among WM patients should be avoided.
patients in this study following discontinuation of The use of bendamustine in combination with rit-
therapy, and patients benefitted with pregabalin. An uximab was explored by Rummel et al. in the frontline
increased incidence of herpes zoster was also observed, therapy of WM. As part of a randomized study, patients
with BDR prompting the use of prophylactic antiviral received six cycles of bendamustine plus rituximab
therapy. An alternative schedule for bortezomib (Benda-R) or CHOP-R. A total of 546 patients were
administration (i.e. weekly at 1.6 mg/m2) in combina- enrolled in this study for patients with indolent NHL,
tion with rituximab and/or dexamethasone has been and included 41 patients with WM. For patients receiv-
investigated in several studies, with overall response ing Benda-R, bendamustine was administered at 90 mg/
rates of 80–90%. A lower incidence of peripheral neu- m2 on days 1 and 2, and rituximab at 375 mg/m2 on day
ropathy was observed in two studies using bortezomib 1. The overall response rate was 96% for Benda-R-, and
once a week. The impact of once- versus twice-a-week 94% for CHOP-R-treated patients. With a median
bortezomib administration on PFS remains to be clari- observation period of 45 months, the median PFS was
fied, as does the use of subcutaneous bortezomib, 69.5 months for Benda-R versus 28.1 months for 149
Chapter 9: Waldenström’s macroglobulinemia/lymphoplasmacytic lymphoma
CHOP-R. Importantly, Benda-R was associated with a autologous SCT had a positive impact on PFS. When
lower incidence of grade 3 or 4 neutropenia, infectious used as consolidation at first response, autologous
complications, and alopecia. In the salvage setting, the transplantation provided a PFS of 44% at 5 years.
outcome of 30 WM patients with relapsed/refractory In the allogeneic SCT experience from the EBMT,
disease who received bendamustine alone, or with a the long-term outcome of 86 WM patients was reported
CD20-directed antibody, was reported by Treon et al. by Kyriakou. A total of 86 patients received allograft
An overall response rate of 83.3% was reported. The by either myeloablative (n = 37) or reduced-intensity
median estimated PFS for all patients was 13.2 months. (n = 49) conditioning. The median age of patients in
Overall therapy was well tolerated. Prolonged myelo- this series was 49 years, and 47 patients had three or
suppression was more common in patients who more previous lines of therapy. Eight patients failed
received prior nucleoside analogs. prior autologous SCT. Fifty-nine patients (68.6%) had
chemotherapy-sensitive disease at the time of allogeneic
SCT. Non-relapse mortality at 3 years was 33% for
Maintenance therapy patients receiving a myeloablative transplant, and 23%
A role for maintenance rituximab in WM patients for those who received reduced-intensity conditioning.
following response to a rituximab-containing regimen The overall response rate was 75.6%. The relapse rates at
was raised in a study examining the outcome of 248 3 years were 11% for myeloablative, and 25% for
WM rituximab-naïve patients who were either reduced-intensity conditioning recipients. Five-year
observed or received maintenance rituximab. In this PFS and OS for WM patients who received a myelo-
retrospective study, categorical responses improved in ablative allogeneic SCT were 56% and 62%, and for
16/162 (10%) observed patients, and in 36/86 (42%) patients who received reduced-intensity conditioning
patients who received maintenance rituximab follow- were 49% and 64%, respectively. The occurrence of
ing induction therapy. Both PFS (56.3 versus 28.6 chronic graft-versus-host disease was associated with
months) and OS (>120 versus 116 months) were lon- improved PFS, and suggested the existence of a clini-
ger in patients who received maintenance rituximab. cally relevant graft-versus-WM effect in this study.
Improved PFS was evident despite previous treatment
status, induction with rituximab alone, or in combi-
nation therapy. Best serum IgM response was lower, Response criteria in Waldenström’s
and hematocrit was higher in those patients receiving macroglobulinemia
maintenance rituximab. Among patients receiving As part of the Second and Third International
maintenance rituximab, an increased number of infec- Workshops on WM, consensus panels developed
tious events, predominantly sinusitis and bronchitis, guidelines for uniform response criteria in WM. The
were observed, though these were mainly grade 1 or 2. category of minor response was adopted at the Third
International Workshop of WM, given that clinically
High-dose therapy and stem cell meaningful responses were observed with newer bio-
logical agents, and this category is based on ≥25 to
transplantation < 50% decrease in serum IgM level, which is used as a
The use of stem cell transplantation (SCT) therapy has surrogate marker of disease in WM. At the Sixth
also been explored in patients with WM. Kyriakou et al. International Workshop on WM, the categorical
reported on the outcome of WM patients in the response of very good partial response (VGPR), i.e.
European Bone Marrow Transplant (EBMT) registry 90% reduction in IgM levels, was adopted given
who received either an autologous or allogeneic SCT. reports of improved clinical outcome associated with
Among 158 patients receiving an autologous SCT, VGPR or better response achievement. In distinction,
which included primarily relapsed or refractory the term major response is used to denote a response
patients, the 5-year PFS and OS rates were 39.7% and of ≥50% in serum IgM levels, and includes partial or
68.5%, respectively. Non-relapse mortality at 1 year was better responses. Response categories and criteria for
3.8%. Chemorefractory disease and the number of prior progressive disease in WM based on consensus rec-
lines of therapy at the time of the autologous SCT were ommendations are summarized in Table 9.4.
the most important prognostic factors for PFS and OS. An important concern with the use of IgM as a
150 The achievement of a negative immunofixation after surrogate marker of disease is that it can fluctuate,
Chapter 9: Waldenström’s macroglobulinemia/lymphoplasmacytic lymphoma
Table 9.4 Summary of updated response criteria adopted at the Sixth International Workshop on Waldenstrom’s macroglobulinemia.
Complete CR IgM in normal range, and disappearance of monoclonal protein by immunofixation; no histological
response evidence of bone marrow involvement, and resolution of any adenopathy/organomegaly (if present
at baseline), along with no signs or symptoms attributable to WM. Reconfirmation of the CR status is
required by repeat immunofixation studies.
Very good partial VGPR A >90% reduction of serum IgM and decrease in adenopathy/organomegaly (if present at baseline) on
response physical examination or on CT scan. No new symptoms or signs of active disease.
Partial response PR A >50% reduction of serum IgM and decrease in adenopathy/organomegaly (if present at baseline) on
physical examination or on CT scan. No new symptoms or signs of active disease.
Minor response MR A >25% but <50% reduction of serum IgM. No new symptoms or signs of active disease.
Stable disease SD A <25% reduction and <25% increase of serum IgM without progression of adenopathy/organo-
megaly, cytopenias, or clinically significant symptoms due to disease and/or signs of WM.
Progressive PD A >25% increase in serum IgM by protein confirmed by a second measurement or progression of
disease clinically significant findings due to disease (i.e. anemia, thrombocytopenia, leukopenia, bulky aden-
opathy/organomegaly) or symptoms (unexplained recurrent fever >38.4°C, drenching night sweats,
>10% body weight loss, or hyperviscosity, neuropathy, symptomatic cryoglobulinemia or amyloidosis)
attributable to WM.
independent of tumor cell killing, particularly with lymphocytic leukemia (CLL) with Richter’s syndrome but
newer biologically targeted agents such as rituximab, not in common CLL. Blood, 1995;85:1913–1919.
bortezomib, and everolimus. Rituximab induces a Buske C, Hoster E, Dreyling MH, et al. The addition of
spike or flare in serum IgM levels, which can occur rituximab to front-line therapy with CHOP (R-CHOP)
when used as monotherapy or in combination with results in a higher response rate and longer time to
other agents, including cyclophosphamide, nucleoside treatment failure in patients with lymphoplasmacytic
lymphoma: results of a randomized trial of the German
analogs, thalidomide, and lenalidomide, and last for Low-Grade Lymphoma Study Group (GLSG). Leukemia,
several weeks to months. Bortezomib and everolimus 2009;23:153–161.
can suppress IgM levels independent of tumor cell
Chang H, Qi C, Trieu Y, et al. Prognostic relevance of 6q
killing in certain patients. Moreover, in patients treat- deletion in Waldenstrom’s macroglobulinemia. Clin
ed with selective B-cell-depleting agents such as ritux- Lymph Myeloma, 2009;9:36–38.
imab and alemtuzumab, residual IgM-producing Chassande B, Leger JM, Younes-Chennoufi AB, et al.
plasma cells are spared and continue to persist, thus Peripheral neuropathy associated with IgM monoclonal
potentially skewing the relative response and assess- gammopathy: correlations between M-protein antibody
ment to treatment. Therefore, in circumstances where activity and clinical/electrophysiological features in 40
the serum IgM levels appear out of context with the cases. Muscle Nerve, 1998;21:55–62.
clinical progress of the patient, a bone marrow biopsy Chen CI, Kouroukis CT, White D, et al. Bortezomib is active
should be considered to clarify the patient’s underlying in patients with untreated or relapsed Waldenstrom’s
disease burden. Soluble CD27 may serve as an alter- macroglobulinemia: a phase II study of the National
native surrogate marker in WM, and remains a faithful Cancer Institute of Canada Clinical Trials Group. J Clin
marker of disease in patients experiencing a Oncol, 2007;25:1570–1575.
rituximab-related IgM flare, as well as plasmapheresis. Cheson BD, Rummel MJ. Bendamustine: rebirth of an old
drug. J Clin Oncol, 2009;27:1492–1501.
Ciccarelli BT, Yang G, Hatjiharissi E, et al. Soluble CD27
Further reading is a faithful marker of disease burden and is
Anagnostopoulos A, Zervas K, Kyrtsonis M, et al. Prognostic unaffected by the rituximab induced IgM flare, as
value of serum beta 2-microglobulin in patients with well as plasmapheresis in patients with
Waldenstrom’s macroglobulinemia requiring therapy. Waldenstrom’s macroglobulinemia. Clin Lymphoma
Clin Lymph Myeloma, 2006;7:205–209. Myeloma, 2009;9:56–58.
Aoki H, Takishita M, Kosaka M, Saito S. Frequent somatic Daoud MS, Lust JA, Kyle RA, Pittelkow MR. Monoclonal
mutations in D and/or JH segments of Ig gene in gammopathies and associated skin disorders. J Am Acad
Waldenström’s macroglobulinemia and chronic Dermatol, 1999;40:507–535. 151
Chapter 9: Waldenström’s macroglobulinemia/lymphoplasmacytic lymphoma
Dhodapkar MV, Jacobson JL, Gertz MA, et al. Prognostic rituximab in untreated patients with Waldenström
factors and response to fludarabine therapy in patients macroglobulinemia. Am J Hematol, 2010;85:670–674.
with Waldenström macroglobulinemia: results of United Gobbi PG, Bettini R, Montecucco C, et al. Study of prognosis
States intergroup trial (Southwest Oncology Group in Waldenström’s macroglobulinemia: a proposal for a
S9003). Blood, 2001;98:41–48. simple binary classification with clinical and
Dimopoulos MA, Alexanian R. Waldenstrom’s investigational utility. Blood, 1994;83:2939–2945.
macroglobulinemia. Blood, 1994;83:1452–1459. Gordon PH, Rowland LP, Younger DS, et al.
Dimopoulos MA, Kantarjian H, Weber D, et al. Primary Lymphoproliferative disorders and motor neuron
therapy of Waldenström’s macroglobulinemia with 2- disease: an update. Neurology, 1997;48:1671–1678.
chlorodeoxyadenosine. J Clin Oncol, 1994;12:2694–2698. Hanzis C, Ojha RP, Hunter Z, et al. Associated malignancies
Dimopoulos MA, Zomas A, Viniou NA, et al. Treatment of in patients with Waldenström’s macroglobulinemia and
Waldenström’s macroglobulinemia with thalidomide. their kin. Clin Lymphoma Myeloma Leuk, 2011;11:88–92.
J Clin Oncol, 2001;19:3596–3601. Hellmann A, Lewandowski K, Zaucha JM, et al. Effect of a 2-
Dimopoulos MA, Zervas C, Zomas A, et al. Treatment of hour infusion of 2-chlorodeoxyadenosine in the
Waldenstrom’s macroglobulinemia with rituximab. treatment of refractory or previously untreated
J Clin Oncol 2002;20:2327–2333. Waldenström’s macroglobulinemia. Eur J Haematol,
Dimopoulos M, Gika D, Zervas K, et al. The international 1999;63:35–41.
staging system for multiple myeloma is applicable in Ho A, Leleu X, Hatjiharissi E, et al. CD27-CD70 interactions
symptomatic Waldenstrom’s macroglobulinemia. Leuk in the pathogenesis of Waldenstrom’s
Lymph, 2004;45:1809–1813. macroglobulinemia. Blood, 2008;112:4683–4689.
Dimopoulos MA, Anagnostopoulos A, Kyrtsonis MC, et al. Ioakimidis L, Patterson CJ, Hunter ZR, et al. Comparative
Treatment of relapsed or refractory Waldenstrom’s outcomes following CP-R, CVP-R and CHOP-R in
macroglobulinemia with bortezomib. Haematologica, Waldenstrom’s macroglobulinemia. Clin Lymphoma
2005;90:1655–1657. Myeloma, 2009;9:62–66.
Dimopoulos MA, Anagnostopoulos A, Zervas C, et al. Kimby E, Treon SP, Anagnostopoulos A, et al. Update on
Predictive factors for response to rituximab in recommendations for assessing response from the Third
Waldenstrom’s macroglobulinemia. Clin Lymphoma, International Workshop on Waldenstrom’s
2005;5:270–272. macroglobulinemia. Clin Lymphoma Myeloma,
Dimopoulos MA, Anagnostopoulos A, Kyrtsonis MC, et al. 2006;6:380–383.
Primary treatment of Waldenstrom’s macroglobulinemia Kyle RA, Greipp PR, Gertz MA, et al. Waldenström’s
with dexamethasone, rituximab and cyclophosphamide. J macroglobulinaemia: a prospective study comparing
Clin Oncol, 2007;25:3344–3349. daily with intermittent oral chlorambucil. Br J Haematol,
Dimopoulos MA, Gertz MA, Kastritis E, et al. Update on 2000;108:737–742.
treatment recommendations from the Fourth Kyle RA, Treon SP, Alexanian R, et al. Prognostic markers
International Workshop on Waldenstrom’s and criteria to initiate therapy in Waldenström’s
Macroglobulinemia. J Clin Oncol, 2009;27:120–126. macroglobulinemia: Consensus Panel Recommendations
Eurelings M, Ang CW, Notermans NC, et al. Antiganglioside from the Second International Workshop on
antibodies in polyneuropathy associated with Waldenström’s macroglobulinemia. Semin Oncol,
monoclonal gammopathy. Neurology, 2003;30:116–120.
2001;57:1909–1912.
Kyriakou C, Canals C, Cornelissen JJ, et al. Allogeneic stem-
Furman RR, Eradat H, Switzky JC, et al. A phase II trial of cell transplantation in patients with Waldenström
ofatumumab in subjects with Waldenstrom’s macroglobulinemia: report from the Lymphoma
macroglobulinemia. Blood, 116:Abstract 1795. Working Party of the European Group for Blood and
Gad A, Willen R, Carlen B, Gyland F, Wickander M. Marrow Transplantation. J Clin Oncol,
Duodenal involvement in Waldenström’s 2010;28:4926–4934.
macroglobulinemia. J Clin Gastroenterol, Kyriakou C, Canals C, Sibon D, et al. High-dose therapy and
1995;20:174–176. autologous stem-cell transplantation in Waldenstrom
Ghobrial I, Gertz M, LaPlant B, et al. Phase II trial of the oral macroglobulinemia: the Lymphoma Working Party of
mammalian target of rapamycin inhibitor everolimus in the European Group for Blood and Marrow
relapsed or refractory Waldenström macroglobulinemia. Transplantation. J Clin Oncol, 2010;28:2227–2732.
J Clin Oncol, 2010;28:1408–1414. Laszlo D, Andreola G, Rigacci L, et al. Rituximab and
152 Ghobrial IM, Xie W, Padmanabhan S, Badros A, et al. Phase subcutaneous 2-chloro-2’-deoxyadenosine combination
II trial of weekly bortezomib in combination with treatment for patients with Waldenstrom
Chapter 9: Waldenström’s macroglobulinemia/lymphoplasmacytic lymphoma
macroglobulinemia: clinical and biologic results of a Ogmundsdottir HM, Sveinsdottir S, Sigfusson A, et al.
phase II multicenter study. J Clin Oncol, Enhanced B cell survival in familial macroglobulinaemia
2010;28:2233–2238. is associated with increased expression of Bcl-2. Clin Exp
Leblond V, Ben Othman T, Deconinck E, et al. Activity of Immunol, 1999;117:252–260.
fludarabine in previously treated Waldenström’s Ogmundsdottir HM, Steingrimsdottir H, Haraldsdottir V.
macroglobulinemia: a report of 71 cases. Groupe Familial paraproteinemia: hyper-responsive B-cells as
Cooperatif Macroglobulinemie. J Clin Oncol, endophenotype. Clin Lymphoma Myeloma Leukemia,
1998;16:2060–2064. 2011;11:82–84.
Leleu X, Jia X, Runnels J, et al. The Akt pathway regulates Owen RG, Treon SP, Al-Katib A, et al. Clinicopathological
survival and homing in Waldenstrom definition of Waldenström’s macroglobulinemia:
macroglobulinemia. Blood, 2007;110:4417–4426. Consensus Panel Recommendations from the Second
Leleu X, O’Connor K, Ho A, et al. Hepatitis C viral infection International Workshop on Waldenström’s
is not associated with Waldenstrom’s macroglobulinemia. Semin Oncol, 2003;30:110–115.
macroglobulinemia. Am J Hematol, 2007;82:83–84. Rummel MJ, Niederle N, Maschmeyer GA, et al.
Leleu XP, Manning R, Soumerai JD, et al. Increased Bendamustine plus rituximab versus CHOP plus
incidence of transformation and myelodysplasia/acute rituximab as firstline treatment of patients with
leukemia in patients with Waldenström indolent and mantle cell lymphomas: updated
macroglobulinemia treated with nucleoside analogs. results from the StiL NHL1 study. Proc ASCO, 2012;
J Clin Oncol, 2009;27:250–255. abstract #3.
Leleu X, Tamburini J, Roccaro A, et al. Balancing risk versus Sahota SS, Forconi F, Ottensmeier CH, et al. Typical
benefit in the treatment of Waldenstrom’s Waldenström macroglobulinemia is derived from a
macroglobulinemia patients with nucleoside analogue B-cell arrested after cessation of somatic mutation but
based therapy. Clin Lymph Myeloma, 2009;in press. prior to isotype switch events. Blood,
2002;100:1505–1507.
Levy Y, Fermand JP, Navarro S, et al. Interleukin 6
dependence of spontaneous in vitro differentiation of B Schop RF, Kuehl WM, Van Wier SA, et al. Waldenström
cells from patients with IgM gammopathy. Proc Natl macroglobulinemia neoplastic cells lack immunoglobulin
Acad Sci USA, 1990;87:3309–3313. heavy chain locus translocations but have frequent 6q
deletions. Blood, 2002;100:2996–3001.
Liu ES, Burian C, Miller WE, Saven A. Bolus administration
of cladribine in the treatment of Waldenström Shiokawa S, Suehiro Y, Uike N, Muta K, Nishimura J.
macroglobulinaemia. Br J Haematol, 1998;103:690–695. Sequence and expression analyses of mu and delta
transcripts in patients with Waldenström’s
Merlini G, Baldini L, Broglia C, et al. Prognostic factors in macroglobulinemia. Am J Hematol, 2001;68:139–143.
symptomatic Waldenström’s macroglobulinemia. Semin
Oncol, 2003;30:211–215. Silvestri F, Barillari G, Fanin R, et al. Risk of hepatitis C virus
infection, Waldenström’s macroglobulinemia, and
Morel P, Monconduit M, Jacomy D, et al. Prognostic factors monoclonal gammopathies. Blood, 1996;88:1125–1126.
in Waldenström macroglobulinemia: a report on 232
patients with the description of a new scoring system and Singh A, Eckardt KU, Zimmermann A, et al.
its validation on 253 other patients. Blood, Increased plasma viscosity as a reason for
2000;96:852–858. inappropriate erythropoietin formation. J Clin Invest,
1993;91:251–256.
Morel P, Duhamel A, Gobbi P, et al. International prognostic
scoring system for Waldenstrom macroglobulinemia. Strauss SJ, Maharaj L, Hoare S, et al. Bortezomib therapy in
Blood, 2009;113:4163–4170. patients with relapsed or refractory lymphoma: potential
correlation of in vitro sensitivity and tumor necrosis
Ngo HT, Leleu X, Lee J, Jia X, et al. SDF-1/CXCR4 and VLA-
4 interaction regulates homing in Waldenstrom factor alpha response with clinical activity. J Clin Oncol,
macroglobulinemia. Blood, 2008;112:150–158. 2006;24:2105–2112.
Nguyen-Khac F, Lejeune J, Chapiro E, et al. Cytogenetic Tedeschi A, Benevolo G, Varettoni M, et al. Results of a
abnormalities in a cohort of 171 patients with phase II multicenter study of immunochemotherapy with
Waldenström macroglobulinemia before treatment: fludarabine, cyclophosphamide and rituximab (FCR) for
clinical and biological correlations. Blood, 2010;116: symptomatic Waldenstrom’s macroglobulinemia. Blood,
Abstract 801. 2008;112:Abstract 3692.
Ocio EM, Schop RF, Gonzalez B, et al. 6q deletion in Tournilhac O, Santos DD, Xu L, et al. Mast cells in
Waldenstrom’s macroglobulinemia is associated with Waldenstrom’s macroglobulinemia support
features of adverse prognosis. Br J Haematol, lymphoplasmacytic cell growth through CD154/CD40
2007;136:80–86. signaling. Ann Oncol, 2006;17:1275–1282. 153
Chapter 9: Waldenström’s macroglobulinemia/lymphoplasmacytic lymphoma
Treon SP. How I treat Waldenstrom’s macroglobulinemia. in rituximab naïve patients with Waldenstrom’s
Blood, 2009;114:419–431. macroglobulinemia who respond to a rituximab
Treon SP, Emmanouilides C, Kimby E, et al. Extended containing regimen. Br J Haematol, 2011; May [Epub
rituximab therapy in Waldenström’s macroglobulinemia. ahead of print].
Ann Oncol, 2005;16:132–138. Treon SP, Merlini G, Morra E, et al. Report from the Sixth
Treon SP, Hansen M, Branagan AR, et al. Polymorphisms in International Workshop on Waldenstrom’s
FcγRIIIA (CD16) receptor expression are associated with Macroglobulinemia. Clin Lymph Myeloma Leukemia,
clinical responses to rituximab in Waldenstrom’s 2011;11:69–73.
macroglobulinemia. J Clin Oncol, 2005;23:474–481. Treon SP, Xu L, Yang G, et al. MYD88 L265P somatic
Treon SP, Hunter Z, Branagan A. CHOP plus rituximab mutation in Waldenstrom macroglobulinemia. N Engl
therapy in Waldenström’s macroglobulinemia. Clin J Med, 2012;367:826–833.
Lymphoma Myeloma, 2005;5:273–277. Varettoni M, Tedesci A, Arcaini L, et al. Risk of second
Treon SP, Hunter ZR, Matous J, et al. Multicenter clinical cancers in Waldenstrom macroglobulinemia. Ann Oncol,
trial of bortezomib in relapsed/refractory Waldenstrom’s 2011; [Epub ahead of print].
macroglobulinemia: results of WMCTG Trial 03–248. Vital A. Paraproteinemic neuropathies. Brain Pathol,
Clin Cancer Res, 2007;13:3320–3325. 2001;11:399–407.
Treon SP, Hunter Z, Ciccarelli BT, et al. IgA and IgG Wagner SD, Martinelli V, Luzzatto L. Similar patterns of
hypogammaglobulinemia is a constitutive feature in most V kappa gene usage but different degrees of somatic
Waldenstrom’s macroglobulinemia patients and may be mutation in hairy cell leukemia, prolymphocytic
related to mutations associated with common variable leukemia, Waldenström’s macroglobulinemia, and
immunodeficiency disorder (CVID). Blood, myeloma. Blood, 1994;83:3647–3653.
2008;112:3749. Weber D, Treon SP, Emmanouilides C, et al. Uniform
Treon SP, Soumerai JD, Branagan AR, et al. Lenalidomide response criteria in Waldenstrom’s macroglobulinemia:
and rituximab in Waldenström’s macroglobulinemia. consensus panel recommendations from the Second
Clin Cancer Res, 2008;15:355–360. International Workshop on Waldenstrom’s
Treon SP, Soumerai JD, Branagan AR, et al. Thalidomide macroglobulinemia. Semin Oncol, 2003;30:127–131.
and rituximab in Waldenstrom’s macroglobulinemia. Weiss MD, Dalakas MC, Lauter CJ, Willison HJ, Quarles
Blood, 2008;112:4452–4457. RH. Variability in the binding of anti-MAG and anti-
Treon SP, Branagan AR, Ioakimidis L, et al. Long term SGPG antibodies to target antigens in demyelinating
outcomes to fludarabine and rituximab in Waldenstrom’s neuropathy and IgM paraproteinemia. J Neuroimmunol,
macroglobulinemia. Blood, 2009;Epub ahead of print. 1999;95:174–184.
Treon SP, Ioakimidis L, Soumerai JD, et al. Primary therapy Whittaker SJ, Bhogal BS, Black MM. Acquired
of Waldenstrom’s macroglobulinemia with bortezomib, immunobullous disease: a cutaneous manifestation of
dexamethasone and rituximab: results of WMCTG IgM macroglobulinaemia. Br J Dermatol,
Clinical Trial 05–180. J Clin Oncol, 2009;27:3830–3835. 1996;135:283–286.
Treon SP, Hanzis C, Tripsas C, et al. Bendamustine therapy World Health Organization. Classification of Tumours of
in patients with relapsed or refractory Waldenström’s Haematopoietic and Lymphoid Tissues. In Swerdlow SH,
macroglobulinemia. Clin Lymphoma Myeloma Leuk, Campo E, Harris NL, et al. (eds). Lyon: IARC Press,
2011;11:133–135. 2008.
Treon SP, Yang G, Hanzis C, et al. Attainment of complete/ Yang G, Xu L, Hunter ZR, et al. The rituximab and IVIG
very good partial response following rituximab-based related IgM flare in Waldenstrom’s macroglobulinemia is
therapy is an important determinant to progression-free associated with monocytic activation of FCGR2A
survival, and is impacted by polymorphisms in FCGR3A signaling, and triggering of IL-6 release by the PI3K/AKT
in Waldenstrom macroglobulinaemia. Br J Haematol, and MAPK pathways. Blood, 2010;116:2870.
2011;154;223–228. Zinzani PL, Gherlinzoni F, Bendandi M, et al. Fludarabine
Treon SP, Hanzis, C, Manning RJ, et al. Maintenance treatment in resistant Waldenström’s
rituximab is associated with improved clinical outcome macroglobulinemia. Eur J Haematol, 1995;54:120–123.
154
Chapter
10
Mantle cell lymphoma
Martin Dreyling and Michael E. Williams
Pathology: Andrew Wotherspoon Molecular cytogenetics: Andreas Rosenwald and German Ott
Lymphoma, Second Edition, ed. Robert Marcus, John W. Sweetenham, and Michael E. Williams. Published by 155
Cambridge University Press. © Cambridge University Press 2014.
Chapter 10: Mantle cell lymphoma
Pathology
Designations in the older literature used descriptive designation of “mantle cell lymphoma” was proposed
terms such as “lymphocytic lymphoma of intermediate by the International Lymphoma Study Group in 1992
differentiation” and “mantle zone lymphoma.” The first and further delineated in the WHO Classification.
specific recognition was by Lennert, who described “dif- The lymph node architecture is usually effaced by
fuse germinocytoma” in 1973, renamed “centrocytic a diffuse proliferation of lymphoid cells with monot-
lymphoma” in the 1974 Kiel classification. As immuno- onous appearance (Figure 10.1). In some cases there
phenotypic characterization became more refined and, may be a more nodular growth pattern, or the cells
importantly, the strong correlation with the t(11;14) may be seen surrounding residual reactive germinal
156 (q13;q32) chromosomal translocation permitted differ- centers in a mantle-zone pattern. In classical cases
entiation from other small cell lymphomas, the current (80–90%) the cells are small to intermediate size
Chapter 10: Mantle cell lymphoma
with scanty cytoplasm and with nuclei that show lack this cytogenetic abnormality; even more rarely,
variable irregularity. The nuclear chromatin is cyclin D2 or D3 is expressed rather than D1. Such cyclin
granular and cells may have small, rather indistinct D1-negative cases, confirmed by gene expression profil-
eosinophilic nucleoli. Rarely the cells may resemble ing, may present a considerable diagnostic challenge.
small lymphocytes with round nuclei or have a Nuclear expression of the neuronal transcription factor
more monocytoid appearance. Large cells resembling Sox11 is present in up to 90% of cases of MCL, includ-
centroblasts or immunoblasts are not a feature. ing in some cases that are cyclin D1-negative. Although
The background frequently contains hyalinized expression of Sox11 is not specific to MCL, as it can be
small blood vessels and may contain histiocytes. found in a proportion of hairy cell leukemias, lympho-
Proliferation centers are absent. Plasma cells may be blastic and Burkitt lymphomas, it may prove a useful
seen in the background but are reactive rather than diagnostic tool. Cytoplasmic rather than nuclear
part of the neoplastic population. expression of SOX11 expression has been suggested to
In a proportion of cases the morphology is not be associated with a worse outcome in nodal MCL,
typical. In the blastoid variant the cells resemble lym- while lack of Sox 11 has been associated with leukemic
phoblasts with scanty cytoplasm and fine nuclear chro- presentation of MCL associated with a more indolent
matin associated with a striking mitotic rate. The course. Proliferation marker expression is variable, but
pleomorphic variant is composed of more pleomorphic a high proliferative index has negative prognostic impli-
cells, which are larger than the typical cells of MCL and cations. Staining for follicular dendritic cells usually
have large nuclei with irregular outlines and prominent demonstrates an expanded and disrupted meshwork
nucleoli. These morphological variants are associated most obvious in cases with more nodular growth, but
with more aggressive disease. usually present even in those with a diffuse pattern. In
Immunophenotypically the cells have surface IgM contrast to other B-cell lymphomas, MCL is more fre-
and IgD. They are positive for CD20 and CD79a, and quently associated with lambda rather than kappa light
typically coexpress CD5, CD43, and FMC7. Although chain expression.
up to 94% of cases express CD43, as many as 20% of Detailed examination of reactive appearing lymph
cases in some series are reported to be CD5-negative. nodes has revealed the infrequent appearance of
MCL are negative for CD10 and BCL-2 and are usually small numbers of cyclin D1-positive cells. These
CD23-negative. The cells are positive for BCL-2 cases are exceptionally rare and are characterized by
protein. There is no staining for MUM1/IRF4. the presence of small numbers of cyclin D1-positive
Staining for nuclear cyclin D1 is characteristically pos- cells in a proportion of the follicles of the affected
itive, reflecting the underlying t(11;14)(q13;q32) that is nodes with a characteristic localization in the inner
characteristic of this lymphoma (Figure 10.2). Cyclin part of the mantle zone adjacent to the germinal
D1 positivity is also present in rare cases that appear to center. These cells contain the classical t(11;14)
which can be demonstrated by fluorescent in situ
hybridization (FISH) analysis. The significance of
the finding of minimal infiltration by mantle cell
lymphoma type cells, currently termed in-situ mantle
cell lymphoma, remains uncertain.
Molecular pathogenesis
The molecular pathogenesis of MCL is a paradigm of
dysregulated control of the cell cycle machinery and the
response to DNA damage in human cancer. Most, if not
all, molecular and cytogenetic alterations described to
date in MCL appear to affect these two crucial cellular
pathways. The hallmark chromosomal translocation,
the t(11;14)(q13;q32), can be detected in the vast major-
ity of MCL cases and results in the overexpression of
cyclin D1, which plays an important role in the regu-
Figure 10.2 Mantle cell lymphoma (lymph node). Immunostaining
lation of the G1-/S-phase transition of the cell cycle. The 157
for cyclin D1 gives a characteristic nuclear pattern.
Chapter 10: Mantle cell lymphoma
juxtaposition of the cyclin D1 chromosomal locus on pathway. In particular, the chromosomal region
11q13, designated BCL-1 (B-cell lymphoma/leukemia 11q22–23 is frequently deleted in MCL, affecting the
1), to the enhancer of the IgH joining region at 14q32 ataxia-telangiectasia mutated (ATM) gene. In addition
leads to upregulation of this D-type cyclin that is usually to hemizygous ATM deletions, many MCL cases show
not expressed in B-cells. Cyclin D1 forms a complex concomitant mutations of the ATM gene. ATM encodes
with the cyclin-dependent kinases-4 (CDK4) and -6 for a kinase that belongs to the PI-3 kinase-related
(CDK6), and overexpression of cyclin D1 in MCL superfamily and plays a pivotal role in the cellular
leads to increased levels of cyclin D1/CDK complexes response to DNA damage. ATM mutations frequently
which in turn phosphorylate the retinoblastoma protein affect the kinase domain of the gene or result in a
(Rb) and thus directly facilitate cell cycle progression. In truncated version of the ATM protein. Besides frequent
addition, increased levels of cyclin D1/CDK complexes ATM alterations, other molecules involved in the DNA
may sequester the CDK inhibitors p27kip1 and p21 that damage response and specifically targets of the ATM
are usually bound to cyclin E/CDK2 complexes which, kinase itself are also occasionally altered in MCL.
once activated, promote S-phase entry. These studies Specifically, rare alterations of the kinases CHK2 and
suggest that aberrant expression of cyclin D1 plays a CHK1 are observed in MCL. CHK2 stabilizes and acti-
crucial role in the pathogenesis of MCL; moreover, the vates p53, which plays a central role in the response to
level of cyclin D1 expression is directly correlated with DNA damage by initiating cell cycle arrest, apoptosis,
the proliferation rate of the tumor cells and therefore or DNA repair. Decreased protein levels and occasional
with the clinical aggressiveness of this lymphoma. CHK2 mutations present in a subset of MCL
Besides the deregulation of cyclin D1 as a result of may compromise p53 function and are associated
the translocation t(11;14), other genetic alterations with a high number of chromosomal imbalances.
have been reported that disrupt proper function of Importantly, p53 is inactivated in approximately 30%
the cell cycle. Specifically, a significant proportion of of MCL cases with blastoid morphology and high pro-
patients with MCL harbor hemizygous or homozy- liferation rates, while this event is uncommon in MCL
gous deletions in the chromosomal locus 9p21 (as cases with classic morphology and low proliferative
detected by conventional karyotyping or FISH), activity.
affecting the CDK inhibitor p16INK4a. p16INK4a serves Several studies suggest that the proliferation rate of
as an inhibitor of CDK4 and CDK6 and thus helps to the tumor cells is associated with the clinical aggres-
maintain the Rb protein in its dephosphorylated, siveness of the lymphoma. In particular, a low prolif-
antiproliferative state. Functional inactivation of eration rate, e.g. as measured immunohistochemically
this CDK inhibitor may therefore cooperate with by the Ki-67 index, corresponds to a more favorable
aberrant cyclin D1 expression in MCL and enhance clinical course, while a high Ki-67 index is associated
cell cycle progression. It is noteworthy that MCL with short survival times. A gene expression profiling
cases with inactivated p16INK4a usually display study in MCL provided a quantitative measurement of
increased proliferative activity. Rare cases of MCL tumor cell proliferation, termed proliferation signa-
with a wild-type (WT) status of the p16INK4a locus ture, allowing for the definition of prognostic sub-
have been described with overexpression and groups that differ in their median survival by more
genomic amplification of BMI-1 that belongs to the than 5 years. Since the gene expression-based meas-
Polycomb group of genes and acts as a transcriptional urement of proliferation was superior in predicting the
repressor of the p16INK4a locus. BMI-1 overexpres- length of survival than individual molecular param-
sion in a small subset of MCL cases may therefore eters alone or in combination (e.g. level of cyclin D1
represent a pathogenetic alternative to the more fre- expression, p16INK4a alterations), the proliferation sig-
quently observed deletions of its transcriptional tar- nature may be viewed as a quantitative integrator of
get p16INK4a. Finally, a few MCL cases with genomic oncogenic events in MCL, which could be helpful in
amplification and protein overexpression of CDK4 the future to test targeted therapeutic approaches and
have been reported which may represent yet another to guide treatment decisions. More indolent forms of
mechanism of disturbing the G1-/S-phase cell cycle MCL have attracted attention recently, and these cases
checkpoint. appear to harbor fewer genetic alterations compared to
As described above, a second major target of genetic conventional MCL and are characterized by mutated
158 alterations in MCL is the DNA damage response IgVH genes in the majority of cases.
Chapter 10: Mantle cell lymphoma
0.8 10% to less than 30% (≥10), and 30% or more (≥30) Ki-67-
0.7 positive cells.
0.6
0.5
0.4
0.3
< 10, median = 112
0.2 >= 10, median = 59
0.1 >= 30, median = 35
P = .0002
0.0
0 12 24 36 48 60 72 84 96 108 120
numbers of patients at risk months since registration
< 10 38 37 33 30 24 18 13 9 6 2
>= 10 59 55 49 42 32 21 14 9 2 0
>= 30 19 16 11 8 7 3 2 0
B 1.0
0.9
probability of overall survival
0.8
0.7
0.6
0.5
0.4
0.3 < 10, median not reached
>= 10, median not reached
0.2
>= 30, median = 52
0.1 P = .0126
0.0
0 12 24 36 48 60 72 84 96 108 120
numbers of patients at risk months since registration
< 10 38 27 20 11 7 0
>= 10 50 48 40 19 6 3 1 0
>= 30 16 15 11 7 1 0
30 untranslated region of cyclin D1 mRNA, itself a 10.1. The rapid advances in understanding MCL
marker of poorer prognosis in MCL, leads to loss of pathogenesis and prognostic markers may lead to
miR-16–1 binding sites which in turn impairs normal more individualized treatment approaches in the
cell cycle regulation. Taken together, elucidation of years to come.
the complex physiologic dysregulation in MCL via Clinical prognostic factors. The International
gene mutation and miRNA alterations represents an Prognostic Index (IPI) developed for diffuse large
exciting new avenue of investigation that promises B-cell lymphoma has limited predictive value in
improved pathogenic insights, prognostic stratifica- MCL, as the majority of patients present with
tion, and target identification. advanced-stage disease and frequent leukemic or
Activating mutations have recently been identified extranodal sites of involvement. Patients who present
in NOTCH1, a ligand-activated transcription factor with predominantly peripheral blood, bone marrow,
that is also mutated in some cases of chronic lympho- and splenic involvement without significant lymph-
cytic leukemia (CLL) and T-cell acute lymphoblastic adenopathy often experience an indolent clinical
leukemia. MCL patients with these mutations course; these cases need to be carefully distinguished
appeared to have a poorer OS. Additional biomarkers from CLL, splenic lymphoma with villous lympho-
160 correlated with prognosis are summarized in Table cytes (SLVL), and prolymphocytic leukemia.
Chapter 10: Mantle cell lymphoma
The Mantle Cell International Prognostic Index advanced-stage disease the benefit of local radiation
(MIPI) was developed as a risk assessment tool utiliz- therapy is not proven.
ing readily available clinical and laboratory para-
meters: patient age, ECOG (Eastern Cooperative
Oncology Group) performance status, total leukocyte
Conventional chemotherapy
count, and serum lactate dehydrogenase. Based upon In advanced-stage disease, conventional immune-che-
the calculated score, patients with stage III and IV motherapy remains a non-curative approach. Because
disease can be stratified into low-, intermediate- and of the aggressive clinical course in the majority of
high-risk groups that directly correlate with OS patients, a watch and wait strategy generally should
following initial chemotherapy. The MIPI has been only be pursued in selected cases with low-risk profile
validated in other patient cohorts treated with immu- based on clinical manifestation (e.g. GI tract or leuke-
nochemotherapy and dose-intensive regimens, mic disease only), histological features, and/or biology
including stem cell transplantation. A further refine- (Ki-67 <10%). However, those patients presenting a
ment of the MIPI incorporates the proliferation history consistent with slow-paced and low-tumor
marker Ki-67, described above, to provide a MIPI burden advanced-stage disease may be considered for
biologic index (MIPIb) with improved predictive “watchful waiting”.
power for OS – patients with low-risk scores had
median OS rates following induction chemotherapy Monoclonal antibody and combined
of more than 6 years, whereas high-risk patients had
median survivals of only 3 years. immunochemotherapy
Minimal residual disease. The presence of low-level Despite high CD20 expression in MCL, rituximab
minimal residual disease (MRD) involving the blood monotherapy achieves only moderate response rates
and bone marrow can be determined by sensitive PCR of 20–35%. In the largest dataset available thus far
assays for clonal IgVH and t(11;14) breakpoints using from the Swiss SAKK study group, four weekly stand-
patient-specific probes. The achievement of molecular ard doses of rituximab achieved complete and overall
remission following induction therapy as reflected by response rates (ORR) of 2% and 27%, respectively in
MRD-negativity has been associated with prolonged 104 patients with newly diagnosed or relapsed MCL,
clinical remission in prospective randomized clinical resulting in a median event-free survival of 6 months
trials. Ninety percent of patients had a detectable only. Thus antibody monotherapy should be applied
molecular marker, and MRD was assayed serially dur- only in very low-risk patients with strict contraindica-
ing and after therapy. The achievement of molecular tions for systemic chemotherapy.
remission strongly and significantly correlated with In contrast, based on a proposed in vitro syner-
response duration, especially among those who gism, a combined immunochemotherapy regimen of
achieved marrow MRD negativity. Further standard- rituximab and CHOP achieved high overall (96%) and
ization and wider availability of MRD assays will per- complete remission rates (48%) in a recent phase II
mit prospective assessment of molecular remission as study. However, even patients achieving a molecular
a therapeutic endpoint and the ability to test early remission displayed a median PFS of only 18.8
“pre-emptive” therapeutic intervention prior to overt months, and there appeared to be no difference in
clinical relapse. duration of response between those who did and did
not achieve molecular remission. A randomized trial
confirmed that the addition of rituximab resulted in
Initial therapy a superior ORR of 94% versus 75% with CHOP alone
In the small number of patients with limited stage (P = 0.0054); CR rates were also improved (34% versus
disease, extended-field (EF-RT) or involved-field 7%; P = 0.00024). After extended follow-up, PFS was
radiation (IF-RT) alone achieves only remissions of doubled from 14 months to 28 months, but constant
limited duration whereas such a radiation consolida- relapses were observed. Thus, additional consolidation
tion after initial systemic therapy potentially results concepts are warranted to translate the high response
in a long-term benefit. In a retrospective study of rates into long-term benefit for the patient.
stage I–II MCL, 5-year PFS and OS was 68% and In another trial by the German Lymphoma Study
71% after IF-RT, respectively, either alone or in Group, the addition of rituximab not only resulted 161
combination with conventional chemotherapy. In in significantly improved complete response rates
Chapter 10: Mantle cell lymphoma
(33% versus 0%) and slightly higher overall response PFS was prolonged with BR (35 months) in comparison
(62% versus 49%, n.s.) in relapsed or refractory MCL, to R-CHOP (22 months; P = 0.043). Thus, especially in
but, more importantly, an improved OS was observed elderly patients not qualifying for subsequent dose
after combined immunochemotherapy with R-FCM intensification, the BR regimen represents one of the
(fludarabine, cyclophosphamide, mitoxantrone) in new standard approaches.
comparison to FCM chemotherapy alone (Table
10.2). Benefit in duration of response was also dem-
onstrated when this regimen was followed by main- Radioimmunotherapy (RIT)
tenance rituximab therapy. Another approach is the application of radio-
In a large phase III study of the European MCL (131iodine or 90yttrium) labeled anti-CD20 antibod-
Network, two different regimens of immunochemo- ies. Conventional dose radioimmunotherapy (RIT)
therapy were compared in 559 first line patients age 60 delivered moderate results, with 32% overall response
years and older who were not candidates for stem cell and rapid progression, resulting in a median PFS of 6
transplant consolidation therapy. Interestingly, after months in relapsed disease. Interestingly, RIT was
the R-FC regimen higher toxicity and significantly not able to overcome either chemotherapy-refrac-
lower 4-year OS rates were observed in comparison toriness or bulky disease in this phase II study. In
to the standard R-CHOP regimen. A strong benefit for contrast, in first line treatment, RIT consolidation
maintenance rituximab over interferon maintenance after systemic induction therapy with rituximab
was observed for patients responding to R-CHOP, but plus CHOP (similar to radiation in early stages)
not to R-FC, supporting the role of maintenance rit- increased the CR/CRu rates from 20% to 55% using
uximab in this setting. standard (non-PET-confirmed) response criteria and
Bendamustine, a compound with characteristics seemed to extend duration of remission, with a
of both alkylators and purine analogs, was compared median time to treatment failure of 34 months and
to the R-CHOP regimen in a subset analysis of MCL an estimated 5-year OS of 73%. Similarly, myeloabla-
patients in another phase III trial. The bendamustine tive regimens incorporating RIT followed by autolo-
monochemotherapy plus rituximab (BR) displayed gous stem cell transplantation achieved high ORRs of
significantly less acute myelotoxicity, resulting in a 100% (91% CR) and an estimated 3-year OS of 93%
25% reduction of infectious episodes. Remarkably, even in relapsed disease.
162
Chapter 10: Mantle cell lymphoma
1. relapse
high tumor load: immunochemotherapy
immunochemotherapy (e.g. BR, R-BAC, R-FC) watch & wait ?
(e.g. BR, R-DHAP) molecular approaches ?
well-tolerated therapy:
e.g. R-Chlorambucil, BR
allo-transplant ? autologous SCT ?
radioimmunotherapy ? radioimmunotherapy ? molecular approaches
Rituximab maintenance Rituximab maintenance
higher relapse
molecular approaches:
Bortezomib, Temsirolimus, Thalidomide/Lenalidomide (preferable in combination);
B-cell receptor inhibitors (in studies)
repeat previous therapy (long remissions)
Figure 10.5 Therapeutic approaches for MCL patients (advanced stage). Note: Since no standard therapy has been established for treatment of
newly diagnosed or relapsed disease, treatment on clinical trial should be considered for all patients. (See Table 10.2 for abbreviations.)
in 16 patients with relapsed MCL, including seven and in combination with immunochemotherapy,
refractory to prior therapy, an ORR of 75% was including BR, R-CHOP, high-dose cytarabine, and
observed, including CR in 50%. The median PFS was the modified R-HyperCVAD regimens. Bortezomib
18 months. Grade 3 leukopenia was observed in 16% of can be administered safely to patients with severe
all NHL patients treated in this study, which also renal insufficiency, but is associated with peripheral
included indolent lymphoma. Non-hematologic tox- neuropathy in many patients; this may become dose-
icity was mild. A US phase II trial of BR in 12 relapsed limiting and necessitates dose-modification and vigi-
patients with MCL showed objective responses in 11 lant monitoring when given in combination with vin-
patients, including CR in over half. The median dura- cristine-containing regimens. Reactivation of herpes
tion of response was 19 months. Preliminary results of a zoster is also frequently observed in patients treated
phase III trial of BR versus fludarabine plus rituximab with bortezomib, prompting consideration of antiviral
in patients with relapsed indolent and MCL showed prophylaxis during the course of therapy.
significant benefit in ORR and CR for BR, with similar Radioimmunotherapy. RIT delivers a targeted
toxicity profiles. radiotherapeutic, 90yttrium or 131iodine, via an anti-
Bortezomib. Bortezomib targets the ubiquitin–pro- CD20 murine monoclonal antibody. Several applica-
teasome pathway and appears to provide therapeutic tions of these agents are being investigated in MCL, as
efficacy via effects on multiple cellular mechanisms in described above, including consolidation following
lymphoid neoplasms. The agent has been approved for induction immunochemotherapy and as a component
the treatment of multiple myeloma as well as relapsed of the conditioning regimen prior to stem cell
MCL. Two phase II studies, taken together, showed a transplantation. In a preliminary report of a phase II
44% ORR with 18% complete responses. Relatively single agent study in relapsed and refractory MCL,
durable responses were observed among responding responses were observed in eight of 23 patients
patients, with a median time to next treatment of (35%), with five patients achieving CR or CR uncon-
about 14 months. Bortezomib activity in MCL is firmed. The median response duration was 9.5
being further tested in ongoing multicenter studies months, with delayed myelosuppression the most fre-
166
as a single agent or in combination with rituximab, quent toxicity.
Chapter 10: Mantle cell lymphoma
Novel therapeutic approaches Table 10.4 Novel agents for treatment of relapsed or
refractory mantle cell lymphoma.
MCL responds well in most cases to initial therapy.
Improved overall and complete response rates have Agent Proposed targets
been achieved with a number of chemoimmunotherapy Lenalidomide Angiogenesis, microenvironment,
approaches as compared with previous combination and cytokines
chemotherapy regimens. However, as discussed above, Temsirolimus, mTOR (mammalian target of
most patients relapse within 1–5 years even after induc- everolimus rapamycin)
tion therapy and/or ASCT consolidation. Second line Idelalisib (GS-1101; PI3Kδ (phosphatidylinositol-3-kinase
regimens may show high therapeutic activity, although CAL-101) delta)
the durability of these responses is often short-lived. Ibrutinib (PCI-32765) BTK (Bruton tyrosine kinase)
There is a pressing need for better agents for relapsed Vorinostat HDAC (histone deacetylase)
and refractory patients, and an increasing number of Flavopiridol Cyclin D1/CDK4 (cyclin-dependent
novel agents show clinical activity in this setting. These kinase)
may be targeted to the dysregulated cell cycle elements PD 0332991 CDK4/CDK6
characteristic of this disease, or to other growth and
Obatoclax, BH3 mimetic, apoptosis (BH = BCL-2
proliferation or apoptosis pathways (Table 10.4). navitoclax homology)
Many of these are already being incorporated into front-
line regimens in an effort to improve the complete
response and PFS in MCL, either as a component of regimen with rituximab plus low-dose oral predni-
combination therapy or as maintenance or consolida- sone, etoposide, procarbazine, and cyclophosphamide
tion strategies. The following is a brief summary of (RT-PEPC) in heavily pre-treated MCL patients. An
selected agents now undergoing clinical testing. ORR of 73% and CR rate of 32% were observed, with
Immunomodulatory drugs (IMiDs). Lenalidomide antiangiogenic and microenvironmental effects postu-
is highly active in multiple myeloma and CLL, with lated to enhance therapeutic efficacy.
multiple mechanisms of action, including direct anti- mTOR inhibitors. The mammalian target of rapamy-
proliferative effects, downregulation of tumor cell/ cin (mTOR) is a downstream signaling molecule in the
stromal cell interactions with disruption of essential phosphatidylinositol-3 kinase (PI3K)/Akt pathway that
cytokine loops, immunomodulatory and antiangio- serves a critical role in regulating mRNA translation,
genic effects. Thalidomide, an earlier immunomodu- including that of cyclin D1. Temsirolimus, a derivative
latory agent with activity in MCL, led to testing of of rapamycin, has been tested in phase II single agent
lenalidomide in relapsed or refractory patients. trials in relapsed MCL, based upon the potential ability
Among heavily pre-treated MCL patients, partial to interrupt cyclin D1-dependent pathways. In a phase
and complete responses were observed with single III comparison of two doses and schedules of temsiroli-
agent oral lenalidomide. These encouraging findings mus versus investigators’ choice of therapy among 162
led to an international phase II trial in which 42% of 57 relapsed or refractory MCL patients, a schedule of
relapsed MCL patients responded, with a median PFS 175 mg weekly for 3 weeks followed by 75 mg weekly
of 5.7 months. Toxicity was predominantly reversible was found to be superior for ORR and PFS.
myelosuppression. Responses have been observed in Temsirolimus is being investigated in frontline combi-
patients relapsing after stem cell transplantation, nation regimens and with rituximab and, like the related
including CRs. A study of 134 patients progressing mTOR inhibitor everolimus, as consolidation or main-
after prior R-chemotherapy and bortezomib showed tenance therapy for MCL high-risk aggressive large cell
a 28% ORR with 8% CRs; median response duration lymphomas. Everolimus has also shown single-agent
was 16.6 months. Preclinical data demonstrating syn- efficacy in relapsed or refractory MCL, including
ergy with rituximab has led to current testing of so- patients who are refractory to bortezomib.
called “R2” regimens in CLL and indolent lymphoma, PI3K/AKT pathway. PI3K and AKT are upstream
and has shown responses in over half of patients with of mTOR and are frequently activated in MCL; indeed,
relapsed or refractory MCL. As noted above, lenalido- the delta form of PI3K is expressed in over 90% of B-
mide or the “R2”combination may also find utility as cell lymphomas, providing a logical therapeutic target
maintenance in MCL. Thalidomide also has single for small molecule inhibitors. Idelalisib (GS-1101,
agent activity and has been used in a combination
167
CAL-101) is an oral PI3K inhibitor with phase I
Chapter 10: Mantle cell lymphoma
activity in relapsed or refractory CLL and NHL, BCL-2 inhibitors/BH3 mimetics. The regulation of
including MCL, and is now being studied in a number cell death pathways is highly complex and consists of
of combination and single agent trials. both prosurvival and proapoptotic proteins, the latter
B-cell receptor (BCR) pathway inhibitors. Normal characterized by the presence of a BH3 domain. As
humoral immune response to antigen occurs via the prosurvival proteins such as BCL-2 are strongly
BCR, a signaling pathway constitutively activated in expressed in MCL, a current therapeutic strategy is to
many B-cell lymphomas that includes PI3K delta. utilize agents that mimic the BH3-only proteins which
Signal transduction via this pathway has been thera- will in turn promote apoptosis. Several BH3 mimetics
peutically targeted at several additional points in phase are in clinical trial, including obatoclax (GX 15–070)
I–II studies with oral small molecule inhibitors in and navitoclax (ABT-263).
mantle cell and other lymphomas. These targets and
agents include spleen tyrosine kinase (Syk) with fosta-
matinib, Bruton tyrosine kinase (BTK) with ibrutinib Future directions
(PCI-32765), and protein kinase C-beta (PKC-β) with MCL presents rather unique challenges among the
enzastaurin. NHLs, displaying an array of clinical behavior ranging
High activity has been identified in relapsed MCL from indolent to aggressive, and typically high
using ibrutinib. response rates to induction immunochemotherapy
HDAC inhibitors. Histone acetylation and deacety- with historically brief durations of response and poor
lation (HDAC) are important transcriptional regula- long-term survival. This situation is steadily improv-
tory mechanisms that are altered in most cancers, ing, however, with increasing rates of survival being
including lymphoma. Cyclin D1 protein levels can be realized with the advent of improved treatment strat-
downregulated in MCL cells in vitro by treatment with ification and new therapeutic agents. Current strat-
vorinostat; it also appears to inhibit the PI3K/AKT egies for management outside clinical trials are
pathway. Preliminary studies of vorinostat (already outlined in Figure 10.5, but treatment approaches
approved in the USA for treatment of T-cell lym- continue to evolve rapidly. The use of the MIPI
phoma) have shown clinical responses in MCL, with score, Ki-67 index, and gene expression profiling are
further trials of this and other HDAC inhibitors in being tested as means to risk-adapt therapy, with fur-
progress. Novel combinations with cytotoxic agents ther refinement possible via monitoring of MRD.
and bortezomib are also being pursued. Improved outcomes have been realized for younger
Cell cycle inhibitors. Cell cycle dysregulation is patients via the incorporation of rituximab and high-
present in all MCL – cyclin D1 overexpression in dose cytarabine followed by dose-intensive therapy
most patients, cyclin D2 or D3 in the remainder – with ASCT consolidation frontline. Improved survival
leading to considerable interest in therapies aimed at is also being achieved via sequential application of
these pathways. Flavopiridol is a synthetic flavone with recently available treatments as well as novel agents
mechanisms of action that include downregulation of for newly diagnosed patients not eligible for intensive
cyclin D1 and cyclin D3 as well as competitive inhib- frontline therapy and for those experiencing relapse.
ition of the cyclin-dependent kinases CDK4 and Although challenging, MCL continues to provide
CDK6. A Canadian study found only a modest 11% important insights to cancer pathogenesis, including
response rate in MCL for single agent flavopiridol cell cycle dysregulation, apoptosis, and cell signa-
using a short-infusion protocol, although subsequent ling pathways that ultimately may prove useful in a
pharmacokinetically driven schedules have shown variety of other hematologic and non-hematologic
much higher response rates in refractory CLL that malignancies.
may inform better approaches for MCL (Table 10.4).
Directly targeting CDK4/CDK6, an approach which
theoretically would circumvent the upregulation of D2 Further reading
or D3 that may follow D1 inhibition, is being explored Andersen NS, Pedersen LB, Laurell A, et al. Pre-emptive
with PD 0332991. This oral agent has shown activity in treatment with rituximab of molecular relapse after
relapsed MCL, and is also being tested in combination autologous stem cell transplantation in mantle cell
with bortezomib. lymphoma. J Clin Oncol, 2009;27:4365–4370.
168
Chapter 10: Mantle cell lymphoma
Ansel SM, Tang H, Kurtin PJ, et al. Temsirolimus and Goy A, Bernstein S, Kahl B, et al. Bortezomib in patients with
rituximab in patients with relapsed or refractory mantle relapsed or refractory mantle cell lymphoma: updated
cell lymphoma: a phase 2 study. Lancet Oncol, time-to-event analyses of the multicenter phase 2
2011;12:361–368. PINNACLE study. Annals Oncol, 2008;20:520–525.
Determan O, Hoster E, Ott G, et al. Ki-67 predicts outcome Goy A, Sinha R, Williams ME, et al. Phase II multicenter study
in advanced-stage mantle cell lymphoma patients treated of single-agent lenalidomide in subjects with mantle cell
with anti-CD20 immunotherapy: results from lymphoma who relapsed or progressed after or were
randomized trials of the European MCL Network and the refractory to bortezomib: the MCL-001 Study (EMERGE™
German Low Grade Lymphoma Study group. Blood, Trial). Blood, 2012;120(Suppl.):abstract #905.
2008;111:2385–2387.
Hermine O, Hoster E, Szymczyk M, et al. Alternating courses
Dreyling M, Lenz G, Hoster E, et al. Early consolidation by of 3x chop and 3x dhap plus rituximab followed by a high
myeloablative radiochemotherapy followed by dose ara-c containing myeloablative regimen and
autologous stem cell transplantation in first remission autologous stem cell transplantation (ASCT) is superior
significantly prolongs progression-free survival in mantle to 6 courses CHOP plus rituximab followed by
cell lymphoma – results of a prospective randomized trial myeloablative radiochemotherapy and ASCT in mantle
of the European MCL Network. Blood, cell lymphoma: final results of the MCL Younger Trial of
2005;105:2677–2684. the European Mantle Cell Lymphoma Network. Blood,
Ek S, Dictor M, Jerkeman M, et al. Nuclear expression of the 2012;120(Suppl.): abstract #151.
non B-cell lineage Sox11 transcription factor identifies Hess G, Herbrecht R, Romaguera J, et al. Phase III study to
mantle cell lymphoma. Blood, 2008;111:800–805. evaluate temsirolimus compared with investigator’s
Fernandez V, Salamero O, Espinet B, et al. Genomic and choice therapy for the treatment of relapsed or refractory
gene expression profiling defines indolent forms of mantle cell lymphoma. J Clin Oncol, 2009;27:3822–3829.
mantle cell lymphoma. Cancer Res, 2010;70:1408–1418. Hoster E, Dreyling M, Klapper W, et al. A new prognostic
Forstpointner R, Dreyling M, Repp R, et al. The addition of index (MIPI) for patients with advanced-stage mantle cell
rituximab to a combination of fludarabine, lymphoma. Blood, 2008;111:558–565.
cyclophosphamide, mitoxantrone (FCM) significantly Howard OM, Gribben JG, Neuberg DS, et al. Rituximab and
increases the response rate and prolongs survival as CHOP induction therapy for newly diagnosed mantle-
compared to FCM alone in patients with relapsed and cell lymphoma: Molecular complete responses are not
refractory follicular and mantle cell lymphomas – results predictive of progression-free survival. J Clin Oncol,
of a prospective randomized study of the German low 2002;20:1288–1294.
grade lymphoma study group (GLSG). Blood, Iqbal J, Shen Y, Liu Y, et al. Genome-wide miRNA profiling
2004;104:3064–3071. of mantle cell lymphoma reveals a distinct subgroup with
Forstpointner R, Unterhalt M, Dreyling M, et al. poor prognosis. Blood, 2012;119:4939–4948.
Maintenance therapy with rituximab leads to a Kirschbaum M, Frankel P, Popplewell L, et al. Phase II study
significant prolongation of response duration after of vorinostat for treatment of relapsed or refractory
salvage therapy with a combination of rituximab, indolent non-Hodgkin’s lymphoma and mantle cell
fludarabine, cyclophosphamide and mitoxantrone lymphoma. J Clin Oncol, 2011;28:1198–1203.
(R-FCM) in patients with relapsed and refractory
follicular and mantle cell lymphomas – results of a Kluin-Nelemans HC, Hoster E, Hermine O, et al. Treatment
prospective randomized study of the German Low Grade of older patients with mantle-cell lymphoma. N Engl J
Lymphoma Study Group (GLSG). Blood, Med, 2012;367:520–531.
2006;108:4003–4008. Kridel R, Meissner B, Rogic S, et al. Whole transcriptome
sequencing reveals recurrent NOTCH1 mutations in
Geisler CH, Kolstad A, Laurell A, et al. The Mantle Cell
mantle cell lymphoma. Blood, 2012;119:1963–1971.
Lymphoma International Prognostic Index (MIPI)
is superior to the International Prognostic Index (IPI) Lefrere F, Delmer A, Levy V, et al. Sequential
in predicting survival following intensive first-line chemotherapy regimens followed by high-dose therapy
immunochemotherapy and autologous stem with stem cell transplantation in mantle cell lymphoma:
cell transplantation (ASCT). Blood, 2010;115: an update of a prospective study. Haematologica,
1530–1533. 2004;89:1275–1276.
Ghielmini M, Schmitz Hsu S-F, Cogliatti S, et al. Effect of Lenz G, Dreyling M, Hoster E, et al. Immunochemotherapy
single-agent rituximab given at the standard schedule or with rituximab and CHOP significantly improves
as prolonged treatment in patients with mantle cell response and time to treatment failure, but not long-term
lymphoma: a study of the Swiss Group for Clinical outcome in patients with previously untreated mantle cell
Cancer Research (SAKK). J Clin Oncol, 2005;23:705–711. lymphoma: results of a prospective randomized trial of 169
Chapter 10: Mantle cell lymphoma
the German Low Grade Lymphoma Study Group rituximab-Hyper-CVAD alternating with rituximab-
(GLSG). J Clin Oncol, 2005;23:1984–1992. high dose methotrexate/cytarabine (R-MA) and
Leonard JP, LaCasce AS, Smith MR, et al. Selective CDK4/6 without stem cell transplantation in patients with
inhibition with tumor responses by PD0332991 in untreated aggressive mantle cell lymphoma. Br J
patients with mantle cell lymphoma. Blood, Haematol, 2010;150:200–208.
2012;119:4597–4607. Ruan J, Martin P, Coleman M, et al. Durable responses with
Magni M, Di Nicola M, Carlo-Stella C et al. High-dose the metronomic rituximab and thalidomide plus
sequential chemotherapy and in vivo rituximab-purged prednisone, etoposide, procarbazine, and
stem cell autografting in mantle cell lymphoma: a 10-year cyclophosphamide regimen in elderly patients with
update of the R-HDS regimen. BMT, 2009;43:509–511. recurrent mantle cell lymphoma. Cancer,
2010;116:2655–2664.
Martin P, Chadburn, A, Christos P, et al. Outcome of
deferred initial therapy in mantle cell lymphoma. J Clin Rummel MJ, Niederle N, Maschmeyer G, et al.
Oncol, 2009;27:1209–1213. Bendamustine plus rituximab is superior in respect of
progression free survival and CR rate when compared
Mato AR, Svoboda J, Feldman T, et al. Post-treatment (not
to CHOP plus rituximab as first-line treatment of
interim) positron emission tomography–computed
patients with advanced follicular, indolent, and
tomography scan status is highly predictive of outcome in
mantle cell lymphomas: final results of a
mantle cell lymphoma patients treated with R-
randomized phase III study of the StiL. Blood,
HyperCVAD. Cancer, 2012;118:3565–3570.
2009;114(22):168.
Ondrejka SL, Lai R, Smith SD, His ED. Indolent mantle
Rummel MJ, Kaiser U, Balser C, et al. Bendamustine
cell leukaemia: a clinicopathological variant
plus rituximab versus fludarabine plus rituximab in
characterized by isolated lymphocytosis, interstitial
patients with relapsed follicular, indolent and mantle
bone marrow involvement, kappa light chain
cell lymphomas–final results of the randomized phase
restriction and good prognosis. Haematologica,
III study NHL 2–2003 on behalf of the StiL group.
2011;96:1121–1127.
Blood, 2010;116(21):373.
Perez-Galan P, Dreyling M, Wiestner A. Mantle cell
lymphoma: biology, pathogenesis, and the molecular basis Rummel MJ, Niederle N, Maschmeyer G, et al.
of treatment in the genomic era. Blood, 2011;117:26–38. Bendamustine plus rituximab versus chop plus
rituximab as first-line treatment in patients with
Pott C, Hoster E, Beldjord K, et al. R-CHOP/R-DHAP indolent and mantle cell lymphomas: updated results
compared to R-CHOP induction followed by high dose from the stil nhl1 study on behalf of the StiL (Study
therapy with autologous stem cell transplantation induces Group indolent Lymphomas, Germany). Proc ASCO
higher rates of molecular remission in MCL: results of the 2012; abstract #3.
MCL Younger Intergroup Trial of the European MCL
Network. Blood, 2010;116(21):425. Schaffel R, Hedvat CV, Teruya-Feldstein J, et al. Prognostic
impact of proliferative index determined by quantitative
Pott C, Hoster E, Delfau-Larue M-H, et al. Molecular
image analysis and the International Prognostic Index in
remission is an independent predictor of clinical outcome
patients with mantle cell lymphoma. Annals Oncol,
in patients with mantle cell lymphoma after combined
2010;21:133–139.
immunochemotherapy: a European MCL intergroup
study. Blood, 2010;115:3215–3223. Smith MR, Li H, Gordon L, et al. Phase II study of
Racke F, Simpson S, Christian B, et al. Evidence of long rituximab plus CHOP immunochemotherapy
latency periods prior to development of mantle cell followed by yttrium-90-ibritumomab tiuxetan in
lymphoma. Blood, 116(21):147. untreated mantle-cell lymphoma: ECOG Study E1499.
J Clin Oncol, 2012;30:3119–3126.
Renner C, Zinzani P, Gressin R, et al. A multicenter
phase II trial (SAKK 36/06) of single-agent Swerdlow SH, Williams ME. From centrocytic to
everolimus (RAD001) in patients with relapsed or mantle cell lymphoma: a clinicopathologic and
refractory mantle cell lymphoma. Haematologica, molecular review of 3 decades. Human Pathol,
2012;97:1085–1091. 2002;33:7–20.
Robinson KS, Williams ME, van der Jagt RH, et al. Phase II Swerdlow SH, Campo E, Harris NL, et al. WHO
multicenter study of bendamustine plus rituximab in Classification of Tumours of Haematopoietic and
patients with relapsed indolent B-cell and mantle cell Lymphoid Tissues. 4th edn. WHO Press, 2008.
non-Hodgkin’s lymphoma. J Clin Oncol, Tam CS, Bassett R, Ledesma C, et al. Mature results of the M.
2008;26:4473–4479. D. Anderson Cancer Center risk-adapted transplantation
Romaguera JE, Fayad LE, Feng L, et al. Ten-year follow- strategy in mantle cell lymphoma. Blood,
170 up after intense chemoimmunotherapy with 2009;113:4144–4152.
Chapter 10: Mantle cell lymphoma
Wang M, Fayad L, Wagner-Bartak N, et al. Lenalidomide relapsed or refractory mantle cell lymphoma: durable
in combination with rituximab for patients efficacy and tolerability with longer follow-up. Blood,
with relapsed or refractory mantle-cell 2012;120(Suppl.):abstract #904.
lymphoma: a phase 1/2 clinical trial. Lancet Oncol,
Zhao J-J, Lin J, Lwin T, et al. MicroRNA expression
2012;13:716–723.
profile and identification of miR-29 as a prognostic
Wang M, Rule SA, Martin P, et al. Interim results of an marker and pathogenetic factor by targeting
international, multicenter, phase 2 study of Bruton’s CDK6 in mantle cell lymphoma. Blood,
tyrosine kinase inhibitor, ibrutinib (PCI-32765), in 2010;115:2630–2639.
171
Chapter
11
Burkitt and lymphoblastic lymphoma:
clinical therapy and outcome
Nirali N. Shah, Alan S. Wayne, and Wyndham H. Wilson
Pathology: Andrew Wotherspoon Molecular cytogenetics: Andreas Rosenwald and German Ott
Introduction and presentation Most individuals without frank leukemia present with
advanced-stage disease (III/IV), and bulky adeno-
Burkitt (BL) and lymphoblastic lymphomas (LBL) are pathy, anterior mediastinal mass, pleural effusion,
highly aggressive diseases with distinct natural histories and/or massive organomegaly are common. Patients
and clinical presentations. BL mostly occurs in the first should be closely evaluated for emergent complica-
two decades of life and accounts for 1–2% of all lympho- tions such as airway obstruction, superior vena cava
mas. Three clinical variants are recognized: (1) endemic syndrome, and pericardial tamponade. Marrow
BL, which is primarily found in equatorial Africa; involvement may result in cytopenias and/or hyper-
(2) sporadic BL, which presents worldwide but is leukocytosis, and CNS involvement is frequent
the most common type in western countries; and (5–15%) at diagnosis. Rare individuals with B-lineage
(3) immunodeficiency-associated BL, which is associated LBL present with more confined disease (e.g. bone,
with HIV infection. There are important clinical differ- skin). LBL/T-ALL has historically fared poorer than
ences in these variants (Table 11.1), with endemic BL B-precursor ALL; however, intensified treatment has
involving the jaw, orbit, and paraspinal regions in half of minimized this difference. Currently, the outcome for
the cases as well as the mesentery and gonads, while children with LBL is excellent, although older individ-
sporadic BL mostly involves the distal ileum, cecum, uals have lower relapse-free survival rates.
and/or mesentery, and rarely the jaw. When bulky or
disseminated disease is present, extranodal involvement
of the ovaries, kidney, breasts, and/or CNS may be seen. Pathology
Clinical presentation in a Berlin–Frankfurt–Munster
Group (BFM) series of 152 pediatric patients included Burkitt lymphoma
advanced stage (III/IV) disease in 38%, bone marrow BL is characterized by a diffuse proliferation of rather
involvement in 33%, and central nervous system (CNS) monotonous intermediate-sized cells with little pleo-
disease in 4%. Of the patients in this series, 27% pre- morphism. Within the sheet of neoplastic cells are dis-
sented as acute leukemia, referred to as the L3 subtype of persed macrophages containing apoptotic nuclear
acute lymphoblastic leukemia (ALL) within the French– debris that give the so-called “starry sky” to the low
American–British (FAB) classification, or B-ALL. BL power appearance of this lymphoma (Figure 11.1). The
infrequently presents in adults, but does occur with cells have deeply basophilic cytoplasm that contains
increased frequency in patients with HIV infection. small vacuoles (best seen on imprint/cytology sections,
LBL is most commonly a malignancy of T-cell but which can sometimes be seen at the edge of tissue
precursor cells, and, as such, it is identical to T-cell sections) that contain fat. The nuclei are not usually
acute lymphoblastic leukemia (T-ALL). The World larger than those of the associated macrophages and are
Health Organization (WHO) classifies these as a single characteristically round with a smooth contour and
entity, with the exception that blasts make up <25% of little variation in size or shape. The nuclear chromatin
nucleated bone marrow cells in LBL. T-ALL makes up is granular and there are 2–5 rather indistinct small
approximately 15% of pediatric and 25% of adult ALL. nucleoli. In some BLs associated with HIV infection
172 Lymphoma, Second Edition, ed. Robert Marcus, John W. Sweetenham, and Michael E. Williams. Published by
Cambridge University Press. © Cambridge University Press 2014.
Chapter 11: Burkitt and lymphoblastic lymphoma
Figure 11.1 Burkitt’s lymphoma. A diffuse proliferation of cells and Figure 11.2 Burkitt’s lymphoma. The proliferation marker Ki-67
scattered macrophages with apoptotic debris giving a “starry sky” stains >98% of tumor cell nuclei.
appearance.
the cells may have slightly eccentric nuclei and a plas- This variation in morphological appearance is more
macytoid appearance. Mitoses are frequent. common in adults and in cases with nodal rather than
Immunophenotypically the cells are positive for extranodal sites of involvement. In some cases the
CD45 and express the B-cell markers CD20 and atypical appearance may be due to poor fixation.
CD79a. They are positive for markers of germinal The current 2008 fourth edition of the WHO clas-
center origin CD10 and BCL-6 but, with exceptionally sification of tumors of the hemopoietic and lymphoid
rare exceptions, lack BCL-2 protein expression. They tissues recognizes an overlap between cases of BL and
are negative for CD5 and CD23. The proliferation diffuse large B-cell lymphoma (DLBCL), referring to
marker Ki-67 stains >98% of tumor cell nuclei such cases as “B cell lymphoma, unclassifiable, with
(Figure 11.2). The cells are negative for terminal de- features intermediate between diffuse large B cell lym-
oxynucleotidyl transferase (TdT). phoma and Burkitt lymphoma.” This group consists of
Previous classifications (Revised European aggressive lymphomas that either have a cellular mor-
American Lymphoma [REAL] classification and the phology resembling BL but with a high proliferation,
third edition of WHO) recognized a group of lympho- starry sky appearance, and immunophenotype consis-
mas designated Burkitt-like or atypical BL. In general, tent with classical BL or have a morphology more
these had a morphology and immunophenotype sim- reminiscent of true BL but atypical immunophenotype 173
ilar to classical BL but showed greater pleomorphism. or genotype.
Chapter 11: Burkitt and lymphoblastic lymphoma
directly 5' upstream, or within the first intron or exon are frequent, with complex karyotypes as evidenced by
of MYC and in the switch region of IGH. Therefore, array comparative genomic hybridization (CGH)
the two tumor types obviously are distinguished by studies.
distinct differentiation stages, in which the transloca-
tions occur, namely either during VDJ rearrangement
and class switch recombination. Precursor B-lymphoblastic leukemia/
Other genes frequently altered in BL comprise
TP53 and CDKN2A. Also, MYC itself is frequently
lymphoma
targeted by mutations. BL usually carry few secondary The translocation t(12;21)(p13;q22), resulting in a
aberrations, most frequently gains in 1q, 7q, 12q, 20q, TEL–AML1 fusion transcript, is detected in up to 25%
and Xq, and losses in 13q. Two landmark studies of childhood ALL. Both TEL (also known as ETV6) and
revealed that BL has a distinct gene expression signa- AML1 (also known as RUNX1) genes are involved in
ture, allowing for construction of a molecular classi- other leukemia-associated translocations as well, e.g. by
fier. As would have been expected, MYC and its target generating fusion transcripts with PDGFRβ, MN1, ABL,
genes are commonly upregulated in BL, but also ger- EVI1, JAK2, and CBFβ. In contrast to childhood ALL,
minal center B-cell-related genes were expressed at the TEL–AML1 translocation represents a relatively
higher levels. In contrast, MHC class I and NF-κB- rare genetic aberration in adult ALL patients (3% of
related genes were expressed at lower levels in BL in adult B-ALL). In both age groups, however, the trans-
comparison to DLBCL. Applying this molecular clas- location confers a highly favorable prognosis, with
sifier to aggressive B-cell lymphomas, BL were clearly event-free survival (EFS) rates close to 90%.
set apart from DLBCL, although in both studies, an Microarray-based gene expression profiling has
“intermediate” or “gray zone” group of cases was iden- revealed a distinctly homogeneous expression pattern
tified, not readily falling into one of the clearly defined in ALL cases with TEL–AML1 translocations, suggest-
core groups. Most strikingly, both groups identified a ing that tumors harboring this aberration represent a
group of aggressive tumors, morphologically indistin- unique leukemia subtype. Interestingly, the tumor cells
guishable from DLBCL but presenting with a BL gene have been shown to overexpress the erythropoietin
expression profile – the so-called “molecular BL.” This receptor, which potentially affects proliferation and
finding clearly raises the question whether these cases survival signaling of the tumor cells.
should be treated with intensified BL-like chemother- Hyperdiploidy (more than 50 chromosomes per
apy regimens. Another finding substantiated in these leukemic cell) is also a common genetic alteration
studies was that up to 10% of DLBCL identified by detected in about 30% of pediatric and 9% of adult
morphology and molecular gene expression profile B-ALL cases. This condition also defines a patient
carry MYC translocations. Preliminary data would subgroup with an excellent clinical outcome. In addi-
suggest that these cases are likely to pursue a worse tion, pediatric cases with trisomies 4, 10, and 17 may
clinical course than MYC-negative DLBCL. have a specifically favorable prognosis. In about 20%
of patients with hyperdiploid chromosome clones,
activating mutations in the receptor tyrosine kinase
B-cell lymphoma, unclassified, with gene FLT3 have been demonstrated, suggesting a pos-
features intermediate between DLBCL sible therapeutic benefit from the administration of
small tyrosine kinase inhibitor molecules in this
and Burkitt lymphoma group.
These cases harbor, both from a morphological as well The translocation t(9;22), encoding the BCR–ABL
as from a genetic view, intermediate features between fusion protein, can be detected in about 30% of adult
BL and DLBCL. Judging from available data, 35–50% of and 5% of pediatric B-ALL tumors, and confers an
cases carry MYC translocations. In contrast to classical adverse prognosis in both age groups. The different
BL, however, MYC in these cases is frequently trans- incidence of the t(9;22) among pediatric and adult
located to non-IG genes. Most cases with dual trans- B-ALL may, at least in part, account for the overall
locations, the so-called “double hit lymphomas,” difference in clinical outcome between the two age
involving mainly MYC and BCL-2, and, to a lesser groups, and bone marrow transplantation during
extent, MYC and BCL-6, fall into this category. In first remission is often recommended in B-ALL 175
contrast to classical BL, secondary genetic abnormalities patients with a BCR–ABL fusion transcript.
Chapter 11: Burkitt and lymphoblastic lymphoma
In contrast to hyperdiploidy, hypodiploidy (fewer LMO1 and LMO2 genes. LMO1 and LMO2 proteins are
than 45 chromosomes) is an infrequent finding in highly expressed in the central nervous system but
B-ALL patients (<2%), and has been associated with nearly inactive in T-cells and their progenitors. The
poor outcome. deregulated expression of HOX11 in T-cells, as a con-
The tumors in about 6% of B-ALL patients may sequence of either the t(10;14)(q24;q11) or the t(7;10)
harbor the t(1;19)(q23;p13) chromosome transloca- (q35;q24), promotes transformation via disruption of
tion that generates the E2A–PBX1 fusion transcript. normal cellular regulatory networks. These aberrations
This chimeric protein interferes with hematopoietic occur in approximately 30% of adult T-ALL patients
differentiation by disrupting the HOX gene-dependent and in only 3% of childhood ALL cases. In both age
regulatory network. groups, abundant ectopic HOX11 protein expression is
Aberrations involving the chromosomal band associated with a favorable clinical outcome. The
11q23 are detected in lymphoblastic and acute myeloid HOX11L2 gene located in the chromosomal band 5q35
leukemias. The mixed-lineage leukemia (MLL) gene in is deregulated in a subset of T-ALL patients by either the
11q23 is involved in translocations with more than 40 t(5;14)(q35;q11) or t(5;14)(q35;q32). The latter trans-
different translocation partners, generating fusion location results in HOX11L2 overexpression by juxta-
proteins harboring the NH2-terminus of the MLL posing it to the BCL11b gene that is differentially
gene and COOH-terminus of the partner genes. expressed during T-cell development. Using molecular
Translocations involving MLL probably play a role in techniques, HOX11L2 translocations are detected in
leukemogenesis via deregulation of the HOX pathway. about 20% of T-ALL patients, but are an infrequent
The most common MLL translocation, the t(4;11) finding in adults. While the NOTCH1 gene is rarely
(q21;q23), results in an AF4/MLL fusion protein. It is targeted by chromosomal translocations, activating
present in almost 50% of infant ALL cases and, less mutations in this gene have been detected in approx-
frequently, in other age groups, and is associated with imately 50% of T-ALL tumors of all different molecular
poor prognosis. Approximately 20% of the cases with subtypes. The NOTCH1 gene encodes a transmembrane
MLL translocations show a high expression of the receptor involved in the regulation of normal T-cell
FLT3 receptor tyrosine kinase due to a mutation in development, and chromosomal aberrations involving
the activation loop region. This finding suggests that NOTCH1 have been shown to induce T-ALL in mouse
MLL fusion proteins may cooperate with activated models. New ABL1 fusion genes have been identified
tyrosine kinases in promoting leukemogenesis. In that provide proliferation and survival advantage to
keeping with this concept, FLT3 inhibitors are anti- lymphoblasts. Recently, microarray based gene expres-
tumor effective in vitro and in mouse models. sion studies revealed that T-ALL cases overexpressing
HOX11 and HOX11L2 as well as T-ALL with MLL
fusion genes share the same expression pattern, charac-
Precursor T-lymphoblastic leukemia/ terized by a global HOXA gene deregulation signature.
lymphoma The translocation t(11;19)(q23;p13) resulting in an
Chromosomal rearrangements in precursor MLL–ENL fusion transcript occurs in acute myeloid
T-lymphoblastic leukemia/lymphoma frequently result leukemia patients, but also in patients with B-ALL and
in proto-oncogene deregulation by juxtaposing them to also T-ALL. In T-ALL patients, the translocation is
T-cell receptor (TCR) regulatory elements. Important associated with a favorable prognosis and long-term
genes involved in such rearrangements are TAL1, MYC, survival. T-ALL with a CALM–AF10 fusion, generated
LYL1, LMO1, LMO2, HOX11, and HOX11L2. TAL1 by the t(10;11)(p13;q14), characteristically display
overexpression is induced by the translocation t(1;14) immature phenotypic features, either lacking TCR
(1p32;q11) or by deletion of upstream regulatory expression or expressing TCRγ/δ. Using genome-wide
elements of the gene, and occurs in about 25% of pedia- profiling strategies in T-ALL, multiple genetic abnor-
tric T-ALL. The deregulated expression of TAL1 prob- malities have been identified, including deletions and/or
ably activates a set of genes that are normally quiescent mutations targeting TAL1, RB1, PTEN, and others.
in T-cell progenitors or, alternatively, may inactivate Analysis of copy-neutral loss of heterozygosity (CN-
E2A homodimers and heterodimers. The translocations LOH) revealed frequent CN-LOH in the vicinity of the
t(11;14)(p15;q11) and t(11;14)(p13;q11) are thought to CDKN2A/B gene in 9p that, however, also usually har-
176 affect similar pathways via the ectopic expression of the bors focal deletions.
Chapter 11: Burkitt and lymphoblastic lymphoma
Early treatment strategies for BL were modeled on schedules, indicating the empiric nature of the actual
ALL regimens which employed dose-intense and pro- combinations (Table 11.3). Indeed, the optimal dose
longed treatment with induction, consolidation, and and schedule of cyclophosphamide, methotrexate, and
maintenance phases. These approaches stand in con- cytarabine remain unknown and vary among the major
trast to the significantly less dose-intense regimens regimens (Table 11.4).
used in adults with “intermediate-grade” lymphoma, The risks of tumor lysis syndrome and propensity
such as CHOP and CHOP-based regimens, that only for CNS dissemination in BL both have important
produced a 50–60% event free survival (EFS). While treatment implications. To reduce tumor lysis, which
dose intensity and dose density are important treat- can produce life-threatening electrolyte imbalances
ment components for BL, later studies indicated that and renal failure, many regimens employ a pre-phase
shorter treatment durations were equally successful whereby relatively low-dose cyclophosphamide and
and less toxic. Furthermore, the recognition that prednisone are administered. This strategy has been
tumor volume is an important prognostic feature led incorporated into the regimens of the French Society
to the use of risk-adaptive approaches and a further of Pediatric Oncology (SFOP), German Multicenter
reduction in treatment for early stage patients. ALL Group (GMALL), and BFM, but not CODOX-M
Several biological characteristics of BL have helped or hyper-CVAD (Table 11.4). The high risk of CNS
guide treatment strategies, including its high prolifera- involvement is handled by the use of relatively high-
tive fraction. It has been recognized for years that BL is dose intravenous methotrexate and cytarabine, both of
sensitive to multiple chemotherapy classes and, in which have CNS penetration, and intrathecal chemo-
endemic BL, apparent cures were occasionally achieved therapy administration. An important advancement
with single agent cyclophosphamide. Despite initial sen- has been to reduce intrathecal treatment and eliminate
sitivity, however, patients frequently relapsed, particu- whole brain radiation for prophylaxis, which has sig-
larly those with higher volume disease. This apparent nificantly reduced CNS toxicity.
dichotomy can potentially be explained by the high
tumor proliferative rate. The role of tumor cell kinetics
was raised some 30 years ago by Skipper et al., who Treatment of Burkitt lymphoma
observed that the fraction of cells undergoing DNA Beginning in 1981, the SFOP conducted a series of
replication, termed the “growth fraction,” greatly influ- protocols to refine the treatment of BL and B-ALL
enced drug sensitivity, a finding that likely reflects the (LMB). These studies demonstrated that short dose-
greater sensitivity during S-phase to many drug classes. intensive treatment was effective in patients without
Although a high growth fraction would predict greater CNS involvement, and that dose intensification of
drug sensitivity, it could also lead to greater tumor methotrexate, cytarabine, and etoposide with cranial
proliferation between cycles. Depending on the relative irradiation improved the EFS of patients with CNS
impact of these two effects, cure could increase owing to involvement to 75%. Based on these findings, a risk-
higher fractional cell kill or decrease if tumor prolifer- adapted protocol (LMB 89 protocol) was developed
ation between cycles leads to a “kinetic failure.” One in which treatment was based on tumor burden and
strategy to overcome such “kinetic failure” is to increase early response to chemotherapy (Table 11.4). Using
dose density through frequent chemotherapy adminis- the St. Jude (Table 11.2) staging, group A included
tration, a strategy employed by most BL regimens. stage I or II with abdominal resection; group B
Another strategy is to increase the fractional cell kill or included unresected stage I, non-abdominal stage II,
efficacy of chemotherapy, thereby reducing the number any stage III or IV disease, and/or B-ALL (CNS-
of tumor cells that can survive and proliferate between negative and <70% marrow blasts); and group C
cycles. Hence, common BL regimens are relatively high patients had CNS involvement and/or >70% marrow
dose and employ multiple chemotherapy agents admin- blasts. Based on these risks, group A only received
istered in alternating cycles. They typically include induction, group B received pre-phase, induction,
anthracylines, epipodophyllotoxins, vinca alkaloids, consolidation, and limited maintenance, and group
and alkylators, as well as methotrexate and cytarabine, C also received extended maintenance and cranial
which are cell cycle active agents and take advantage of irradiation if the CNS was involved. If a CR was not
the high tumor proliferation. These agents, however, are achieved in groups B and C after the third or fourth
178 administered in a variety of combinations and induction-consolidation course, patients underwent
Chapter 11: Burkitt and lymphoblastic lymphoma
Table 11.3 Comparison of dose and dose intensity among Burkitt lymphoma regimens.
autologous transplant. This strategy was highly success- version in which intrathecal methotrexate was reduced
ful in pediatric patients, with 5-year EFS and OS of 92% and cranial radiation was only administered to high-risk
(Table 11.4). The success of LMB 89 in pediatrics led to patients (Table 11.4). These patients were compared to
its testing in adults with minor modifications another cohort who received the standard CNS prophy-
(Table 11.4). The outcome in 72 adult patients, mostly laxis. Overall, 92 patients enrolled on the two cohorts,
with advanced disease, was favorable, with EFS and and there were no differences in outcome with EFS and
overall survival (OS) of 65% and 70%, respectively, at OS of 45–52% and 50–54%, respectively, at 3 years. This
2 years. Toxicity remains a problem for advanced stage study suggested that short-duration treatment with less
patients because of the treatment intensity. Treatment- intensive CNS prophylaxis was equally effective, with
related deaths from infection were higher in adults significantly less neurotoxicity in comparison to more
compared to pediatrics, with incidences of 5% and intensive prophylaxis.
1.6%, respectively, among patients in groups B and Another highly effective regimen for BL was devel-
C. Myelosuppression was the primary treatment com- oped by the BFM (Table 11.4). Like other regimens for
plication, with over 40% of adults experiencing febrile BL, the BFM approach was based on short, intensive
neutropenia. Tumor lysis syndrome requiring dialysis cycles and, over the course of several protocols, led to a
was prevented by the pre-phase treatment and associ- reduction in the number of cycles based on risk strat-
ated supportive care. ification. The BFM 90 protocol continued to refine the
The GMALL reported excellent results with an risk stratification further and to improve the outcome
intensive 18-week regimen in adult patients with of patients who had an incomplete initial response with
B-ALL. In an effort to reduce CNS toxicity, the Cancer further treatment intensification. Among 322 pediatric 179
and Leukemia Group B (CALGB) studied a modified patients with BL or B-ALL treated with the BFM 90
Chapter 11: Burkitt and lymphoblastic lymphoma
Regimen Patients Histology Median age Stage (%) CR (%) EFS (%) OS (%)
(No.) (No.) (years)
(range)
LMB 89 561 Burkitt and 8 (0.17–18) III–IV 97% 92% (5-year) 92%
B-ALL (420) 79% (5-year)
Modified LMB 72 Burkitt and 33 (18–76) III–IV 72% 65% (2-year) 70%
B-ALL 67% (2-year)
GMALL 35 B-ALL 36 (18–65) N/A 74% 71% (4-year) 51%
B-NHL 86 (DFS) (4-year)
Modified GMALL 92 Burkitt and 47 (17–78) III–IV 74% 45–52% 50–54%
B-ALL 89% (3-year) (3-year)
BFM 90 413 Burkitt and 9 (1.2–17.9) III–IV N/A 89% (6-year) 14
B-ALL (322) 60% deaths
CODOX-M/ IVAC 21 peds Burkitt 12 (3–17) III–IV 95% 85% (peds) 2 deaths
20 adult 25 (18–59) 78% 100% (adults)
(2-year)
CODOX-M/ 52 Burkitt 35 (15–60) III–IV 77% 65% (2-year) 73%
IVAC 61% (2-year)
CODOX-M/IVAC 53 Burkitt 37 (17–76) III–IV 64% (2-year) 67%
(lower dose 76% (PFS) (2-year)
methotrexate 3 g/m2)
Hyper-CVAD 26 B-ALL 58 (17–79) N/A 81% 61% (3-year) 49%
(DFS) (3-year)
Hyper-CVAD/R 31 Burkitt / 46 (17–77) IV 45% 86% 80% (3-year) 89%
B-ALL >60 (29%) (3-year)
DA-EPOCH-R 25 Burkitt 30 (18–66) III–IV 56% 100% 96% (3-year) 100%
(3-year)
5-drug induction 27 Burkitt 36 (15–64) III–IV 44% 81% 73% (5-year) 81%
followed by BEAM (5-year)
and autologous SCT
regimen, EFS was 89% at 6 years. Importantly, this developed in which CODOX-M was alternated with
represented a significant advance for advanced-stage IVAC and administered for four cycles in high-risk
patients compared to the previous BFM 86 protocol. patients. This approach yielded an overall EFS of 85%
Two clinical trials conducted at the National Cancer compared to 55% for all patients treated on CODOX-M
Institute (NCI protocols 77–04 and 89-C-41) diverged alone, with no difference in low- and high- risk disease.
from the leukemia model of treatment and investigated When the CODOX-M/IVAC regimen was tested in an
the role of more limited treatment (Table 11.4). In the adult cooperative group trial, however, only a 65%
77–04 protocol, three cycles of intravenous cyclophos- 2-year EFS was achieved. A recent prospective trial
phamide, vincristine, doxorubicin, methotrexate, and incorporated risk stratification and dose-modification
leucovorin rescue with cytarabine intrathecal prophy- of CODOX-M in adults with or without IVAC based
laxis (CODOX-M) showed that intensive combination on staging. The methotrexate dose was reduced from 6.7
chemotherapy produced an EFS of 66% in patients with g/m2 to 3 g/m2. EFS and OS at 2 years were 64% and
low-risk disease. However, patients with high-risk dis- 67%, respectively. Although there was no significant
ease (i.e. St. Jude stage IV) fared poorly, with a 19% EFS. reduction in toxicity, as outcomes in this trial were
Based on the observation that some patients were sal- similar to results with higher dose methotrexate, the
vaged with combination ifosfamide, etoposide, and authors suggested that the lower dose methotrexate
cytarabine (IVAC), a successor protocol 89-C-41 was should be the new standard of care in this treatment arm.
180
Chapter 11: Burkitt and lymphoblastic lymphoma
The hyper-CVAD regimen was based on a mod- chemotherapy). The Children’s Oncology Group
ification of a regimen developed by Murphy et al. for (COG) has also recently completed a prospective trial
pediatric B-ALL (Table 11.4). In this regimen, hyper- that incorporated rituximab into the LMB 96 regimen.
fractionated cyclophosphamide, vincristine, doxoru- Analysis of 45 patients (56% with BL) revealed that the
bicin, and dexamethasone were alternated with addition of rituximab appeared to be safe, and that the
methotrexate and cytarabine for a total of eight cycles. 3-year EFS and OS were both 95%. Preliminary results
The results in 26 adults with B-ALL were favorable, with the NCI dose-adjusted pharmacodynamic-based
with an EFS of 73% at 5 years. Based on these results in infusional regimen with etoposide, prednisone, vincris-
B-ALL, this regimen is also used for BL. tine, cyclophosphamide, doxorubicin, and rituximab
In an effort to improve the outcome of BL, inves- (DA-EPOCH-R) suggest that optimizing the schedule
tigators have explored the role of autologous stem may significantly improve the therapeutic index of che-
cell transplant (SCT) consolidation. One study motherapy and reduce toxicity. Preliminary results in 25
investigated the role of two courses of intensive patients (17 HIV-negative, 8 HIV-positive) with
chemotherapy, without high-dose methotrexate or untreated BL have shown a 3-year EFS and OS of 96
cytarabine, followed by BEAM (carmustine (BCNU), and 100%, respectively. Such results suggest that DA-
etoposide, cytarabine, and melphalan) and autologous EPOCH-R may be an excellent and well-tolerated alter-
SCT (Table 11.3). In a multiple center study of 27 native to dose-intense regimens, especially for HIV-
adult patients with BL without CNS or extensive bone positive patients and older adults with BL where toxicity
marrow involvement, the EFS was 73% at 5 years. is a major concern. Multiple new targeted agents are also
While these results are encouraging, they are not currently under development and may have utility in
significantly different from other regimens, and BL. However, these agents are in early clinical
question the added benefit of autologous SCT over development.
current approaches.
While current treatments are effective in BL, Management of relapse in Burkitt
improvements are needed in patients in advanced-
stage disease. Furthermore, toxicity of the regimens for lymphoma
advanced-stage disease is high, especially in older adults, The salvage treatment of relapse or refractory BL is
and treatment-related mortality is excessive. Because all usually unsuccessful. Early-stage patients who relapse
BL regimens are very aggressive, there are no specific after limited treatment, however, may still be curable,
recommendations regarding which ones to use in adults. whereas advanced-stage patients who have received
However, the treating physician must be cognizant of intensive treatment or are primary induction failures
their toxicity, particularly in patients over 50 years of are rarely cured. A COG prospective trial evaluating
age. Hence, new strategies are needed to improve the the use of rituximab, ifosfamide, carboplatin, and eto-
therapeutic index of treatment and to increase efficacy. poside in pediatric relapsed/refractory BL and other
The success of rituximab in DLBCL has prompted its mature B-cell lymphomas was recently conducted.
testing in BL. Rituximab has been incorporated in the Twenty patients with DLBCL and BL were treated
hyper-CVAD regimen with a 3-year EFS of 80% and OS with an objective response rate for the entire group
of 89% in 31 patients, most of whom had advanced- of 60% with a median follow-up of 10.5 months. The
stage disease. In this study, favorable outcomes were CR and PR rates for the 14 patients with BL and B-ALL
seen even in the elderly subgroup (age >60 years), were 29% (four patients) and 36% (five patients),
which represented approximately 30% of the study pop- respectively. A review from the European Group for
ulation. Similarly, rituximab is being studied in pediatric Blood and Marrow Transplantation (EBMT) by
BL. A phase II window study evaluated the activity and Sweetenham et al. reported an OS of 37% at 3 years
safety of single agent rituximab in children <19 years of for patients with chemosensitive relapse, but only 7%
age with newly diagnosed BL and other mature B-cell for those with resistant disease following autologous
NHLs. Of evaluable patients with BL, 27 of 67 (40%) SCT. An analysis of allogeneic transplantation in BL
were considered “responders” to rituximab alone did not find any relationship between survival and
(defined as ≥25% decrease of at least one lesion or blasts graft-versus-host disease, suggesting no graft-versus-
and no disease progression at other sites 5 days after tumor effect, and the OS was lower than patients
receiving rituximab and prior to receiving systemic treated with autologous transplant. 181
Chapter 11: Burkitt and lymphoblastic lymphoma
HIV-associated Burkitt lymphoma regimens are considered too toxic to warrant their gen-
eral use in HIV-positive patients, and have led to the use
BL comprises 13–18% of HIV-associated lymphomas,
of CHOP-based treatment. The excellent results
a significant decrease since the development of highly
achieved with the DA-EPOCH-R regimen in HIV-
active antiretroviral therapy (HAART). Clinically, BL
associated lymphoma overall and BL in HIV-negative
in HIV-infected patients is similar to sporadic BL and
patients suggest it may be an excellent regimen in HIV-
typically occurs before the development of severe
positive patients because of its relatively low toxicity.
immunodeficiency. Unlike other HIV-associated lym-
Preliminary results from a trial at the NCI of DA-
phomas, there has been no improvement in the out-
EPOCH-R in 25 patients (eight HIV-positive) with
come of BL since the development of HAART. A
untreated BL have shown a 3-year EFS and OS of 96
retrospective analysis of HIV-associated BL diagnosed
and 100%, respectively. Similar studies conducted by
between 1982 and 2003 from the University of
other groups have revealed tolerability and efficacy, but
Southern California showed no change in median
with increased infection-related deaths in patients with
survivals of 6.4 and 5.7 months, respectively, in the
CD4 counts <50/μL.
pre- and post-HAART eras. This contrasts with the
outcome of DLBCL during these periods, when
the median survival improved from 8.3 to 43.2 Treatment recommendations in
months, respectively. This improvement is likely
because of a more favorable DLBCL pathobiology
Burkitt lyphoma
that occurs at higher CD4 cell counts in the HAART Multiple regimens, as discussed above, have shown
era. This does not seem to be the case with BL, which equivalent efficacy in BL and may be used. Current
occurs at a high median CD4 cell count and does not treatment plans should include risk-adaptive strategies
have such a favorable pathobiology. and reduced CNS prophylaxis to minimize toxicity.
The relatively poor outcome of HIV-associated BL Adults may be effectively treated with pediatric regi-
is likely due to the use of CHOP-based regimens, which mens, although the CODOX-M/IVAC, CALGB and
are known to have a poor outcome in this disease. In hyper-CVAD regimens have been specifically investi-
contrast, dose-intense regimens like hyper-CVAD have gated in adults and shown to be effective and relatively
shown encouraging results in HIV-associated BL, with well tolerated. Treatment with DA-EPOCH-R is prom-
92% complete remission, but are quite toxic. The LMB ising in both HIV-negative and HIV-positive patients.
86 regimen was prospectively evaluated in the treatment
of 63 HIV-infected patients with stage IV BL, with a Treatment principles of
CR rate of 70% and 2-year OS of 47%. Like other
intensive therapies, this was associated with significant lymphoblastic lymphoma
treatment-related toxicities, including pancytopenia Therapy consists predominantly of combination che-
and deaths from neutropenia-associated sepsis. motherapy. The most commonly employed regimens
Patients with higher CD4 counts and better perform- include the following phases: induction, consolidation,
ance status tolerated this therapy better and had CNS treatment, and maintenance. Induction with ≥4
improved outcomes. Until recently, the use of rituxi- drugs is often given as a 28-day course, although there
mab in HIV-related malignancies has been somewhat are alternative approaches. A variety of consolidation,
controversial. However, recent prospective trials have also known as intensification, blocks have been shown
demonstrated both safety and efficacy, with the incor- to improve long-term outcome for high-risk pediatric
poration of rituximab in HIV-related BL. A prospective subgroups, including those with slow or incomplete
trial of rituximab with a GMALL chemotherapy back- responses to induction. All patients with LBL require
bone in 19 patients with HIV-related BL showed higher CNS-directed therapy to decrease the risk of menin-
incidences of grade 3–4 mucositis and more severe geal relapse, and a prolonged maintenance phase.
infectious episodes and outcomes comparable to the Shortening therapy to <2 years appears to compromise
non-HIV cohort, with a 2-year OS of 82% and disease outcome for both adults and children.
free survival (DFS) of 93%. These studies highlight the Most patients with LBL present with advanced-
necessity to balance treatment efficacy and toxicity in stage (III/IV) disease. Although less intense treatment
patients with HIV-associated NHL. Dose-intense BL is commonly employed for patients with low-stage
182
Chapter 11: Burkitt and lymphoblastic lymphoma
(I/II) LBL, randomized clinical trials confirm the need chemotherapy in comparison to the most effective
to treat with aggressive, prolonged ALL-type therapy. pediatric BFM regimen.
For example, an early Pediatric Oncology Group A number of specific chemotherapy agents, doses,
(POG) trial demonstrated an advantage of the addition and schedules are particularly effective against LBL/T-
of a 24-week maintenance phase for pediatric patients ALL. Increasing the intensity of induction above and
with low-stage disease. Reducing therapy for low-stage beyond the standard four-drug regimen (corticoste-
LBL increases the risk of relapse, although salvage rates roids, vincristine, anthracyclines, asparaginase) by the
are better for such patients in comparison to those addition of other agents (e.g. cyclophosphamide)
who relapse after more intensive regimens. No appa- appears to improve CR rates in children and adults.
rent difference in outcome has been seen within low- High doses of methotrexate are required to achieve
stage (i.e. I versus II) or advanced-stage (i.e. III versus adequate intracellular concentrations of methotrexate
IV) LBL in pediatric series. Bone marrow involvement polyglutamates in T-lymphoblasts. Early French stud-
was associated with poor outcome in the LMT 89 trial ies (LMT 81) added high-dose methotrexate to the
in adults. However, subgroup numbers are small in all LSA2L2 regimen with improved outcome (80% EFS).
series, preventing firm conclusions. Rare individuals The value of intensive L-asparaginase in LBL/T-ALL
with B-precursor phenotype LBL are treated according has been demonstrated in POG and Dana Farber
to B-precursor ALL-specific regimens, which are not Cancer Institute (DFCI) cooperative group studies.
considered in detail in this chapter. High-dose cytarabine has also been shown to have
specific activity against T-NHL/ALL. Results of series
with the best outcomes are summarized in Table 11.5.
Treatment of lymphoblastic lymphoma Currently, treatment can be expected to result in EFS
Sequential pediatric cooperative group trials in the rates of approximately 80–90% for children and
setting of LBL/T-ALL have led to the development of 50–60% for adults. Hyper-CVAD and variants have
highly effective specific regimens (Table 11.5). Initial also been evaluated in a predominantly adult group
studies in the 1970s demonstrated the advantage of with LBL. Forty-nine patients (median age 31, range
ALL-type therapy over CHOP-like regimens. In a 17–59), with de novo LBL received hyper-CVAD-
trial conducted by the Children’s Cancer Group based regimens, CNS prophylaxis, radiation therapy
(CCG) in North America, the Memorial Sloan as indicated for bulky mediastinal disease, and main-
Kettering Cancer Center (MSKCC) LSA2L2 regimen tenance therapy. A CR rate of 96% was seen in 46
was shown to be superior to COMP (cyclophospha- evaluable patients with 5-year DFS and OS of 72%
mide, vincristine, methotrexate, and prednisone), and 68%, respectively.
with 64% versus 35% EFS, respectively. An early To maximize therapeutic benefits, and in the
study conducted by the BFM group also demonstra- absence of prohibitive toxicity, attempts should be
ted efficacy of ALL-type therapy for children with made to deliver specified doses of all prescribed chemo-
LBL (70% EFS). Serial BFM trials have led to steady therapy agents. Furthermore, doses of methotrexate
improvements, with EFS exceeding 90% on the most and 6-mercaptopurine (6-MP) during maintenance
current pediatric protocols. The BFM regimen “back- should be dose-escalated to achieve a targeted degree
bone” has been incorporated into North American of myelosuppression. In the event of significant
studies through the CCG and COG. chemotherapy-related toxicity, specific agents should
Clinical trials have also demonstrated the advant- be dose-reduced or discontinued as clinically indicated.
age of ALL-type therapy for adults with LBL. The Those individuals with thiopurine S-methyltransferase
pediatric BFM regimen has been applied successfully deficiency (~10% incidence) require dose reduction of
to adult patients in serial trials of the GMALL. 6-MP to avoid toxicity. Results from EORTC 58951, a
Investigators at the MD Anderson Cancer Center randomized phase III trial with a BFM backbone for
have achieved good results with the hyper-CVAD reg- pediatric patients with LBL and ALL, demonstrated
imen. However, although remission induction rates improved OS and DFS in patients who were treated
on these trials exceed 90%, relapse-free survival rates with pulses of vincristine and corticosteroids during
(44–62%) are lower than in pediatric series. In the maintenance.
case of GMALL studies, this is likely due at least in Recent studies have been designed to eliminate local
part to the use of less intensive and shorter duration radiation to bulky sites of disease. A randomized trial 183
Table 11.5 Selected regimens for lymphoblastic lymphoma.
Pediatric trials
Regimen Treatment Patients (no.) Outcomes by stage
All I/II III/IV III IV
BFM 86 7-drug induction, 54 83% EFS (7-year)
consolidation, reinduction,
maintenance; 24 months
BFM 90 7-drug induction, 105 90% EFS (5-year) 90% EFS (5-year) 90% EFS 95% EFS (5-year)
consolidation, reinduction, (5-year)
maintenance; 24 months
BFM 95 7-drug induction, 156 82% EFS
consolidation, reinduction, 88% DFS 85% OS
maintenance; 24 months (5-year)
CCG 123 – New 4–6 drug induction, 371 67% EFS (7-year)
York I and BFM consolidation, reinduction,
arms maintenance; 24 months
COG A5971 4–5 drug induction, 254 80–84%
consolidation EFS (5-year)
DFCI 87/91/95 4–5-drug induction, 15 93% OS 87% EFS
consolidation, maintenance; 24 71% CCR (5-year)
months (5-year)
EORTC 58881 4-drug induction, 119 86% OS Stage I: 86% OS 84% OS 88% OS
consolidation, reinduction, 78% EFS (6-year) 86% EFS 78% EFS 77% EFS (6-year)
maintenance; 24 months Stage II: (6-year)
100% OS
64% EFS
(6-year)
EORTC 58951 4-drug induction, 411 94% OS 91% DFS
consolidation, reinduction, 83% DFS (6-year)
maintenance; 24 months (6-year) intervention intervention
arm arm
LMT 81 5-drug induction, 82 76% OS 73% EFS 79% EFS 72% EFS (5-year)
consolidation, maintenance; 24 75% EFS (5-year) (5-year) (5-year)
months
LSA2L2 4-drug induction, 95 79% OS 87% EFS 87% 90% OS IVA: 79% OS;
consolidation, maintenance; 24– 75% EFS (5-year) OS (5-year) 85% EFS 73% EFS (5-year)
36 months (5-year) IVB: 74% OS
70% EFS (5-year)
POG 8704 - 6-drug induction, 84 78% CCR (4-year)
Intensive consolidation, maintenance; 24
L-asparaginase months
arm
NHL 13 7-drug induction, 41 90% OS
consolidation, maintenance; 24 83% EFS
months (5-year)
Adult trials
GMALL 89/93 7-drug induction, 45 51% OS 56% OS (7-year) 48% OS (7-year) 49% OS (7-year)
consolidation, reinduction, 62% DFS (7-year)
maintenance; 6–12 months
Hyper-CVAD 4-drug induction, 26 62% PFS (3-year)
consolidation, repeated × 4,
maintenance; 24–36 months
LMT 89 5-drug induction, 27 63% OS 85% OS 37% OS
consolidation, maintenance; 12– 44% PFS (5-year) 67% PFS 21% PFS (5-year)
24 months (5-year)
CCR, Continuous complete remission; DFS, disease-free survival; EFS, event-free survival; OS, overall survival; PFS, progression-free survival.
Chapter 11: Burkitt and lymphoblastic lymphoma
conducted by the POG failed to show benefit for The 5-year DFS was 67% for 36 patients who received
involved field radiation for pediatric patients with low- SCT versus 42% for the 120 patients treated with
stage disease. Mediastinal radiation was eliminated chemotherapy alone, with OS rates of 67% and 47%,
from the treatment of advanced-stage disease on the respectively. This study demonstrates that, for high-
pediatric BFM 90 study without an apparent increase risk T-cell ALL, SCT in CR1 results in excellent out-
in the local recurrence rate (7%). Notably, the media- comes. In light of the excellent outcome for children
stinum was the primary site of relapse (17%) in adults and adolescents with LBL with current chemotherapy
treated on recent GMALL trials despite the fact that regimens, allogeneic SCT is rarely used for pediatric
most had received 2400 cGy of local radiation, suggest- patients except for those with poor response to or
ing that chemotherapy is the critical determinant of relapse after standard treatment.
bulky site disease control. In contrast, adjuvant radia-
tion did appear to decrease the risk of mediastinal CNS treatment of lymphoblastic
recurrence for adults treated with the hyper-CVAD
regimen. Although males with testicular ALL/LBL lymphoma
have historically been treated with radiation, recent LBL has a high rate of CNS involvement compared to
results with regimens containing intermediate- or other subtypes of lymphoma and leukemia.
high-dose methotrexate suggest that radiation may not Approximately 5–15% of patients have meningeal dis-
be needed. Thus, in the absence of a protocol-specified ease at presentation; additionally, recent studies from
role, radiation should be reserved for emergency man- COG A5971 suggest that these patients have the worst
agement of life- or organ-threatening complications. prognosis. CNS-directed prophylaxis is required for all
Even in that setting, however, corticosteroids are com- patients with LBL because of the high risk of CNS
monly adequate to induce rapid disease reduction. relapse, even in those without overt meningeal involve-
There have been few studies designed to evaluate the ment. Historically, craniospinal irradiation has been
relative efficacy of high-dose therapy with autologous relied upon for CNS prophylaxis for patients with
SCT for LBL. Most published series represent single- advanced-stage LBL/T-ALL. To reduce the neurocogni-
institution or registry studies that often include mixed tive dysfunction encountered in survivors of childhood
lymphoma subtypes and/or patients with high-risk fea- ALL, regimens that limit radiation exposure have been
tures. The European Group for Blood and Bone designed over the past two decades. Initial trials led to
Marrow Transplantation and the United Kingdom reductions in the cranial radiation dose and replacement
Lymphoma Group conducted a randomized trial of of spinal irradiation with intensive intrathecal chemo-
autologous SCT versus conventional chemotherapy as therapy. More recently, cranial radiation has been elim-
post-remission treatment for adults with LBL. Although inated in all but the highest risk patients. Thus, CNS
there was a trend towards improved relapse-free sur- prophylaxis for patients with ALL now consists primar-
vival on the autologous SCT arm, no OS benefit was ily of intrathecal chemotherapy in combination with
demonstrated. Thus, there is no current evidence to systemic agents that have good CNS penetration, most
suggest that autologous SCT is more effective than notably dexamethasone and high-dose methotrexate. It
standard LBL-specific chemotherapy regimens. must be emphasized that most of these data are derived
Studies of allogeneic SCT are also limited in the from patients with B-precursor ALL. However, initial
setting of LBL. Most prospective trials compare allo- results from EORTC 58881, BFM 95, and COG A5971
geneic to autologous stem cell rescue. In general, suggest that this approach may be applied effectively to
relapse rates are lower after allogeneic SCT; however, LBL/T-ALL. EORTC 58951 incorporated the use of tri-
much of this advantage is offset by increased ple intrathecal therapy without any craniospinal irradi-
treatment-related mortality. Consequently, allogeneic ation and no CNS relapses have occurred to date. Other
SCT is usually reserved for those with specific high- more recent studies, including the St. Jude NHL 13 trial,
risk features. A recent prospective trial based on the also eliminated prophylactic cranial irradiation and had
BFM 90 and BFM 95 regimens employed matched excellent outcomes and OS comparable to other regi-
sibling donor SCT in CR1 for pediatric patients with mens. Nonetheless, follow-up remains relatively short
high-risk T-cell ALL. High-risk was defined as a poor for radiation-free regimens and cranial irradiation is
response to prednisone, non-response on day 33, still considered the standard treatment for individuals
186 or chromosomal translocations t(9;22) or t(4;11). with active meningeal leukemia. It is notable that in the
Chapter 11: Burkitt and lymphoblastic lymphoma
BFM 95 study, the elimination of cranial radiation pro- Treatment recommendations for
phylaxis was associated with an increase in bone marrow
relapse, the cause of which is not completely understood. lymphoblastic lymphoma
Finally, in light of both the poorer overall outcome and Therapy should be instituted as soon as possible after
the lower risk of neurologic toxicity in older individuals, diagnosis and should be directed by practitioners
cranial irradiation is still commonly employed for adults experienced with the specific regimen and the manage-
with ALL and LBL. ment of expected complications. It is recommended
A number of practical aspects of intrathecal that patients be treated on clinical trials whenever
administration are worthy of comment. To minimize possible. The best reported outcome for LBL is with
the risk of meningeal contamination from traumatic aggressive ALL-type regimens (Table 11.5), and recent
lumbar puncture, spinal taps should be performed by BFM regimens appear superior. In general, treatment
experienced practitioners. Furthermore, it is recom- is similar for adult and pediatric patients, although, as
mended that intrathecal chemotherapy be adminis- detailed above, therapy is often stratified based on risk
tered at the time of the initial diagnostic lumbar of relapse and organ toxicity.
puncture, especially when there are circulating blasts
in the peripheral blood. To improve chemotherapy
distribution, the volume of cerebrospinal fluid (CSF) Supportive care of Burkitt and
removed should equal the volume administered and
patients should remain prone for approximately 30
lymphoblastic lymphoma
minutes after administration. The use of an Ommaya Aggressive supportive care and surveillance for com-
reservoir may improve delivery to the ventricles, plications during and after treatment is critical to
although this approach is not routinely employed. successful outcome.
Finally, intrathecal chemotherapy should be dosed Antiemetics: routine prophylaxis and treatment of
according to age owing to age-related CSF volume nausea and vomiting are required during induction,
changes. consolidation, and CNS-directed therapy phases.
Tumor lysis syndrome: as LBL and BL frequently
present with a large tumor burden and cell turnover is
Management of relapse in lymphoblastic extremely rapid, tumor lysis syndrome is commonly
encountered during initiation of therapy. Some regi-
lymphoma mens include a “pre-phase” of treatment to allow grad-
In general, the outcome after relapse is poor for chil- ual induction, for example with a few day courses of
dren and adults with LBL. In a retrospective analysis of corticosteroids alone. Tumor lysis prophylaxis should
pediatric patients with LBL treated by the BFM be started as soon as possible after diagnosis and at least
between 1990 and 2003, the salvage rate for those 12 hours prior to the initiation of induction chemo-
with progressive disease or relapse was poor, with a therapy. Prophylaxis and monitoring should continue
14% OS rate. Long-term survival was only achieved in until disease burden is reduced and it is apparent that
those few patients who were able to undergo an allo- no complications of lysis have developed (usually 3–7
geneic SCT. Those patients with low-stage disease who days). The following is recommended:
relapse after abbreviated therapy have higher salvage
rates. Allogeneic SCT is recommended for patients * Allopurinol: 100 mg/m2/dose PO TID. Urate
with HLA-matched related donors. In addition, oxidase (Rasburicase) is an alternative for
matched unrelated donor transplantation should be management of extreme hyperuricemia.
considered for young individuals. Agents targeted to * Hydration: IV fluids at a rate of ≥2 times
T-cell malignancies hold promise. Nelarabine (com- maintenance (≥120 ml/m2/hour) adjusted to
pound 506U78), a water-soluble prodrug of ara-G, maintain urine specific gravity ≤1.010 and normal
which was approved by the FDA for treatment of urine output. Because of the risk of hyperkalemia,
relapsed T-LBL/ALL in 2005, has single agent activity potassium should be avoided.
in pediatric and adult patients with relapsed/refractory * Serum potassium, phosphorus, calcium, creatinine,
LBL/T-ALL. An objective response rate of 33% was blood urea nitrogen (BUN), and uric acid should be
observed in pediatric patients (CR 26%; PR 7%) and a assayed every 4–6 hours for the first 24–48 hours,
31% CR rate in adults. then less frequently once stable. 187
Chapter 11: Burkitt and lymphoblastic lymphoma
Transfusions: blood transfusion therapy is usually * Myeloid growth factors: myeloid growth factor
required to counter severe cytopenias associated with support (e.g. G-CSF) is commonly employed as
induction and consolidation phases of therapy. part of treatment of BL in both children and adults.
Specialized blood products should be employed in However, the benefits of such have not been proved
attempt to decrease the risk of transfusion-associated in the setting of LBL therapy.
complications. Chemotherapy-prophylaxis: agent-specific prophylaxis
* Platelets: to prevent bleeding, platelet counts should be utilized as clinically indicated (e.g. gastritis
should routinely be maintained above 10 000/μL. prophylaxis during corticosteroids).
Higher levels may be needed to manage active Nutritional support: nutritional status should be
bleeding, prior to invasive procedures, and to monitored and supplementation provided as needed.
reduce the risk of leukostasis-induced CNS Routine folic acid use should be avoided around the
hemorrhage in the setting of hyperleukocytosis. time of methotrexate administration as this may coun-
Single donor platelets are recommended whenever teract the therapeutic efficacy of folate antagonism.
possible to decrease donor exposure and the risk of In contrast, leucovorin rescue is required after inter-
HLA-alloimmunization. mediate- and high-dose methotrexate.
* Red blood cells: concomitant anemia often partially Response and toxicity evaluations: serial monitor-
offsets the hyperviscosity associated with severe ing for response- and therapy-associated toxicity
hyperleukocytosis. Thus, red blood cell transfusion should be conducted during and after treatment.
should be avoided if possible when the white blood Life-long follow-up for possible late effects is indicated
count is >100 000/μL. If red cell transfusion is for long-term survivors, including the following:
needed, the hemoglobin should be increased slowly * Cardiomyopathy: to decrease the risk of
using small aliquots of packed red cells until the cardiotoxicity, cumulative anthracycline doses are
peripheral blast count is reduced. usually limited to <400 mg/m2. Left ventricular
* Irradiation: all cellular blood products should be function should be monitored.
irradiated to prevent transfusion-associated graft- * Neurologic toxicity: children are at high risk of
versus-host disease. neurotoxicity from CNS-directed therapy and
* Leukodepletion: platelets and red cells should be neurodevelopmental assessment is required.
leukocyte-reduced to decrease the risk of febrile * Endocrinologic dysfunction: patients should be
reactions and platelet-refractoriness due to HLA- monitored for endocrinopathies, including growth
alloimmunization. retardation, infertility, and hormone deficiencies.
Infectious prophylaxis: aggressive surveillance, prophy- * Osteonecrosis: corticosteroids are associated with a
laxis, and treatment for bacterial, fungal, viral, and high incidence of osteonecrosis.
opportunistic infections is required throughout * Secondary malignancy: long-term survivors are at
therapy. risk of secondary malignancies even beyond the
first decade after treatment.
* Pneumocystis jiroveci pneumonia (PCP): all
patients should receive prophylaxis against PCP
until approximately 6 months after completion of Further reading
chemotherapy. Amylon MD, Shuster J, Pullen J, et al. Intensive high-dose
* Neutropenic fever: patients with fever in the setting asparaginase consolidation improves survival for
of an absolute neutrophil count (ANC) <500/μL pediatric patients with T cell acute lymphoblastic
require emergent evaluation and management for leukemia and advanced stage lymphoblastic lymphoma: a
possible infection. Immediate, empiric, broad- Pediatric Oncology Group study. Leukemia,
spectrum, intravenous antibiotics are indicated. 1999;13:335–342.
Antifungal therapy should be initiated for Burkhardt B, Woessmann W, Zimmermann M, et al.
persistent neutropenic fever beyond 5–7 days. In Impact of cranial radiotherapy on central nervous
system prophylaxis in children and adolescents
general, antibiotics should be continued until the
with central nervous system-negative stage III or
ANC rises to >500/μL, fever resolves, cultures are IV lymphoblastic lymphoma. J Clin Oncol,
188 negative, and any infection is fully treated. 2006;24:491–499.
Chapter 11: Burkitt and lymphoblastic lymphoma
Dave SS, Fu K, Wright GW, et al. Molecular diagnosis Kingdom Lymphoma Group LY06 study. Ann Oncol,
of Burkitt’s lymphoma. N Engl J Med, 2002;13:1264–1274.
2006;354:2431–2442. Millot F, Suciu S, Philippe N, et al. Value of high-dose
Divine M, Casassus P, Koscielny S, et al. Burkitt lymphoma cytarabine during interval therapy of a Berlin–Frankfurt–
in adults: a prospective study of 72 patients treated with Munster-based protocol in increased-risk children with
an adapted pediatric LMB protocol. Ann Oncol, acute lymphoblastic leukemia and
2005;16:1928–1935. lymphoblastic lymphoma: results of the European
Gaynon PS, Trigg ME, Heerema NA, et al. Children’s Cancer Organization for Research and Treatment of Cancer
Group trials in childhood acute lymphoblastic leukemia: 58881 randomized phase III trial. J Clin Oncol,
1983–1995. Leukemia, 2000;14:2223–2233. 2001;19:1935–1942.
Goldberg JM, Silverman LB, Levy DE, et al. Childhood T-cell Mora J, Filippa DA, Qin J, Wollner N. Lymphoblastic
acute lymphoblastic leukemia: the Dana-Farber Cancer lymphoma of childhood and the LSA2-L2
Institute acute lymphoblastic leukemia consortium protocol: the 30-year experience at Memorial-
experience. J Clin Oncol, 2003;21:3616–3622. Sloan-Kettering Cancer Center. Cancer,
2003;98:1283–1291.
Goldman S, Smith L, Anderson JR, et al. Rituximab and
FAB/LMB96 chemotherapy in children with Stage III/IV Mullighan CG, Downing JR. Global genomic
B-cell non-Hodgkin lymphoma: a Children’s Oncology characterization of acute lymphoblastic leukemia.
Group report. Leukemia, 2013;27:1174–1177. Semin Hematol, 2009;46:3–15
Hoelzer D, Ludwig WD, Thiel E, et al. Improved outcome in Murphy SB, Bowman WP, Abromowitch M, et al. Results of
adult B-cell acute lymphoblastic leukemia. Blood, treatment of advanced-stage Burkitt’s lymphoma and B
1996;87:495–508. cell (SIg+) acute lymphoblastic leukemia with high-dose
fractionated cyclophosphamide and coordinated high-
Hoelzer D, Gokbuget N. Treatment of lymphoblastic dose methotrexate and cytarabine. J Clin Oncol,
lymphoma in adults. Best Pract Res Clin Haematol, 1986;4:1732–1739.
2002;15:713–728.
Patte C, Kalifa C, Flamant F, et al. Results of the LMT81
Kluin PM, Harris NL, Stein H, et al. B-cell lymphoma, protocol, a modified LSA2L2 protocol with high dose
unclassifiable, with features intermediate between diffuse methotrexate, on 84 children with non-B-cell
large B-cell lymphoma and Burkitt lymphoma. In (lymphoblastic) lymphoma. Med Pediatr Oncol,
Swerdlow SH, Campo E, Harris NL, et al. (eds). WHO 1992;20:105–113.
Classification of Tumours of Haemopoietic and Lymphoid
Tissues. Lyon: IARC, 2008;265–266. Patte C, Auperin A, Michon J, et al. The Societe Francaise
d’Oncologie Pediatrique LMB89 protocol: highly
Hummel M, Bentink S, Berger H, et al. A biologic definition effective multiagent chemotherapy tailored to the tumor
of Burkitt’s lymphoma from transcriptional and genomic burden and initial response in 561 unselected children
profiling. N Engl J Med, 2006;354:2419–2430. with B-cell lymphomas and L3 leukemia. Blood,
Jabbour E, Koscielny S, Sebban C, et al. High survival rate 2001;97:3370–3379.
with the LMT-89 regimen in lymphoblastic lymphoma Pilozzi E, Pulford K, Jones M et al. Co-expression of CD79a
(LL), but not in T-cell acute lymphoblastic leukemia (JCB117) and CD3 by lymphoblast lymphoma. J Pathol,
(T-ALL). Leukemia, 2006;20:814–819. 1998;186:140–143.
Jaffe ES, Harris NL, Stein H, Vardiman JW. World Health Reiter A, Schrappe M, Ludwig WD, et al. Intensive
Organization Classification of Tumours. Pathology and ALL-type therapy without local radiotherapy provides a
Genetics of Tumours of Hematopoietic and Lymphoid 90% event-free survival for children with T-cell
Tissues. Lyon: IARC Press, 2001. lymphoblastic lymphoma: a BFM group report. Blood,
Leoncini L, Raphael M, Stein H, et al. Burkitt lymphoma. In 2000;95:416–421.
Swerdlow SH, Campo E, Harris NL, et al. (eds). WHO Reiter A, Schrappe M, Parwaresch R, et al.
Classification of Tumours of Haemopoietic and Lymphoid Non-Hodgkin’s lymphomas of childhood and adolescence:
Tissues. Lyon: IARC, 2008; 262–264. results of a treatment stratified for biologic subtypes
Magrath I, Adde M, Shad A, et al. Adults and children with and stage: a report of the Berlin–Frankfurt–Munster Group. J
small non-cleaved-cell lymphoma have a similar excellent Clin Oncol, 1995;13:359–372.
outcome when treated with the same chemotherapy Reiter A, Schrappe M, Tiemann M, et al. Improved
regimen. J Clin Oncol, 1996;14:925–934. treatment results in childhood B-cell neoplasms with
Mead GM, Sydes MR, Walewski J, et al. An international tailored intensification of therapy: a report of the Berlin–
evaluation of CODOX-M and CODOX-M alternating Frankfurt–Munster Group Trial NHL-BFM 90. Blood,
with IVAC in adult Burkitt’s lymphoma: results of United 1999;94:3294–3306.
189
Chapter 11: Burkitt and lymphoblastic lymphoma
Rizzieri DA, Johnson JL, Niedzwiecki D, et al. Intensive Thomas DA, Cortes J, O’Brien S, et al. Hyper-CVAD program
chemotherapy with and without cranial radiation for in Burkitt’s-type adult acute lymphoblastic leukemia. J Clin
Burkitt leukemia and lymphoma: final results of Cancer Oncol, 1999;17:2461–2470.
and Leukemia Group B Study 9251. Cancer, Thomas DA, O’Brien S, Cortes J, et al. Outcome with the
2004;100:1438–1448. hyper-CVAD regimens in lymphoblastic lymphoma.
Schrappe M, Reiter A, Ludwig W, et al. Improved Blood, 2004;104:1624–1630.
outcome in childhood acute lymphoblastic leukemia
despite reduced use of anthracyclines and cranial Uyttebroeck AS, Suciu S, Laureys G, et al.
radiotherapy: results of trial ALL-BFM 90. Blood, Treatment of childhood T-cell lymphoblastic
2000;95:3310–3322. lymphoma according to the strategy for acute
lymphoblastic leukaemia, without radiotherapy:
Steinherz PG, Gaynon PS, Breneman JC, et al. Treatment of long term results of the EORTC CLG 58881 trial. Eur J
patients with acute lymphoblastic leukemia with bulky Cancer, 2008;44(6):840–846.
extramedullary disease and T-cell phenotype or other
poor prognostic features: randomized controlled trial van Imhoff GW, van der Holt B, MacKenzie MA, et al.
from the Children’s Cancer Group. Cancer, Short intensive sequential therapy followed by
1998;82:600–612. autologous stem cell transplantation in adult Burkitt,
Termuhlen AM, Smith LM, Perkins SL, et al. Br J Hematol. Burkitt-like and lymphoblastic lymphoma. Leukemia,
2013;ePub ahead of print. 2005;19:945–952.
190
Chapter
12
Therapy of diffuse large B-cell lymphoma
John W. Sweetenham
Pathology: Andrew Wotherspoon Molecular cytogenetics: Andreas Rosenwald and German Ott
Lymphoma, Second Edition, ed. Robert Marcus, John W. Sweetenham, and Michael E. Williams. Published by 191
Cambridge University Press. © Cambridge University Press 2014.
Chapter 12: Diffuse large B-cell lymphoma
Figure 12.1 Diffuse large B-cell lymphoma. A sheet of large Figure 12.2 Diffuse large B-cell lymphoma. A sheet of large cells
lymphoid cells with pale cytoplasm and large pleomorphic nuclei with abundant pale eosinophilic cytoplasm and nuclei that show
containing prominent nucleoli. marked lobulation.
nucleolus. The cells of this variant of DLBCL have T-cell/histiocyte-rich B-cell lymphoma
abundant cytoplasm. Their nuclei are pleomorphic
and may resemble Reed–Sternberg cells or cells of ana- (TCHRBCL)
plastic T-cell lymphoma. Multinucleate giant cells may In this variant the neoplastic B-cells form a small per-
be seen. There may be a sinusoidal pattern of infiltration centage (usually <10%) of the cellular population while
within nodes. the majority of the cells are T-lymphocytes and/or
Characteristically, DLBCL express CD45 with the histiocytes. The large cells may have a variety of appear-
pan-B-cell markers CD19, CD20, CD22, CD79a, and ances, including centroblastic or immunoblastic, and in
PAX5. Monotypic surface and/or cytoplasmic immu- some cases they may resemble Reed–Sternberg cells or
noglobulin is demonstrable in a majority of cases. the cells of lymphocyte-predominant Hodgkin lym-
Combinations of staining for CD10, BCL-6, and phoma. Small B-cells are essentially absent. The cells
MUM1 have been used as an immunocytochemical express the pan-B-cell markers (CD19, CD20, CD22,
surrogate for the identification of germinal center- CD79a, and PAX5). Staining for BCL-6 is frequently
like phenotype (CD10+ or CD10–, BCL-6+ and seen but the cells are usually negative for CD10.
MUM1–) as opposed to non-germinal center (acti- Staining for BCL-2 is seen in a proportion of cases.
vated) B-cell-like DLBCL recapitulating the separa- Infrequently there is expression of CD30 but the cells
tion seen by studies using microarray techniques. The lack CD15 and there is expression of epithelial mem-
separation of germinal center-like and activated brane antigen (EMA) in about one-third of cases.
B-cell-like categories can be enhanced by the addition These cases need to be distinguished from classical
of staining for FOXP1 and GCET. Expression of and lymphocyte-predominant Hodgkin lymphoma.
BCL-2 protein is present in 30–50%. A proportion Immunophenotypic studies help distinguish these
of cases express CD5 but these are negative for cyclin cases from classical Hodgkin lymphoma, as the latter
D1. Expression of CD30 is seen in occasional cases express CD30 and CD15 and lack the full B-cell-
with non-anaplastic morphology. Lymphomas with related antigen profile, while TCHRBCL express
anaplastic morphology stain for CD30 but are CD20 with CD79a with staining for OCT2 and
negative for CD15 and ALK1, distinguishing them BOB1 and they are positive for CD45 with staining
from classical Hodgkin and anaplastic large cell for EMA in a proportion of cases. The presence of
lymphoma. Staining for CD23 may also be seen in nodular areas with large cells within areas containing
some cases (up to 16% in some series). Expression small B-cells and follicular dendritic cell meshworks
of CD138 may be seen in cases showing plasmablastic raises the possibility of lymphocyte-predominant
192 morphology but is rarely seen in other cases. Hodgkin lymphoma.
Chapter 12: Diffuse large B-cell lymphoma
in their global gene expression profiles, additional treatment regimens, specifically R-CHOP, are used.
molecular features of the tumors point to a profound Moreover, the composition of the microenvironment
biological difference between GCB and ABC DLBCL. (“stromal signature”) appears to be of prognostic
In particular, as mentioned above, translocations of importance. In the prognostically less favorable ABC
the BCL-2 oncogene are strongly associated with the DLBCL subtype, activating or inactivating mutations of
GCB-type of DLBCL, as are amplifications of a chro- genes involved in the NF-κB pathway (A20, CARD11,
mosomal region in 2p that harbors the REL and BCL- TRAF2, TRAF5, RANK) have been described in a sig-
11A loci. Additional genetic differences between GCB nificant proportion of cases. Additional mutations were
and ABC DLBCL include frequent chromosomal gains discovered in B-cell receptor-associated molecules
of 12q in GCB DLBCL, whereas chromosomal gains/ (CD79B), leading to “chronic active” B-cell receptor
amplifications of 3q and 18q are predominantly signaling and in MYD88, an adaptor protein involved
detected in the ABC DLBCL subtype. In contrast to in toll/interleukin-1 receptor signaling.
GCB DLBCL, ABC DLBCL is characterized by con-
stitutive activation of the oncogenic NF-κB pathway. Staging investigations
Cell lines representative of this DLBCL subtype can be
Recommended staging investigations for newly pre-
killed by blocking NF-κB, suggesting that this pathway
senting patients with DLBCL are summarized in
may represent a potential new therapeutic target, e.g.
Table 12.1.
for proteasome inhibitors (PS341).
Present treatment approaches to DLBCL are still
Functional studies of GCB and ABC DLBCL cell
largely based upon anatomic stage, and, particularly in
lines have revealed additional molecular differences.
the context of clinical trials, on the risk factors
For example, phosphodiesterase 4B (PDE4B), an inac-
described by the International Prognostic Index
tivator of cyclic AMP (cAMP) which mediates apop-
(IPI), as described below. Emerging data on the poten-
tosis in B-cells via AKT inactivation, is highly
tial value of 18fluoro-deoxyglucose positron emission
expressed in ABC DLBCL. As a result, activation of
tomography (FDG–PET) as a staging and response
the cAMP pathway in GCB DLBCL cell lines leads to
evaluation technique in this disease is likely to modify
apoptotic cell death whereas there is no effect in ABC
determination of disease extent in the future. It is
DLBCL lines that have PDE4B expression. Likewise,
therefore important to emphasize that new treatment
the response to interleukin-4 stimulation appears to be
strategies may emerge as new data on the value of
different between the two subtypes.
FDG–PET appear. This subject is discussed more
The transcriptional heterogeneity of DLBCL is fur-
fully in Chapter 3.
ther highlighted by the description of additional gene
expression signatures that are variably expressed
between these tumors. Monti and colleagues, for Table 12.1 Standard staging investigations in DLBCL.
example, defined “oxidative phosphorylation,” “B-cell
receptor proliferation,” and “host response” clusters Complete history and physical examination
that show variable expression in DLBCL tumors. In Determination of performance status
the clinical setting, the gene expression profiling Complete blood count and differential white cell count
approach may also be used to predict survival of Serum biochemistry profile
DLBCL at the time of diagnosis. By combining gene
Lactate dehydrogenase
expression signatures that capture the germinal-center
phenotype, the proliferative activity, MHC class II Serum β2 microglobulin
expression, and the host response, a powerful mathe- Serum protein electrophoresis and immunofixation
matical predictor can be constructed that identifies Chest X-ray
DLBCL patients with particularly favorable or poor CT scan neck, chest, abdomen, pelvis
survival. Genetic alterations, e.g. gains of 3p, may
Bone marrow aspirate and core biopsy for routine
improve such survival predictors, indicating that histopathology, flow cytometry, cytogenetics
genetic alterations may not be generally inherent in
Other imaging techniques, including magnetic resonance
global gene expression profiles. imaging as indicated clinically
The more favorable prognosis of patients with the Potential future role for FDG–PET
GCB DLBCL phenotype is still evident when current 195
Chapter 12: Diffuse large B-cell lymphoma
with localized aggressive NHL (81% of whom had Table 12.2 Stage-adjusted International Prognostic Index.
DLBCL) who were randomized to receive either
three cycles of CHOP chemotherapy followed by IF- Adverse factor Stage-adjusted IPI
RT, or dose-intensive chemotherapy with ACVBP Stage Bulky stage II
(doxorubicin, cyclophosphamide, vindesine, bleomy- Age >60
cin, prednisone) followed by sequential consolidation LDH >Normal
therapy with methotrexate, etoposide, ifosfamide, and
Performance status ≥2
cytarabine. With a median follow-up of 7.7 years, the
5-year event-free survival (EFS) was 82% in the che- Extranodal sites Not applicable
motherapy arm, compared with 74% in the combined
modality arm (P < 0.001). The corresponding figures
for OS were 90% and 81% (P = 0.001). brief duration chemotherapy (CHOP × 3) and IF-RT.
The outcomes for patients with non-bulky disease Such combined modality therapy should therefore be
were analyzed separately, and the difference in EFS and considered the standard approach for this patient
OS in favor of the chemotherapy-only arm was main- group, since future studies are very unlikely to dem-
tained. The upper age limit for the GELA study was 61 onstrate further improvements in outcome.
years, with a median age of 46 years. By comparison, the In contrast, patients with one adverse factor have a
median age in the SWOG was 59 years, with approx- projected 5-year OS rate of around 70%; only 50–60%
imately 50% of the patients being over 60 years. It is of those with three or four adverse factors survive for 5
therefore not clear whether an unselected patient pop- years. New approaches are therefore needed for these
ulation with limited-stage disease would tolerate or groups, although combined modality therapy as
benefit from the dose-intensive chemotherapy used in described above remains the current standard
the GELA study, particularly since this regimen was treatment.
associated with a 25% hospitalization rate with each of Several studies in advanced DLBCL have now dem-
the first three cycles of ACVBP chemotherapy. onstrated improvements in DFS and OS by the addi-
Although the results of these studies suggest that tion of the anti-CD20 monoclonal antibody,
chemotherapy alone may be adequate treatment for rituximab, to combination chemotherapy such as
bulky disease, this was not confirmed by a study from CHOP. No randomized studies have been reported
the Eastern Co-operative Oncology Group (ECOG), in to date in limited-stage disease. A recent phase II
which 352 patients with clinical stage I or II disease study from the SWOG has tested the addition of ritux-
(including bulky disease) were initially treated with imab to three cycles of CHOP chemotherapy plus IF-
eight cycles of CHOP chemotherapy. Patients in com- RT for diffuse aggressive B-cell NHL. The 2-year PFS
plete response after chemotherapy were randomized was 94%. These results compared favorably with an
between 30 Gy of IF-RT or no further treatment. historical series of patients treated with CHOP × 3 plus
Patients in partial remission after chemotherapy IF-RT using identical selection criteria. Randomized
received 40 Gy of IF-RT. For the 172 randomized studies will be required to evaluate the benefit of addi-
patients, the 6-year DFS was 73% for the radiation tion of rituximab to chemotherapy in this context
therapy arm compared with 56% for the observation prospectively.
arm (P = 0.05). No OS difference was observed. The potential use of FDG–PET to determine which
The apparent differences in outcomes between the patients with early stage DLBCL can be managed with-
studies described above are, in part, related to patient out IF-RT has been investigated at the British
selection for the various studies. Since patients with Columbia Cancer Agency. Patients with early-stage
limited-stage disease represent a heterogeneous disease who are PET-negative after three cycles of
patient group, a stage-adjusted IPI has recently been R-CHOP therapy receive one further cycle, while
proposed, to facilitate risk stratification in future stud- PET-positive patients receive IF-RT. Of 47 patients
ies in early-stage disease (Table 12.2). This prognostic who were PET-negative and in whom IF-RT was omit-
model has been validated in other patient populations ted, only one has relapsed to date, suggesting that
with early-stage disease. According to this model, FDG–PET may be a useful tool for directing
patients with no adverse risk factors have a projected de-escalation of therapy, although this will require
5-year OS of approximately 95% when treated with prospective evaluation. 197
Chapter 12: Diffuse large B-cell lymphoma
Primary extranodal disease trial. Preliminary data suggests that it may reduce
the need for IF-RT, but this requires prospective
Extranodal involvement in the context of widespread
evaluation.
disease is a frequent finding in DLBCL. In addition,
* Novel approaches include the addition of
approximately 40% of all cases of DLBCL present with
radiolabeled monoclonal antibodies to standard
primary extranodal involvement. The most common
combined modality therapy. Cooperative group
site for primary extranodal disease is the stomach, but
trials testing this approach are under way at
other common sites include bone (which may be mul-
present.
tifocal), thyroid, kidney, spleen, and testis. Primary
CNS and skin DLBCL are specific entities covered
* Preliminary data suggest that FDG–PET may be
elsewhere in this book. In general, the treatment of used to determine the requirement for radiation
localized primary extranodal DLBCL is exactly as for therapy after abbreviated chemotherapy.
nodal disease – many of the randomized trials for
patients with limited-stage disease have included
patients with nodal and extranodal primaries, with Treatment of advanced-stage DLBCL
no obvious differences in outcome observed. Primary Approximately 75% of patients with DLBCL present
involvement of certain anatomic sites, notably the with bulky stage II, or stage III–IV disease, and should
testis, paranasal sinuses, and the nasopharnyx, are therefore be regarded as having advanced-stage dis-
associated with an increased risk of CNS relapse, and ease. Until recently, CHOP was considered the stand-
CNS prophylaxis is typically given to patients with ard first line therapy for all patients with DLBCL,
these primary sites of involvement. mainly on the basis of the SWOG randomized trial
which compared this regimen with three more inten-
sive regimens, showing no difference in response rates,
Early-stage DLBCL – conclusions PFS, or OS, but a higher toxicity rate in the regimens
A treatment algorithm for early-stage DLBCL is other than CHOP.
shown in Figure 12.5. In recent years, several approaches have been
investigated to improve the outcome for patients
* Combined modality therapy with brief duration with advanced-stage disease.
chemotherapy and IF-RT is the standard of care,
irrespective of risk group according to the stage-
adjusted IPI. Dose-intensified and dose-dense therapy
* The addition of rituximab to initial chemotherapy Intensification of the CHOP regimen, either by redu-
has not been formally evaluated in a randomized cing the treatment duration from 21 to 14 days, or by
IPI = 0
the addition of etoposide to the standard regimen, has The dose-adjusted EPOCH-R (etoposide, predni-
been investigated in two studies from the German non- sone, vincristine, cyclophosphamide, doxorubicin, rit-
Hodgkin’s Lymphoma Study Group. Parallel studies uximab) [DA-EPOCH-R] regimen represents another
were conducted in young and elderly patients, and in approach to intensification of chemotherapy dose in
each study the comparison of CHOP21 with CHOP14 DLBCL. This regimen requires close monitoring of
and CHOP with the same regimen plus etoposide hematologic toxicity, and reduction or increase of
(CHOEP) was performed using a 2 × 2 factorial design. chemotherapy doses according to the neutrophil
The trial in older patients included 689 patients, 71% of nadir. In the initial single institution phase II study
whom had DLBCL. In this study, CHOP14 was shown of this regimen in DLBCL, 5-year PFS and OS rates
to be significantly superior to CHOP21. The 5-year were 79% and 80%, respectively, and did not differ
event-free rate for CHOP21 was 33% compared with according to tumor proliferation rates or BCL-2
44% for CHOP14 (P = 0.003). The corresponding rates expression. This regimen is now being compared
for OS were 41% versus 53% (P < 0.001). The addition with R-CHOP in an ongoing intergroup randomized
of etoposide had no effect in this study. trial in the USA.
The corresponding study included 710 younger
patients, aged 18–60 years, with stage I–IV disease,
with a normal lactate dehydrogenase (LDH) level. No High-dose therapy and autologous stem
EFS or OS benefit was seen in this group from reduction cell transplantation as a component of
of the treatment interval from CHOP21 to CHOP14.
However, an EFS benefit was seen for patients receiving
first line therapy
CHOEP14 or 21 (5-year EFS = 69%) compared with The reported effectiveness of high-dose therapy and
those receiving CHOP14 or 21 (% year EFS = 58%; autologous stem cell transplantation (ASCT) as a sal-
P = 0.004). No OS benefit was observed, possibly due vage treatment in aggressive NHL has prompted many
to the ability to “salvage” patients relapsing after first groups to investigate the use of ASCT as a component
line therapy. of first line therapy, particularly for patients identified
The clinical relevance of these findings is unclear, as having “poor-risk” disease at presentation. Many
and they must be evaluated in the context of trials are now published, although comparison of these
rituximab-containing therapy. studies is complicated by the variable inclusion crite-
Based on the results of these studies, the German ria, and variable definition of poor-risk disease. Some
non-Hodgkin’s Lymphoma Study Group has con- of these studies are retrospective, subset analyses of
ducted a randomized trial comparing CHOP chemo- clinical trials that were not initially stratified according
therapy given at 14 day intervals with R-CHOP, also to risk groups and not statistically powered to detect
given at 14 day intervals in elderly patients with differences in subgroup analysis.
advanced DLBCL. This was a four-arm study which Many prospective studies have subsequently been
also compared six versus eight cycles of therapy. The conducted, using the aaIPI to define poor-risk patients
rituximab-containing arms of this study were both (Table 12.3). Results have been variable, although
shown to have superior EFS and PFS when compared most studies have failed to show an advantage for
to the arms using CHOP only. Only the patients high-dose versus conventional dose remission
receiving R-CHOP for six cycles showed an OS consolidation.
advantage in this study, primarily because of late A randomized study compared eight cycles of
toxic deaths observed in patients receiving eight cycles CHOP chemotherapy with two cycles of a more
of the same therapy. This trial, known as the dose-intensive first line regimen, followed by high-
“RICOVER 60” trial, established R-CHOP for six dose therapy (HDT), and ASCT for responding
cycles at 14 day intervals as the standard approach in patients with advanced diffuse aggressive NHL, has
Germany and parts of the rest of Europe. recently been reported by Milpied et al. (Table 12.3).
This approach has not been adopted more widely Eligible patients were aged 15–60 years with low, low–
because of a lack of randomized trials directly compar- intermediate, or high–intermediate risk disease
ing the R-CHOP regimen given at 21 and 14 day according to the aaIPI, and therefore represented a
intervals. Two such studies have now been conducted relatively favorable group compared with many other
(see below). studies. 199
Chapter 12: Diffuse large B-cell lymphoma
Table 12.3 Results of high-dose therapy and ASCT in first remission for DLBCL and other aggressive NHLs.
An intent-to-treat analysis was performed, demon- DLBCL), irrespective of IPI risk group, and the addi-
strating a significantly higher EFS in the HDT arm tion of ASCT to standard induction therapy should
(55% versus 37%; P = 0.037), although there was no still be considered experimental, and used only in the
difference in OS. Subset analysis showed a significant context of clinical trials.
difference in OS in patients with high–intermediate Since most of the studies reported above were
risk disease (P = 0.001). conducted prior to the introduction of rituximab (see
In view of the conflicting results that have been below), or dose-dense chemotherapy regimens, their
reported in studies of first remission transplantation in relevance to current management of diffuse aggressive
aggressive NHL, a meta-analysis has recently been NHL is unclear. The SWOG 9704 study may help to
conducted. This study included data from 2018 clarify the role of remission consolidation with ASCT
patients from 13 randomized trials who were evaluable in patients receiving R-CHOP as first line therapy.
for outcome data. Although a significantly higher This study is now completed although results are not
complete response (CR) rate was reported for HDT yet available.
and ASCT, no differences in EFS or OS were observed.
No difference in outcome was seen according to IPI
group, and analyses according to the number of Addition of rituximab to chemotherapy
patients with DLBCL, transplant conditioning regi- The benefit of adding rituximab to CHOP chemother-
men, and response status prior to ASCT also failed to apy for DLBCL was initially demonstrated in a random-
clearly identify a group with superior outcome after ized trial from the GELA. In this study, 399 previously
HDT. Data regarding the benefit of ASCT for patients untreated patients aged between 60 and 80 years with
with IPI poor risk disease was inconclusive and there DLBCL were randomized to receive eight cycles of
was some evidence that patients with favorable risk CHOP chemotherapy at 21-day intervals, or the same
disease might have an inferior outcome after ASCT. chemotherapy plus rituximab given on day 1 of each
Based on these results, HDT and ASCT should not cycle. The R-CHOP arm was shown to be superior to
200 be considered a component of first line therapy in CHOP in terms of complete response rate (76% versus
patients with diffuse aggressive lymphoma (including 63%; P = 0.005), 2-year EFS (61% versus 43%; P = 0.002)
Chapter 12: Diffuse large B-cell lymphoma
and 2-year OS (70% versus 57%; P = 0.007). The survival These results suggest that the addition of rituxi-
advantage for R-CHOP in this trial was observed in all mab to chemotherapy is likely to benefit all risk
IPI risk groups. The results of this trial have been groups. Further studies will be required to determine
updated recently. The 10-year PFS rates were 20% and whether biological markers such as BCL-2 protein
36.5% for CHOP and R-CHOP, respectively. The cor- expression will reliably predict those patients likely
responding OS rates were 27.6% and 43.5%. to benefit from the addition of rituximab to
The addition of rituximab to CHOP in elderly chemotherapy.
patients with DLBCL has also been investigated in an The potential role of first remission HDT and
intergroup study in the USA, including 632 patients ASCT is uncertain for patients receiving rituximab as
aged over 60 years with previously untreated, advanced part of their induction regimen. The SWOG 9704
DLBCL. Patients were randomized to receive either six study described above may help to clarify its benefit
or eight cycles of CHOP, according to response, or the in this context.
same chemotherapy plus rituximab. A second random- Other ongoing trials are comparing the R-CHOP
ization was included for responding patients between combination given according to a standard 21 day
observation only, and maintenance rituximab, given or accelerated 14 day schedule. Preliminary data
once per week for 4 weeks at 6-month intervals for a have been presented from two studies. Neither
total of 2 years. In all, 632 patients were randomized, of study yet shows a clear difference in response rates,
whom 415 responders were subsequently randomized or PFS or OS endpoints, but further follow-up is
to maintenance rituximab or observation only. There required for both.
was a significant improvement in 3-year failure-free
survival (FFS) in the R-CHOP arm compared with
CHOP (53% versus 46%; P = 0.04) and in the main-
Addition of other new agents
tenance rituximab arm compared with observation to chemotherapy
alone. The advantage of maintenance rituximab Early results from phase II trials incorporating other
appeared to be limited to patients who did not receive novel agents with first line therapy for DLBCL have
this agent as part of their induction regimen. No OS been reported recently. The anti-CD22 monoclonal
differences were observed in the study, possibly because antibody, epratuzumab, has been combined with the
around 40% of patients randomized to receive CHOP R-CHOP regimen in a phase II study in 107 patients
alone received rituximab in the second randomization. with advanced DLBCL. The 12-month EFS and OS
Further evidence for the benefit of addition of rates were 79% and 89%, respectively, and no signifi-
rituximab to chemotherapy has been reported from a cant additional toxicity was observed compared
retrospective, population-based study from British with historical experience with R-CHOP alone.
Columbia, Canada, which has demonstrated higher Comparison with historical controls suggested that
EFS and OS rates for patients with DLBCL since the there may be a benefit for the epratuzumab rituximab
introduction of rituximab. (ER)-CHOP regimen in patients with poor-risk dis-
A benefit for rituximab in younger, low-risk ease by IPI, although this requires confirmation in a
patients has also been shown in the MInT randomized trial. Bortezomib has also been combined
(MabThera International Trial). This study included with first line chemotherapy in DLBCL. Initial results
patients with DLBCL with bulky stage I or stage suggest that this combination can be delivered safely,
II–IV disease, aged 18–60 years with IPI scores of 0 although the incidence and severity of peripheral neu-
or 1. Treatment comprised six cycles of CHOP, ropathy appears to vary according to the dose of bor-
CHOEP, or a comparable chemotherapy regimen tezomib. Early data also suggest that the benefit of
with or without rituximab administered on day 1 of adding bortezomib may be limited to patients with
each chemotherapy cycle. Preliminary results have non-germinal-center DLBCL.
been reported for the first 326 randomized patients.
The 2-year time to treatment failure (TTF) was 76%
for patients receiving chemotherapy plus rituximab,
Use of involved-field radiation therapy
compared with 60% for those receiving chemother- in advanced DLBCL
apy alone (P < 0.001). The corresponding rates for OS The role of IF-RT in limited-stage DLBCL has been
were 94% versus 87% (P = 0.001). well documented (see above). In the setting of 201
Chapter 12: Diffuse large B-cell lymphoma
Rituximab-CHOP x 8
CR/PR
observe NR/PD
NR/PD Not eligible for
Eligible for ASCT ASCT
Relapse/PD
response
NR/PD
High-dose therapy/autologous
SCT
frequently invades surrounding structures such as the in which rituximab-based second line regimens are
pleura and pericardium. It typically affects young compared are in progress. Results from the CORAL
adults, with a female predominance. Disseminated dis- study, comparing R-DHAP with R-ICE in patients
ease beyond the chest is relatively uncommon. Gene with relapsed DLBCL, have been reported recently.
expression profiling studies have demonstrated similar- No difference was observed in response rates or
ities with Hodgkin lymphoma. 3-year EFS according to second line regimen.
Recent clinical studies have demonstrated that this The National Cancer Institute of Canada is leading
entity has a favorable prognosis, although optimal a similar study comparing R-DHAP with another sec-
treatment strategies remain unclear. Several single ond line regimen, R-GDP (rituximab, gemcitabine,
center phase II studies have reported favorable treat- dexamethasone, cisplatin). This study is still in
ment results for relatively intensive chemotherapy progress.
regimens such as MACOP-B (methotrexate, doxoru- Patients who do not achieve at least a partial remis-
bicin, cyclophosphamide, vincristine, prednisone, sion to second line therapy have a poor outcome when
bleomycin) compared with CHOP chemotherapy. treated with HDT and ASCT in most series. These
However, these studies were conducted prior to the patients should be considered for trials of novel treat-
introduction of rituximab. Recent data have suggested ment approaches. A recent retrospective analysis from
that R-CHOP is an effective regimen for this condi- the UK has demonstrated that patients who do not
tion. The role of radiation therapy is also uncertain. respond to one conventional dose salvage regimen are
Although consolidative radiation therapy is widely not rescued by subsequent conventional dose salvage
used in this disease, its benefit has been questioned in therapy.
an analysis from the MInT trial and in a retrospective
study from British Columbia. The potential use of
FDG–PET after induction chemotherapy to determine High-dose therapy and autologous
which patients may benefit from consolidative IF-RT
will be investigated in a prospective study in Europe.
stem cell transplantation
The use of HDT and ASCT has been regarded as the
standard of care for patients with relapsed DLBCL
Management of relapsed DLBCL for over a decade. This is based on the results of
the PARMA randomized trial. This study included
Second line therapy prior to ASCT 215 patients with relapsed aggressive NHL (mostly
Although HDT and ASCT remains the standard of DLBCL), initially treated with two cycles of salvage
care for patients with relapsed DLBCL, a survival chemotherapy with DHAP. Responding patients were
benefit for this approach has only been demonstrated randomized to receive further DHAP chemotherapy,
convincingly in patients with disease that is responsive or to proceed to HDT using BEAC (carmustine,
to second line salvage chemotherapy. Commonly used etoposide, cytarabine, cyclophosphamide) and autolo-
second line regimens, including DHAP (dexametha- gous bone marrow transplantation. Significantly supe-
sone, high-dose cytosine arabinoside, cisplatin), ICE rior 5-year EFS (46% versus 12%; P = 0.0001) and OS
(ifosfamide, carboplatin, etoposide) and mini-BEAM (53% versus 32%; P = 0.038) rates were observed for
(carmustine, etoposide, cytarabine, melphalan) pro- the transplant arm compared with the conventional
duce overall response rates (ORR) of 40–60%, and chemotherapy arm. No formal follow-up analysis of
CR rates of only 25–35%. Analysis of the effectiveness the PARMA study has been published. Although this
of ASCT from the date of transplantation therefore study established ASCT as the standard approach
overestimates the effectiveness of this approach in the for patients with relapsed chemosensitive DLBCL,
entire population of relapsed patients. When analyzed the results should be interpreted cautiously. Of the
by intent to treat, the EFS for relapsing patients with 215 patients entered onto the study, only 109 were
DLBCL treated with second line salvage therapy fol- randomized, most commonly because the patients
lowed by ASCT is between 20% and 35%. did not achieve an adequate response to second line
The development of more effective second line therapy with DHAP (only 56% of patients responded
regimens has the potential advantage of increasing to this chemotherapy). All subsequent survival anal-
response rates and therefore increasing the number yses were restricted to randomized patients only, and 203
of patients eligible for ASCT. Two randomized studies no intent-to-treat analysis was performed.
Chapter 12: Diffuse large B-cell lymphoma
In addition to these limitations, the relevance Allogeneic stem cell transplantation in DLBCL
of the study in the present context is unclear. Current data on the role of allogeneic SCT in aggres-
Improved supportive care, including the use of sive lymphoma, either using myeloablative or non-
peripheral blood progenitor cells, has reduced the myeloablative conditioning regimens, are limited.
morbidity associated with HDT and extended its Comparative studies of allogeneic and autologous
use to older patient groups, typically up to 70–75 SCT in aggressive NHL have not shown a survival
years old. Most centers will now accept patients for advantage for allogeneic SCT, despite the lower relapse
transplantation if they achieved partial response (PR) rate in allogeneic recipients. The lower relapse rate has
to prior therapy, unlike the PARMA study, in which been offset by the increased regimen-related mortality
a previous CR was required for eligibility. The associated with allogeneic transplantation. In the
population of patients now receiving ASCT is there- absence of clear evidence of a clinically relevant
fore less defined than that in the original PARMA graft-versus-lymphoma effect in aggressive NHL, the
study, raising questions concerning the current rele- use of allogeneic SCT should be restricted to research
vance of this trial. The addition of rituximab to protocols.
combination chemotherapy regimens, and the Allogeneic transplantation for patients who relapse
advent of accelerated 14 day regimens for first line after autologous transplantation is increasing in use,
treatment, have improved DFS and OS in DLBCL. although there are few data to confirm its benefit. A
It is not clear whether patients whose disease relapses recent retrospective study from the International Bone
after one of these regimens will have the same Marrow Transplant Registry analyzed results for 114
salvage rates as those treated without monoclonal patients with various subtypes of NHL who received
antibodies as part of their initial treatment. allogeneic SCT after relapse following autologous
Retrospective comparisons of outcomes following SCT. All patients underwent myeloablative condition-
ASCT for patients receiving first line CHOP or ing. The regimen-related mortality was 22% at 3 years,
CHOP-R have recently been reported. No difference and the 5-year OS and PFS were 24% and 5%, respec-
in EFS or OS following ASCT was observed accord- tively. No analysis was performed according to NHL
ing to initial therapy in these studies. By contrast, subtype, but the study suggests that the curative poten-
the CORAL study identified prior therapy with rit- tial for this approach is low and that its use should be
uximab as an adverse prognostic factor. The 3-year restricted to patients in prospective trials.
EFS for patients previously treated with rituximab
was 21% compared with 47% for those without
previous rituximab exposure. Prognostic factors for relapsed DLBCL
A recent report of dose-adjusted EPOCH-R as Multiple early single institution and registry studies of
primary therapy in DLBCL has shown a 2-year PFS ASCT in aggressive NHL demonstrated the impor-
rate of 83%, with a very similar OS, indicating the high tance of sensitivity of disease to second line therapy
activity of this regimen, and the apparent inability to as a predictive factor for outcome after transplanta-
salvage relapsed patients with a standard transplant tion. Short (less than 1 year) remission duration and
approach. disease bulk at the time of ASCT were also identified as
Although HDT and ASCT remains the standard of adverse factors in many studies.
care for patients with relapsed DLBCL which is still The aaIPI has been shown to have predictive value
sensitive to second line chemotherapy, the true benefit in a follow-up report of the PARMA study. It proved
of this approach in the context of modern first line highly predictive of response to DHAP. Patients with
therapies is unclear and requires re-evaluation. an aaIPI score of 0 had an ORR of 77% compared to
The potential benefit of rituximab maintenance only 42% for those with three adverse factors. The
after HDT and ASCT has also been investigated in aaIPI was predictive of OS for the entire patient
the CORAL study, in which a second randomization cohort.
to maintenance rituximab versus no further therapy When randomized patients were analyzed sepa-
was included for patients undergoing ASCT. Results rately, the aaIPI was predictive for those receiving
for this second randomization have not yet been DHAP, but not in those undergoing ASCT. In a subset
reported and this approach should be considered analysis, there was no difference in OS or PFS accord-
204 experimental. ing to the randomized arm for patients with an aaIPI
Chapter 12: Diffuse large B-cell lymphoma
score of 0, although a significant difference remained without additional toxicity or prolongation of engraft-
for those with scores of 1–3. Similar results have been ment times. In view of the emerging data on the use of
reported from other centers. rituximab as maintenance therapy in DLBCL, new
In the CORAL study (see above), in addition to studies are currently being planned to assess the use
prior rituximab exposure, other adverse factors for of radioimmunotherapy to consolidate remission after
EFS were a prior remission duration of less than 12 induction therapy with R-CHOP and other rituximab-
months after diagnosis, and an IPI score of 2 or higher. chemotherapy combinations in DLBCL.
These subset analyses should be interpreted cau-
tiously, but question whether ASCT offers a survival Relapsed DLBCL – conclusions
advantage to low-risk patients. * HDT and ASCT is the standard of care for patients
One recent study has also addressed the potential
with chemosensitive relapse.
value of cell of origin, as defined by tissue microarrays, * The true impact of SCT in patients relapsing after
as a prognostic factor for patients undergoing salvage
rituximab-containing therapy is unclear.
therapy with HDT and ASCT. In this study, no differ-
* Allogeneic SCT has no proven benefit in relapsed
ence in OS was observed when patients with germinal-
DLBCL and should be used only in the context of
center or non-germinal-center phenotypes were
prospective clinical trials.
compared.
The potential value of functional imaging using
* Patients with refractory disease do not benefit from
FDG–PET following salvage therapy but prior to SCT – other, experimental, strategies should be
HDT and ASCT has also been assessed. In patients considered.
with Hodgkin’s lymphoma, PET scan positivity prior
* New treatment approaches are required for
to ASCT has been shown to be highly predictive of patients who relapse after SCT or who are ineligible
outcome. In the case of DLBCL, limited published data for transplantation.
are available from some small series, and the predictive
value of PET in this context therefore remains unclear. New therapeutic targets in DLBCL
The use of molecular techniques, including GEP, has
Other treatment approaches for relapsed identified many potential rational therapeutic targets
which are now under investigation in DLBCL. Some of
disease these are listed in Table 12.4.
Radiolabeled monoclonal antibodies Histone deacetylase (HDAC) inhibitors may have
Both 131I tositumomab (Bexxar) and 90Y-ibritumomab several potential mechanisms of action in DLBCL.
tiuxetan (Zevalin) are active agents in indolent and Acetylation of histones in the nucleosome is a major
transformed CD20-positive B-cell lymphomas. Few determinant of the regulation of many genes.
data are available for these agents in DLBCL. A phase Deacetylation of histones results in condensed chro-
II study of 90Y-ibritumomab tiuxetan in 104 patients matin structure and repression of gene transcription.
with relapsed or refractory DLBCL, not eligible for Inhibitors of HDAC such as suberoylanilide
HDT and ASCT, has shown an ORR of 44%. The
response rate was higher in patients who had not had Table 12.4 Potential therapeutic targets in DLBCL.
prior therapy with rituximab, compared with those who
had previously been treated with rituximab and chemo- CD22
therapy as their primary treatment. Median PFS was Histone deacetylase
around 6 months for patients who were rituximab- Proteasome
naïve, compared with only about 2 months for those
BCL-6
previously treated with rituximab.
These agents have also been studied in high-dose BCL-2
regimens used with ASCT. Most studies to date have mTOR/AKT
included patients with mixed histologic subtypes of PKC-β
NHL. Early results indicate that both 131I tositumo- Syk
mab and 90Y-ibritumomab tiuxetan can be combined 205
Btk
with standard high-dose chemotherapy regimens
Chapter 12: Diffuse large B-cell lymphoma
hydroxamic acid (SAHA) have been shown to induce Many of these agents have shown promising clin-
differentiation and/or apoptosis in various tumor cell ical activity, with non-overlapping toxicities compared
lines, and this agent has demonstrated clinical activity with conventional chemotherapy. As a result, phase II
in heavily pre-treated patients with NHL. studies adding these agents to standard induction reg-
HDAC inhibitors may exert some of their activity imens such as R-CHOP for previously untreated
in DLBCL through BCL-6. This gene is thought to be patients with DLBCL are now being developed.
important in the pathogenesis of DLBCL and has been
shown to be antiapoptotic in tumor cells, through
inhibition of transcription of the p21WAF-1 gene. In Further reading
the presence of HDAC inhibitors, p21WAF-1 tran- Calaminici M, Piper K, Lee AM, Norton AJ. CD23
scription results in growth inhibition, apoptosis, and expression in mediastinal large B-cell lymphomas.
differentiation. Various HDAC inhibitors are now in Histopathology, 2004;45:619–624.
early-phase clinical trials in DLBCL with confirmed Choi WWL, Weisenburger DD, Greiner TC, et al. A new
activity. immunostain algorithm classifies diffuse large B-cell
lymphoma into molecular subtypes with high accuracy.
Mammalian target of rapamycin (mTOR) is a ser-
Clin Cancer Res, 2009;15:5494–5502.
ine/threonine protein kinase involved in the regula-
tion of cell growth in response to multiple nutrients. It Coiffier B, Thieblemont C, Van Den Neste E, et al. Long-
term outcome of patients in the LNH-98.5 trial, the first
is phosphorylated via the PI3/AKT pathway and plays
randomized study comparing rituximab-CHOP to
a role in the regulation of apoptosis. It also appears to standard CHOP chemotherapy in DLBCL patients: a
be involved in the regulation of angiogenesis through study by the Groupe d’Etudes des Lymphomes de
effects on production of vascular endothelial growth l’Adulte. Blood, 2010;116:2040–2045.
factor (VEGF). Phase I and II studies of various agents Compagno M, Lim WK, Grunn A, et al. Mutations of
which target mTOR, including everolimus and temsir- multiple genes cause deregulation of NF-kappaB in
olimus, have confirmed that these agents have clinical diffuse large B-cell lymphoma. Nature,
activity in relapsed and refractory DLBCL. 2009;459:717–721.
The expression of protein kinase C β (PKCβ) by Copie-Bergman C, Plonquet A, Alonso MA, et al. MAL
immunohistochemistry has been shown to be an expression in lymphoid cells: further evidence for MAL as
adverse prognostic factor in DLBCL, and in vitro a distinct molecular marker of primary mediastinal large
studies of inhibitors of PKCβ, including bryostatin, B-cell lymphomas. Mod Pathol, 2002;15:1172–1180.
have validated this molecule as a target in DLBCL. Davis RE, Ngo VN, Lenz G, et al. Chronic active B-cell-
A phase II study of the PKCβ inhibitor, enzastaurin, receptor signalling in diffuse large B-cell lymphoma.
Nature, 2010;463:88–92.
has confirmed the activity of this agent in DLBCL.
A phase III study of this agent as a maintenance Fisher RI, Gaynor ER, Dahlberg S, et al. Comparison of a
therapy after completion of R-CHOP is currently in standard regimen (CHOP) with three intensive
chemotherapy regimens for advanced non-Hodgkin’s
progress.
lymphoma. N Engl J Med, 1993;328:102–106.
Data have also been reported recently for the clin-
Greb A, Bohlius J, Trelle S, et al. High-dose chemotherapy
ical activity of agents which target the B-cell receptor
with autologous stem cell support in first-line treatment
and downstream signaling molecules in DLBCL. of aggressive non-Hodgkin lymphoma – results of a
Fostamatinib, an agent that targets spleen tyrosine comprehensive meta-analysis. Cancer Treat Rev,
kinase (syk), has been shown to produce responses in 2007;33:338–346.
patients with relapsed and refractory DLBCL. Gisselbrecht C, Glass B, Mounier N, et al. Salvage regimens
Similarly, PCI-37265, an inhibitor of Bruton’s tyrosine with autologous transplantation for relapsed large B-cell
kinase (Btk), has been shown to produce responses in a lymphoma in the rituximab era. J Clin Oncol,
phase I study, including in patients with relapsed 2010;28:4184–4190.
aggressive NHL, including DLBCL. Habermann TM, Weller EA, Morrison VA, et al.
Immunomodulatory drugs, including lenalido- Rituximab-CHOP versus CHOP alone or with
mide, have demonstrated activity in DLBCL in phase maintenance rituximab in older patients with diffuse
I and II clinical trials, and these are now being eval- large B-cell lymphoma. J Clin Oncol, 2006;24:3121–3127.
uated in several clinical contexts, including as main- Hans CP, Weisenberger DD, Greiner TC, et al.
206 tenance therapy. Conformation of the molecular classification of diffuse
Chapter 12: Diffuse large B-cell lymphoma
large B-cell lymphoma by immunohistochemistry using a or without rituximab: an exploratory analysis of the
tissue microarray. Blood, 2004;103:275–282. MabThera International Trial Group (MInT) study.
Horning SJ, Weller E, Kim K, et al. Chemotherapy with or Lancet Oncology, 2008;9:435–444.
without radiotherapy in limited-stage diffuse aggressive Pfreundschuh M, Schubert J, Ziepert M, et al. Six versus
non-Hodgkin’s lymphoma: Eastern Cooperative eight cycles of bi-weekly CHOP-14 with or without
Oncology Group Study 1484. J Clin Oncol, rituximab in elderly patients with aggressive CD20+
2004;22:3032–3038. B-cell lymphomas: a randomised controlled trial
Linderoth J, Jerkemen M, Cavallin-Stahl E, et al. (RICOVER-60). Lancet Oncol, 2008:9:105–116.
Immunohistochemical expression of CD23 and CD40 Phan J, Mazloom A, Medeiros JL, et al. Benefit of
may identify prognostically favorable subgroups of consolidative radiation therapy in patients with diffuse
diffuse large B-cell lymphoma: a Nordic lymphoma large B-cell lymphoma treated with R-CHOP
group study. Clin Cancer Res, 2003;9:722–728. chemotherapy. J Clin Oncol, 2010;28:4170–4176.
Lossos IS. Molecular pathogenesis of diffuse large B-cell Philip T, Guglielmi C, Hagenbeek A, et al. Autologous bone
lymphoma. J Clin Oncol, 2005;23:6351–6357. marrow transplantation as compared with salvage
Miller TP, Dahlberg S, Cassady JR, et al. Chemotherapy chemotherapy in relapses of chemotherapy-sensitive
alone compared to chemotherapy plus radiotherapy for non-Hodgkin’s lymphoma. N Engl J Med,
localized intermediate- and high-grade non-Hodgkin’s 1995;333:1540–1545.
lymphoma. N Engl J Med, 1998;339:21–26. Reyes F, Lepage E, Ganem G, et al. ACVBP versus CHOP
Ngo VN, Young RM, Schmitz R, et al. Oncogenically active plus radiotherapy for localized aggressive lymphoma. N
MYD88 mutations in human lymphoma. Nature, Engl J Med, 2005;352:1197–1205.
2011;470:115–119. Rieger M, Osterborg A, Pettengell R, et al. Primary
Ott G, Ziepert M, Klapper W, et al. Immunoblastic mediastinal B-cell lymphoma treated with CHOP-like
morphology but not the immunohistochemical GCB/ chemotherapy with or without rituximab: results of the
nonGCB classification predicts outcome in diffuse large Mabthera International Trial Group study. Ann Oncol,
B-cell lymphoma in the RICOVER-60 trial of the 2011;22:664–670.
DSHNHL. Blood, 2010;116:4916–4925. Rosenwald A, Wright G, Chan WC, et al. The use of
Persky DO, Unger JM, Spier CM, et al. Phase II study of molecular profiling to predict survival after
rituximab plus three cycles of CHOP and involved-field chemotherapy for diffuse large B-cell lymphoma. N Engl J
radiotherapy for patients with limited-stage aggressive Med, 2002;346:1937–1947.
B-cell lymphoma: Southwest Oncology Group study Savage KJ, Al-Rajhi N, Voss N, et al. Favorable outcome of
0014. J Clin Oncol, 2008;26:2258–2263. primary mediastinal large B-cell lymphoma in a single
Pfreundschuh M, Trumper L, Kloess M, et al. Two-weekly institution: the British Columbia experience. Ann Oncol,
or 3-weekly CHOP chemotherapy with or without 2006;17:123–130.
etoposide for the treatment of young patients with Sehn LH, Donaldson J, Chhanabhai M, et al. Introduction
good-prognosis (normal LDH) aggressive lymphomas: of combined CHOP plus rituximab therapy
results of the NHL-B1 trial of the DSHNHL. Blood, dramatically improved outcome of diffuse large B-cell
2004;104:626–633. lymphoma in British Columbia. J Clin Oncol,
Pfreundschuh M, Trümper L, Osterborg A, et al. 2005;23:5027–5033.
CHOP-like chemotherapy plus rituximab versus Sehn LH, Savage KJ, Hoskins P, et al. Limited stage DLBCL
CHOP-like chemotherapy alone in young patients with patients with a negative PET scan following 3 cycles of
good-prognosis diffuse large-B-cell lymphoma: a R-CHOP have an excellent outcome following
randomised controlled trial by the MabThera abbreviated immunochemotherapy alone. Ann Oncol,
International Trial (MInT) Group. Lancet Oncol, 2008;19 (suppl): (abstract 052).
2006;7:379–391. Wilson WH, Dunleavy K, Pittaluga S, et al. Phase II study of
Pfreundschuh M, Ho A, Cavallin-Stahl E, et al. Prognostic dose-adjusted EPOCH and rituximab in untreated diffuse
significance of maximum tumour (bulk) diameter in large B-cell lymphoma with analysis of germinal center
young patients with good-prognosis diffuse large B-cell and post-germinal center biomarkers. J Clin Oncol,
lymphoma treated with CHOP-like chemotherapy with 2008;26:2717–2724.
207
Chapter
13
Central nervous system lymphomas
Andrés J. M. Ferreri and Lisa M. DeAngelis
Pathology: Andrew Wotherspoon Molecular cytogenetics: Andreas Rosenwald and German Ott
A variety of lymphomas can involve the central nervous Rare cases of small lymphocytic lymphoma, dural-
system (CNS) at different phases of their evolution, based mucosa-associated lymphoid tissue (MALT)
both in immunocompetent and immunocompromised lymphoma, and T-/NK-cell lymphoma similar to
individuals. They represent a heterogeneous group those seen in tissue outside the CNS have been
of malignancies, with variable clinical and behavioral described. Secondary involvement by lymphoma orig-
characteristics, and require different therapeutic inating elsewhere is also encountered.
approaches. In this chapter, the therapeutic manage-
ment of these malignancies will be analyzed separately
in three main entities: primary CNS lymphomas Conventional therapeutic strategies
(PCNSL), secondary CNS lymphomas (SCNSL), and PCNSLs are aggressive malignancies arising within
other, less common, forms of CNS lymphomas. and confined to the CNS. They comprise 3% of intra-
cranial neoplasms, and in the past two decades their
incidence has risen both in immunocompromised and
Histopathology immunocompetent individuals, especially in those
Primary diffuse large B-cell lymphoma (DLBCL) of over 50 years of age. The main patient characteristics
the CNS shows diffuse parenchymal growth with at presentation are summarized in Table 13.1. Current
dense cuffing in perivascular spaces. The latter is best therapeutic knowledge in PCNSL results from three
appreciated towards the periphery of the tumor. Large randomized trials, multiple single arm phase II trials,
areas of necrosis may be present, especially in those meta-analyses, and large retrospective, multicenter
treated with steroids prior to biopsy. Towards the series. The few randomized trials limit comparison
periphery of the tumor the neoplastic cells are often between new approaches, but therapeutic progress
admixed with astrocyte gliosis. Cytologically, the neo- has been made by these studies as well as single arm
plastic cells resemble those of DLBCL encountered phase II trials. However, the use of divergent study
elsewhere. Similarly, the immunophenotype of pri- designs and entry criteria in all prospective trials, as
mary CNS DLBCL is similar to those outside the well as the presence of some methodological pitfalls,
CNS and is positive for B-cell markers (CD20, make comparisons difficult. Despite the improved
CD79a, and PAX-5). Strong IRF4/MUM1 staining is outcome reported in prospective trials, progress in
seen in 90% of cases. CD10 is expressed by a minority the treatment of PCNSL has not been reflected in
of cases, while many express BCL-6. BCL-2 is often studies of population-based cohorts, and survival has
expressed. As these lymphomas (by definition) occur not improved consistently in the past three decades.
in immunocompetent patients, Epstein–Barr virus This finding highlights one of the most important
(EBV) is generally absent. limitations of all prospective trials, which is potential
Many lymphomas involving the CNS in immu- patient selection bias.
nocompromised patients are EBV-positive and show The optimal treatment of PCNSL requires a multi-
features of DLBCL, many with an immunoblastic disciplinary approach. If neuroimaging suggests the
morphology. Burkitt lymphoma, a common lym- possibility of PCNSL (Figure 13.1), then a stereotactic
phoma in the immunocompromised patient, may biopsy is the most appropriate surgical approach.
involve the CNS. Aggressive surgical resection should be avoided
208 Lymphoma, Second Edition, ed. Robert Marcus, John W. Sweetenham, and Michael E. Williams. Published by
Cambridge University Press. © Cambridge University Press 2014.
Chapter 13: Central nervous system lymphomas
Table 13.2 Staging work-up and pre-treatment evaluations Methotrexate and cytarabine
in PCNSL.
Antimetabolites such as MTX and araC constitute the
Staging backbone of most chemotherapy combinations used in
Physical examination PCNSL, with demonstrated efficacy in prospective trials
Routine blood studies (Table 13.3). Different regimens have been used in an
Whole-brain MRI
Contrast total body CT scan attempt to improve efficacy; MTX doses from 1 to 8 g/
Ophthalmologic evaluation (including slit-lamp m2 are feasible in PCNSL; both total dose and infusion
examination) rate are important for MTX delivery to the brain paren-
Cerebrospinal fluid cytology
Cerebrospinal fluid biochemical examination chyma and CSF. MTX doses ≥1 g/m2 result in tumor-
Bone marrow biopsy icidal levels in the brain parenchyma and doses ≥3 g/m2
Testicular ultrasonography (older men) yield tumoricidal levels in the CSF. Conversely, doses
18
FDG–PET (investigational role)
Suspicion of vitreal infiltration may require confirmation up to 8 g/m2 delivered in a 24-hour continuous infusion
by vitrectomy do not reliably achieve an adequate CSF level. HD-
Prognostic factors MTX, when used alone, is a safe treatment even in
Age patients older than 70. Calculated or measured creati-
Performance status
Lactate dehydrogenase serum level
nine clearance and glomerular filtration rates have been
Cerebrospinal fluid protein concentration proposed to tailor MTX dose. Individualized dosing of
Involvement of deep regions of the brain HD-MTX might have the potential to improve outcome
Pre-treatment assessment in patients with PCNSL, even when administered con-
Neurological examination currently with high-dose (HD)-araC. In the future, this
Biochemical serum profile
Baseline neuropsychiatric tests
could be carried out by using first-cycle pharmacoki-
Renal and hepatic functionality tests (creatinine clearance) netic modeling with determination of potential dose
Cardiac functionality tests* adaptations for later cycles using Bayesian analysis.
HIV, hepatitis B and C virus evaluation
HD-MTX (8 g/m2) monochemotherapy yielded an
* Echocardiography with left ventricular ejection fraction overall response rate (ORR) of 51–68%, with a 3-year
evaluation is advisable in elderly patients eligible for HD-MTX-
based chemotherapy considering that the administration of OS of 35%, which is similar to the 3-year OS of 32–
this drug requires adequate hydration (near 3000 mL/day on 47% recorded in trials evaluating polychemotherapy
the days before and after MTX). HIV, Human immunodeficiency combinations with a MTX dose of 3.5 g/m2
virus.
(Table 13.3). Noteworthy, dose reduction because of
impaired creatinine clearance was needed in 45% of
patients in trials using MTX 8 g/m2, whereas in the 3.5
agents). The combination of drugs from the first group g/m2 trials few patients needed reduction in MTX
(i.e. the CHOP regimen) represents the backbone of dose. Thus, tolerability and activity of 3.5 g/m2 suggest
treatment of extra-CNS DLBCL but exhibits negligible that this might be a good compromise between safety
activity in PCNSL. While many patients have an imme- and efficacy for combination regimens.
diate radiographic response, most experience progres- A large meta-analysis and a retrospective series
sion after two to three cycles. This may be explained by have suggested a survival improvement when araC is
positron emission tomography (PET) studies that dem- added to MTX. In a recently published randomized
onstrate normalization of the disrupted BBB 3–4 weeks trial, 79 patients with newly diagnosed PCNSL were
after initial chemotherapy, suggesting that the bulky assigned to four cycles of MTX (3.5 g/m2, day 1) alone
tumor not protected by the BBB responds, but micro- or combined with araC (2 g/m2/12 h, days 2 and 3);
scopic tumor is not adequately treated and progresses. both arms were followed by WBRT. The addition of
Importantly, the addition of CHOP chemotherapy both araC resulted in significantly improved complete
to high-dose methotrexate (HD-MTX) and whole-brain remission rate and OS compared with MTX alone.
radiotherapy (WBRT) did not improve outcome with Hematologic toxicity was higher in the MTX-araC
respect to the use of these strategies as an exclusive arm, while non-hematologic toxicities were uncom-
approach. As a consequence, CHOP and CHOP-like mon. This MTX-araC combination may become the
regimens have been abandoned, and chemotherapy current standard chemotherapeutic approach for de
combinations for PCNSL are currently designed by novo PCNSL since it is supported by the only random-
210 using drugs from the other two subgroups. ized study thus far.
Table 13.3 Published prospective trials on PCNSL in immunocompetent patients treated with chemotherapy alone or combined treatment.
Ref. No. Median Treatment Therapy Conditioning WBRT Outcome Neurotoxicity (%) Median TRM
of age line (induction → regimen follow-up (%)
pts. (range) intensification) (months)
1 22 53 (27–64) Salvage araC + VP16 TT/Bu/Cy No 3-year 32 41 4
OS 64%
2 43 52 (23–65) Salvage araC + VP16 TT/Bu/Cy No 2-year 5 36 7
OS 45%
3 6 53 (30–66) First line MBVP → IFO + araC BEAM Yes 2-year 33 41 0
OS 40%
4 25 52 (21–60) First line MBVP → IFO + araC BEAM Yes 4-year 8 34 4
OS 64%
5 28 53 (25–71) First line HD-MTX → araC BEAM No 2-year 0 28 4
OS 55%
6 23 55 (18–69) First line HD-MTX → – Bu/TT Yes§ 2-year 39 15 13
OS 48%
7 7 56 (41–64) First line* HD-MTX → araC Bu/TT/Cy No 3-year 0 28 14
OS 50%
8 30 54 (27–64) First line HD-MTX → araC + TT BCNU/TT Yes 5-year 17 63 3
OS 69%
9 13 54 (38–67) First line HD-MTX → araC + TT BCNU/TT Yes§ 3-year 0 23 0
OS 77%
araC, cytarabine; BCNU, carmustine; BEAM, carmustine, etoposide, cytarabine, and melphalan; Bu, busulfan; Cy, cyclophosphamide; IFO, ifosfamide; MBVP (regimen), methotrexate,
carmustine, etoposide, and methylprednisolone; OS, overall survival; TRM, treatment-related mortality; TT, thiotepa; VP16, etoposide; WBRT, whole-brain irradiation.
* One patient received the treatment as salvage therapy.
§ Only for patients not achieving a complete remission.
References: (1) Soussain C, et al. J Clin Oncol, 2001;19(3):742–749; (2) Soussain C, et al. J Clin Oncol, 2008;26(15):2512–2518; (3) Brevet M, et al. Eur J Haematol, 2005;75(4):288–292; (4) Colombat P,
et al. Bone Marrow Transplant, 2006;38(6):417–420; (5) Abrey LE, et al. J Clin Oncol, 2003;21(22):4151–4156; (6) Montemurro M, et al. Ann Oncol, 2007;18(4):665–671; (7) Cheng T, et al. Bone Marrow
Transplant, 2003;31(8):679–685; (8) Illerhaus G, et al. J Clin Oncol, 2006;24(24):3865–3870; (9) Illerhaus G, et al. Haematologica, 2008;93(1):147–148.
Chapter 13: Central nervous system lymphomas
leptomeningeal lymphoma in the absence of a paren- Finally, therapeutic MTX concentrations can be
chymal mass represents less than 5% of all PCNSL achieved in CSF using intravenous doses ≥3 g/m2, and
(Figure 13.3). Meningeal dissemination is detected by preliminary data suggest that systemic HD-MTX is able
conventional CSF cytology examination in 16% of to clear the CSF of neoplastic cells.
patients, by polymerase chain reaction (PCR) in 11%,
and by magnetic resonance imaging (MRI) in 4%.
However, the rate of discordant PCR and cytomor- Intraocular lymphoma
phologic results is high, and cytology and PCR should In 5–20% of patients, PCNSL involves the eyes.
be regarded as complementary methods. In effect, an Intraocular lymphoma (IOL) can occur in isolation
autopsy study demonstrating meningeal involvement (primary ocular lymphoma) or as a component of
in 100% of cases suggests that leptomeningeal dissem- more extensive PCNSL. Since the eye is an extension
ination is underestimated with current methods. of the CNS, its involvement is not considered systemic
Prognosis of leptomeningeal lymphoma is variable, dissemination, even if bilateral. The neoplastic cells
owing to the difficulty of delivering effective treatment can infiltrate the vitreous humor, the retina, the cho-
to the subarachnoid space, which could be optimized by roid, and, less frequently, the optic nerve. Patients
using an Ommaya reservoir. However, indications for typically present with floaters and blurred vision and
intrathecal/intraventricular chemotherapy are debat- are often misdiagnosed as having a benign ophthal-
able considering that its efficacy has not been studied mologic condition. Patients treated for isolated IOL
prospectively, while a large PCNSL retrospective study have an 80% risk of developing cerebral involvement
showed no benefit from intrathecal therapy. Conversely, up to 10 years or more after initial diagnosis.
the comparison of two phase II trials seems to provide The prognosis of IOL is similar to that of PCNSL
indirect evidence of the benefit of intraventricular drug without ocular involvement. There is no standard treat-
delivery in PCNSL. In fact, the exclusion of intraven- ment for isolated IOL. Systemic administration of MTX
tricular chemotherapy was associated with worse results and cytarabine can yield therapeutic levels of drug in the
with respect to the same chemotherapy plus intraven- intraocular fluids and clinical responses have been
tricular drug delivery. However, the theoretical advant- documented; however, relapse is common. The efficacy
age of intraventricular drug administration may not of cytostatics is dependent on intraocular pharmacoki-
outweigh the additional risk of infective complications netics, which are not well understood, and preliminary
of the reservoir, and the increased risk of neurotoxicity data suggest that intraocular concentrations of MTX are
216
and chemical meningitis associated with this strategy. achieved when the drug is given at 8 g/m2, with a high
Chapter 13: Central nervous system lymphomas
rate of persistent ocular disease. Better disease control should be treated similarly to other PCNSL. Because
has been reported by combining ocular irradiation with neurolymphomatosis is a condition outside the CNS,
HD-MTX-based chemotherapy. Irradiation of the pos- there may be a rationale to incorporate rituximab (in
terior two-thirds of the globes with 35–45 Gy has been CD20-positive lymphomas) and CHOP chemotherapy
recommended, and more recent experience suggests in combination with agents that penetrate the blood–
using radiotherapy of the entire orbit up to 20 Gy, nerve barrier in the treatment of this entity.
followed by an additional 10 Gy after shielding the
anterior chamber of the eyes. Even in the presence of
unilateral disease, both eyes should be irradiated
Relapsed disease
because microscopic bilateral involvement occurs in Median survival for patients with recurrent or refrac-
80% of patients. The actual incidence of ocular compli- tory PCNSL without treatment is 2 months, but sal-
cations (cataract, dry eyes, punctate keratopathy, retin- vage treatment may substantially prolong survival.
opathy, optic atrophy) is not reported because of the Unfortunately, the optimal salvage regimen has not
short follow-up of published series; however, cataract been defined and a variety of possible strategies have
occurs in virtually all patients. been reported (Table 13.5). The choice of the optimal
The poor results obtained with conventional strat- salvage strategy should take into consideration the
egies have induced investigators to identify new ther- patient’s age, PS, site of relapse, and prior therapy.
apeutic approaches. Intravitreal injections of MTX Radiotherapy may be used at recurrence in previously
have been associated with better intraocular disease non-irradiated patients; salvage WBRT has been
control (ORR 100%) in small series of patients with associated with an ORR of 79% and a median survival
relapsed IOL, but this benefit did not affect OS. after relapse of 16 months in patients who experi-
Moreover, important side effects have been reported enced failure after initial HD-MTX. Salvage chemo-
in up to 73% of treated eyes, with a significant deteri- therapy is often used in an effort to improve disease
oration of visual acuity in 27% of patients. Drugs other control while reducing neurotoxicity. Reinduction
than MTX have been administered by this route in with HD-MTX resulted in a response in approxi-
IOL patients. The administration of rituximab appears mately 50% of patients, with a median PFS of 10
safe and possibly efficacious, whereas intravitreal thi- months. Preliminary experiences showed encourag-
otepa requires further investigation. ing but unconfirmed results with polychemotherapy
combinations (Table 13.5). Patients with relapsed or
refractory PCNSL constitute the optimal context to
Spinal cord lymphoma identify potentially new active drugs.
Primary lymphomas of the spinal cord are the rarest Isolated systemic (extra-CNS) dissemination is
manifestation of PCNSL. These lymphomas often arise observed in 3–7% of failures among patients with
in the upper thoracic or lower cervical regions of the PCNSL; this rare condition seems to be associated
spinal cord. Presenting symptoms depend on the spinal with a significantly better outcome with respect to
cord level involved. Myelography shows a widened CNS relapses, in particular, if treated with conven-
spinal cord. MRI reveals hyperintensity on T2-images tional antilymphoma chemotherapy.
and homogeneous enhancement after gadolinium
administration on T1-weighted images. Increased CSF CNS involvement in HIV-related
protein concentration is common. Lymphomas arising
in the spinal nerves and ganglia (“neurolymphomato- lymphomas
sis”), cauda equina, and the sciatic nerve are extremely PCNSL in immunocompromised patients has differ-
rare and should be distinguished from neural infiltra- ent characteristics and behavior compared to the
tion by a systemic lymphoma. Prognosis of patients immunocompetent population (Table 13.6). EBV
with spinal cord lymphoma is poor, mainly because of infection and c-myc translocation result in the prolif-
delayed diagnosis. Therapeutic results are strongly eration of malignant lymphocytes in HIV patients; a
affected by pre-treatment neurological status. Similar CD4 count <50 cells/μL and high peripheral HIV viral
to other PCNSL, corticosteroids and radiotherapy are load are the most significant risk factors for PCNSL
associated with lymphoma regression and clinical development. Since the introduction of highly active
improvement. Only sparse data on chemotherapy effi- antiretroviral therapy (HAART), the incidence of 217
cacy are available; however, spinal cord lymphomas AIDS PCNSL has declined.
Table 13.5 Salvage regimens for PCNSL.
Treatment Study No. Median Prior PS CR + mPFS mOS 1-yr N PLT Other
age RT 0–1 PR OS toxicities
Topotecan P 27 51 52% 60% 19+14% 2.0 8.4 39% 26% 15% 11%
Temozolomide P 36 60 86% 28% 25+6% 2.8 4.0 31% 6% 3% 3%
Methotrexate R 22 58 14% – 73+19% 26 26 70% 5% 5% 36%
Temozolomide + rituximab R 15 69 13% 67% 40+13% 2.2 10.5 58% 7% 27% 7%
VP16 + ifosfamide + araC R 16 54 100% 37% 37+0% 4.5 6.0 41% 69% 50% 37%
i. a. carboplatin ± VP16 ± CTX ± RT R 37 57 24% 76% 24+11% 3.0 6.8 25% 22% 19% >30%
Radiotherapy R 27 67 – – 37+37% 9.7 10.9 49% – – 15% neuro
Radiotherapy R 20 – – – 60 + NR – 19.0 – – – 58% neuro
%
HD-araC + VP16 → TTP + busulfan P 43 52 33% – 50 + 0% – 18 – – – TRM: 1%
+ CTX
Intrathecal rituximab P 10 56 80% – 0 + 60% – 5.2 30%
Study: P, prospective; R, retrospective; RT, radiotherapy; PS, performance status; CR, complete response; PR, partial response; mPFS, median progression-free survival; mOS, median overall
survival; 1-year OS, 1-year overall survival; N, grade 3–4 neutropenia; PLT, grade 3–4 thrombocytopenia; tox, toxicity; i.a., intra-arterial; TRM, treatment-related mortality; VP16, etoposide; HD-
araC, high-dose cytarabine; TTP, thiotepa; CTX, cyclophosphamide.
Chapter 13: Central nervous system lymphomas
AIDS PCNSL is typically ring-enhancing on MRI. poorly and develop more infectious complications.
Cerebral toxoplasmosis, abscesses, and progressive Bone marrow toxicity and opportunistic infections fre-
multifocal leukoencephalopathy are the main differ- quently complicate treatment regimens. In some series,
ential diagnoses. Single-photon emission computed up to 40% of patients did not receive any treatment.
tomography (SPECT), PET, and proton MR spectro- Before HAART, the mean survival was 1–2 months
scopy may distinguish tumor from infectious pro- without treatment. WBRT improved OS to 3–4 months.
cesses, but are not always definitive. Given the variety Since the advent of HAART, the ability to deliver
of intracranial diseases indistinguishable by imaging, a adequate chemotherapy to AIDS patients with PCNSL
brain biopsy is often required for diagnosis, particu- has improved. However, only 10% of AIDS patients
larly if the patient has rapid neurologic deterioration, with PCNSL are eligible for chemotherapy because of
negative CSF cytology, negative toxoplasma serology, comorbidity or poor neurological condition. A few cases
radiographic features atypical for toxoplasmosis, and of HIV-positive patients with PCNSL successfully treat-
progressing symptoms within the first 1–2 weeks of ed with HAART, with or without other specific antitu-
antitoxoplasmosis therapy. Surgical biopsy could be mor therapy, have been published. Therefore, HIV-
avoided in AIDS patients with brain lesions that have positive patients with newly diagnosed PCNSL should
both increased thallium uptake on SPECT scanning be started on HAART (or have their HAART regimen
and a positive CSF for EBV DNA. The combination of optimized) and receive appropriate prophylaxis for
these procedures is associated with 100% sensitivity opportunistic infections. It is possible that this approach
and specificity. SPECT is insufficient as a sole diag- will be adequate in patients who are antiretroviral drug-
nostic modality and may give a false-negative result in naïve. However, most patients will require additional
lesions <0.6 cm, located near the skull or ependyma, or specific antitumor therapy. Given the likelihood that
after steroid therapy. WBRT may exacerbate or accelerate the risk of HIV-
Because they are profoundly immunocompromised, related dementia, it seems most appropriate to initiate
these patients have a much worse prognosis than non- therapy with MTX-based chemotherapy, similar to the
AIDS patients. They tolerate cytotoxic chemotherapies approach in immunocompetent patients. HD-MTX,
Table 13.6 Clinical, radiological, and histological features of PCNSL in HIV-infected patients.
alone or in combination, with or without intrathecal risk of CNS relapse ranges between 5% and 30% in
drug delivery, has been associated with a response rate all NHL subtypes. The most relevant clinical aspects
of 33–57% and a median survival of 3–4 months. The of this heterogeneous condition are summarized in
combination of oral zidovudine and HD-MTX admin- Table 13.7. Given the high morbidity and mortality
istered weekly for three to six cycles resulted in a 46% associated with CNS dissemination of NHL, a pro-
CRR and a median survival of 12 months. In contrast, phylactic strategy, analogous to that employed suc-
the use of CHOP followed by WBRT has been associ- cessfully for acute lymphoblastic leukemia (ALL), is
ated with disappointing results. In a small group of indicated. However, the incidence of CNS recurrence
highly selected patients (lesions limited to the brain in NHL is not sufficiently high to warrant the use of
and high CD4 counts), WBRT associated with hydroxy- CNS prophylaxis in all patients. The analysis of risk
urea followed by PCV (procarbazine, lomustine, and factors for CNS relapse in NHL is biased by differ-
vincristine) chemotherapy was associated with a 100% ences in the definition criteria and retrospective
response rate and a median survival of 13 months. nature of reported studies. Nevertheless, advanced
Experimental strategies that directly target EBV to disease, increased LDH serum levels, certain extra-
induce lymphoma remission are under investigation. nodal sites of disease, and highly aggressive lympho-
These approaches include direct antiviral therapy (zido- mas (Burkitt and lymphoblastic lymphomas) have
vudine, ganciclovir, and interleukin-2), monoclonal been associated with an increased risk for CNS
antibodies, cytotoxic T-lymphocytes, episomal replica- recurrence. The risk is increased for patients with a
tion inhibition by hydroxyurea or antisense oligonu- high IPI score, particularly those with a high serum
cleotides, pro-drug activation by tumor cells, and LDH level, or involvement of more than one extra-
kinase expression upregulation and lytic induction. nodal site, but CNS recurrence occurs in all the IPI
However, clinical experience is still limited and successes risk groups. Patients displaying involvement of mul-
are exceptional. Depleting the viral reservoir (i.e. tiple extranodal organs and increased LDH serum
B-lymphocytes) in high-risk patients by intravenous levels have a 1-year risk for CNS recurrence of 20%.
rituximab could play a relevant role in this setting. These patients, as well as those with testicular or
paranasal sinus involvement, should be assessed for
CNS involvement at diagnosis.
Secondary CNS lymphomas In highly aggressive lymphomas, the 5-year CNS
recurrence rate was substantially higher among
Prophylaxis against CNS relapse patients who did not receive CNS prophylaxis
CNS dissemination is an almost uniformly fatal (32–78%) in comparison to patients who did (19%).
220 complication of hematological malignancies. The This wide variation may be explained by a variable
Chapter 13: Central nervous system lymphomas
proportion of T-cell lymphoblastic lymphoma (LbL) hydrocortisone) twice a week using an Ommaya res-
patients in the study cohorts. In fact, patients with ervoir. Although this strategy invariably results in a
T-cell LbL or T-cell ALL have a higher risk for CNS decrease in the number of tumor cells in the CSF,
relapse. The inclusion of CNS-directed prophylaxis symptomatic improvement occurs in <20% of cases.
led to a substantial reduction in CNS relapse rates in Slow-release formulation of cytarabine, injected once
LbL patients; 0–36% with intrathecal chemotherapy every 2 weeks, produces a higher response rate and a
alone, 3–21% with intrathecal chemotherapy and better quality of life relative to that produced by free
WBRT, and <5% with intrathecal chemotherapy, cytarabine injected twice a week except for the high
WBRT, and systemic HD-MTX and HD-araC. incidence of chemical meningitis associated with lip-
For aggressive lymphomas, CNS prophylaxis is osomal araC. Patients with focal neurological deficits
currently used in 10–15% of cases; however, there is or intraparenchymal lesions in the brain, cranial
still no uniform practice, which reflects the fact that nerves, or spinal cord are commonly treated with
published data are open to different interpretations. radiation therapy, which results in symptomatic
Intrathecal chemotherapy and the addition of improvement in >65% of cases; however, CNS or
systemic HD-MTX or HD-araC to conventional che- systemic progression is the rule. The optimal approach
motherapy reduces the CNS relapse rate and appears to managing overt CNS disease at diagnosis in chil-
to be as effective as and less toxic than WBRT. dren with large-cell lymphoma is controversial
Incorporating both intrathecal MTX and consolida- because of the low frequency and marked heterogen-
tion with systemic HD-MTX, ifosfamide, and cytar- eity of this group of malignancies. In children with
abine significantly reduced CNS recurrence from Burkitt lymphoma, HD-MTX and cytarabine and
8.3% to 2.7% in NHL patients aged 61–69 years intrathecal chemotherapy are currently used for both
with IPI ≥1, in comparison with standard CHOP. CNS prophylaxis and treatment.
The use of conventional-dose rituximab to prevent Some of the most successful treatment regimens
CNS relapse in patients with DLBCL treated with against highly aggressive lymphomas have excluded
CHOP has been associated with conflicting results WBRT. In children, there is a trend toward reducing
in two randomized trials. the use of radiotherapy for CNS disease and prophy-
laxis to reduce late sequelae such as second cancer,
endocrinopathy, and neuropsychologic defects. In
Treatment of SCNSL recent clinical trials, the prophylactic WBRT dose
In a large proportion of patients, CNS relapse is has been reduced to 12 Gy, and the therapeutic
accompanied or rapidly followed by systemic relapse, dose to 18 Gy, while the safety of WBRT omission
and the patient’s outcome is determined equally by is being assessed in current trials on LbL in children.
the control of systemic and CNS disease. Systemic In recent trials, the elimination of WBRT has not
high-dose chemotherapy is the most effective strategy been associated with an excessive rate of CNS
that meets these requirements. Similar to PCNSL, the relapse.
choice of drugs is based on the ability to cross the
BBB and have antilymphoma activity. Most first line Autologous transplantation
combinations against SCNSL contain HD-MTX or Some reports conclude that 20–40% of adults with
HD-araC, but the most effective administration NHL and CNS disease can achieve durable remissions
schedules for these drugs in SCNSL remain to be after HDC/ASCT, primarily patients with highly
defined. Most lymphoma patients with a high risk aggressive lymphomas in first remission at the time
of CNS relapse may receive these drugs as first line of transplantation. On the other hand, attempts to
treatment, so salvage options may be limited because transplant adults with active CNS disease at the
few other drugs cross the BBB. Some anecdotal expe- moment of autologous transplantation have had dis-
rience with cisplatin-based regimens has been appointing results, with a PFS of 9% at 71 months.
reported, and autologous or allogeneic transplanta- Moreover, it is difficult to estimate the exact contri-
tion may prove beneficial for selected patients. bution of autologous transplantation to survival in
Patients with leptomeningeal involvement without patients with highly aggressive lymphomas and CNS
focal neurological deficits are usually treated with involvement at diagnosis, which may be curable with
intrathecal chemotherapy (MTX, cytarabine and conventional-dose treatment. In some series, overall 221
Chapter 13: Central nervous system lymphomas
cerebral non-Hodgkin lymphoma: cerebral radiotherapy Reni M, Mazza E, Foppoli M, Ferreri AJ. Primary central
with and without cyclophosphamide, doxorubicin, nervous system lymphomas: salvage treatment after
vincristine, and prednisone chemotherapy. Cancer, failure to high-dose methotrexate. Cancer Lett,
2000;89(6):1359–1370. 2007;258(2):165–170.
Ponzoni M, Terreni MR, Ciceri F, et al. Primary brain CD30 Reni M, Zaja F, Mason W, et al. Temozolomide as salvage
+ ALK1+ anaplastic large cell lymphoma (‘ALKoma’): the treatment in primary brain lymphomas. Br J Cancer,
first case with a combination of ‘not common’ variants. 2007;96(6):864–867.
Ann Oncol, 2002;13(11):1827–1832. Shenkier TN, Blay JY, O’Neill BP, et al. Primary CNS
Ponzoni M, Ferreri AJ, Campo E, et al. Definition, diagnosis, lymphoma of T-cell origin: a descriptive analysis from the
and management of intravascular large B-cell lymphoma: international primary CNS lymphoma collaborative
proposals and perspectives from an international group. J Clin Oncol, 2005;23(10):2233–2239.
consensus meeting. J Clin Oncol, 2007;25(21):3168–3173. Smith JR, Rosenbaum JT, Wilson DJ, et al. Role of
Reni M, Ferreri AJ, Villa E. Second-line treatment for intravitreal methotrexate in the management of
primary central nervous system lymphoma. Br J Cancer, primary central nervous system lymphoma with
1999;79(3–4):530–534. ocular involvement. Ophthalmology,
Reni M, Ferreri AJ, Guha-Thakurta N, et al. 2002;109(9):1709–1716.
Clinical relevance of consolidation radiotherapy and Thiel E, Korfel A, Martus P, et al. High-dose methotrexate
other main therapeutic issues in primary central nervous with or without whole brain radiotherapy for primary
system lymphomas treated with upfront high-dose CNS lymphoma (G-PCNSL-SG-1): a phase 3,
methotrexate. Int J Radiat Oncol Biol Phys, randomised, non-inferiority trial. Lancet Oncol,
2001;51(2):419–425. 2010;11(11):1036–1047.
225
Chapter
14
T-cell non-Hodgkin lymphoma
Sheetal M. Kircher, Beverly P. Nelson, Steven T. Rosen,
and Andrew M. Evens
226 Lymphoma, Second Edition, ed. Robert Marcus, John W. Sweetenham, and Michael E. Williams. Published by
Cambridge University Press. © Cambridge University Press 2014.
Chapter 14: T-cell non-Hodgkin lymphoma
Table 14.1 WHO classification of T-cell and NK-cell neoplasms switching, and somatic hypermutation. T-cell neo-
(2008). plasms may have rearrangements involving the site
of TCR α and δ genes on chromosome 14, or, more
Mature T-cell and NK-cell neoplasms
rarely, chromosome 7 (7q34–36 and 7p15), the site of
Peripheral T-cell lymphoma, NOS
TCR β and γ genes. Many of the genes located at the
Angioimmunoblastic T-cell lymphoma breakpoints of recurring chromosome translocations
Anaplastic large-cell lymphoma, ALK+ have been identified. The majority of translocated
Anaplastic large-cell lymphoma, ALK–* genes encode transcription factors. Transcription fac-
tors are involved in the initiation of gene transcrip-
Adult T-cell leukemia/lymphoma
tion and cell differentiation. The most common result
Extranodal NK-/T-cell lymphoma, nasal type
of chromosome translocations in lymphomas that
Enteropathy-associated T-cell lymphoma involve TCR and/or Ig genes is deregulation of gene
Subcutaneous panniculitis-like T-cell lymphoma expression with irregular expression or overexpres-
Primary cutaneous CD8+ gamma-delta T-cell lymphoma* sion in cells that normally do not express this gene.
Systemic EBV+ T-cell lymphoproliferative diseases of childhood
Few lymphomas have been recognized to contain
translocations that produce a fusion protein, such as
Hydroa vacciniforme-like lymphoma
the t(2;5)(p23;q35) translocation in ALCL that results
Aggressive NK-cell leukemia in expression of the nucleophosmin (NPM)–ALK
T-cell prolymphocytic leukemia protein.
T-cell large granular lymphocytic leukemia Inactivation of tumor suppressor genes may also
Chronic lymphoproliferative disorder of NK-cells*
play a role in lymphogenesis. The most common
mechanism of tumor suppressor inactivation occurs
Mycosis fungoides
through the Knudson two-hit model, where a reduc-
Sézary syndrome tion of homozygosity leads to tumor formation (for
Primary cutaneous CD30+ T-cell lymphoproliferative disorders* example, following germline deletion of one allele and
Primary cutaneous CD4+ small/medium T-cell lymphoma somatic mutation of the other). Tumor suppressor
* Provisional entities for which the WHO Working Group felt genes identified with NHL include p53, p15, and p16.
there was insufficient evidence to recognize as distinct diseases Moreover, specific chromosomal deletions that have
at this time. been detected in NHL (including some T-cell lympho-
NOS, not otherwise specified; ALK, anaplastic lymphoma kinase;
NK, natural killer; EBV, Epstein–Barr virus. mas), such as 3p, 6q, 13q, and 17p, may represent sites
of yet to be identified tumor suppressor loci. Other
mediators that may be involved in lymphogenesis
include the cyclin-dependent kinase (cdk) inhibitors,
of the proto-oncogene by a promoter associated with such as p21(Waf1). A function of the p21(Waf1) protein
an immunoglobulin (Ig) or T-cell receptor (TCR) includes the arrest of cells in G1-phase checkpoint by
gene. The two subtypes of TCRs that T-cell lympho- associating with cyclin–CDK complexes, but the exact
cytes express are gamma-delta (γδ) or alpha-beta factors critical for apoptosis have not been clearly
(αβ). Approximately 95% of normal T-lymphocytes defined. The gene for the p21(Waf1) protein has been
express the αβ heterodimer, while the minority of identified as a downstream target of p53 in regulating
T-lymphocytes express the γδ heterodimer. Alpha- cell cycle progression through G1. Induction of
beta T-cells develop predominantly in the thymus, p21(Waf1) has also been demonstrated to occur through
while γδ T-cells may develop in extrathymic locations a p53-independent pathway.
such as the skin, intestinal epithelium, and Gene amplification leads to an increase in the
spleen. The four TCR genes are arranged in germline number of copies of a gene in the genome of a cell,
configuration in non-continuous segments of varia- which may contribute to lymphogenesis. Gene
ble (V), diversity (D), joining (J), and constant (C) amplification has been identified mostly in B-cell
regions. The precise mechanism by which transloca- lymphomas (e.g. REL gene), although amplification
tion of TCR and Ig genes occur is not known, but it of TCR genes in varied T-cell lymphomas has been
appears to in part involve dysfunctional gene remod- described. Random genomic instability, as seen in
eling, including V–D–J recombination, isotype many epithelial cancers, is not a characteristic of 227
Chapter 14: T-cell non-Hodgkin lymphoma
the more stable lymphoma genome. Defects in DNA activated cells as opposed to resting cells. Many
mismatch repair that manifest as genomic microsa- features of NK/TCL indicated perturbation of angio-
tellite instability are also less frequently recognized in genic pathways, including overexpression of platelet-
lymphoma, as compared to various hereditary solid derived growth factor receptor α. In addition, they
tumor syndromes and rare sporadic cancers. identified that there is evidence of dysregulation of
Recently, gene expression profiling has helped to the tumor suppressor HACE1 in the frequently
identify emerging oncogenic pathways and unique deleted 6q21 region. These molecular signatures of
molecular subgroups within PTCL. Iqbal et al. pub- T-cell lymphomas will need to be validated in larger
lished results from 144 cases of PTCL and NK/TCLs, populations.
including AITL (n = 36), ALK + ALCL (n = 20), ALK–
ALCL (n = 6), ATLL (n = 12), NK/TCL (n = 14), PTCL-
NOS (n = 44), and others (n = 10), who had gene
Prognosis
Even in the pre-rituximab era, peripheral T-cell lym-
expression profiling. They were able to identify a unique
phomas had inferior OS rates compared with their
molecular subgroup of PTCL-NOS. This subgroup
aggressive B-cell counterparts (e.g. diffuse large B-cell
showed marked upregulation in normal activated CD8
lymphoma [DLBCL]). As noted before, T-cell lympho-
+ T-cells as well as a distinct set of cytokines associated
mas are relatively uncommon and many clinicopatho-
with CD8+ T-cells and NK-cells. AITL cases had more logic subtypes exist. A number of studies have shown
frequent activation of the NF-κβ pathway as compared that the T-cell immunophenotype alone, with the
with PTCL-NOS (P < 0.001). Immunosuppressive path- exception of ALK+ ALCL, is an independent adverse
ways were enriched with increased IL-6 signaling as well prognostic indicator. Melnyk et al. found that T-cell
as other gene signatures linked with angiogenesis. In lymphomas had a significantly inferior complete remis-
ALK+ ALCL, the expression of IL-17-associated mole- sion, failure-free survival (FFS), and OS when com-
cules was present. ATLL had a homogeneous molecular pared with DLBCL. Over the years, treatment
signature with high expression of HTLV type I genes. paradigms and prognostic indicators have frequently
In this study, they also constructed a molecular been extrapolated from aggressive B-cell NHLs. These
prognosticator of AITL in which they identified genes aforementioned findings suggest that there may need to
that were expressed differently between AITL cases be different prognostic models and treatment regimens
with less than a 3-year overall survival (OS) (poor for T-cell lymphomas. There have been several studies
prognosticator) and those with a greater than 3-year attempting to validate previous prognostic models and
OS (good prognosticator). Genes upregulated in the to create new systems of risk stratification (Table 14.2).
poor prognosis group included two gene signatures The International Prognostic Index (IPI) was ini-
associated with immunosuppressive functions: tol- tially used to predict survival of patients with aggres-
erogenic DCs43 and CD31 stromal cells. Cases with sive B-cell lymphomas but has since been used to
a poor prognosis also showed high expression of risk-stratify T-cell lymphomas as well. The IPI is a
receptors or cell adhesion molecules that mediate prognostic indicator based on disease stage, lactate
proliferative signals, including PDGFR. In contrast, dehydrogenase (LDH), age, extranodal sites, and per-
transcripts that have inhibitory effects on myeloid cell formance status (PS). Sonnen et al. conducted a ret-
functions (CD200, MIF, SERPINB1), associated with rospective study of 125 patients diagnosed with
B-cells (SpiB, BTLA4, SYK), or encoded members of ALCL, PTCL-NOS, and AITL in an effort to deter-
the ribosomal protein synthesis pathway were highly mine whether IPI can be used to predict survival in
expressed in the good prognostic group. Huang et al. T-cell lymphoma patients. It confirmed that patients
published results whereby biopsies and cell lines of with T-cell lymphomas markedly differ in their pre-
NK/TCL, nasal type, underwent combined gene sentation of disease, their response to treatment, and
expression profiling and array-based comparative OS. In multivariate analysis, the IPI, not the histo-
genomic hybridization (CGH) analysis. NK/TCL logic subtype, was the main factor in determining
samples were compared to both normal NK-cells overall prognosis. Furthermore, other studies have
and PTCL-NOS. As compared to PTCL-NOS, NK/ shown that higher IPI scores, elevated LDH,
TCL had greater transcripts for NK-cell and cytotoxic B-symptoms, and elevated LDH were all individually
228 molecules, especially granzyme H. As compared with associated with inferior OS. Most recently, it was
normal NK-cells, tumors more closely resembled shown that tumor score, elevated beta-2
Chapter 14: T-cell non-Hodgkin lymphoma
microglobulin, and bulky disease were associated factors, and group 3 are those patients with three or
with inferior OS. four factors. Median survival for patients in group 1
The prognostic index for PTCL (PIT) is a prog- was 37 months, 23 months for group 2, and 6 months
nostic model used to risk-stratify patients with for group 3.
PTCL-NOS (excluding other clinicopathologic The International Peripheral T-cell Lymphoma
T-cell lymphoma subtypes). It takes into considera- Project (IPTCLP) recently proposed another model
tion four variables: age >60 years, PS (2 or higher), for risk stratification of PTCL and AITL which takes
elevated LDH (>1× upper limit of normal [ULN]), into consideration age >60, PS > 2, and platelet count
and bone marrow involvement, compiling these fac- <150 as adverse indicators. Patients were classified
tors into a score. A score of 1 is assigned for no into low risk with no risk factors, low–intermediate
adverse factors, 2 for patients with one factor, 3 for if they had one risk factor, high–intermediate or
patients with two factors, and 4 for patients with high risk if they had two or three risk factors, respec-
three or four adverse factors. A score of 1 correlated tively. Corresponding 5-year OS rates were 58%
with a 63% 5-year survival rate while a score of 3 for low-risk patients, 15% for low–intermediate
corresponded with a 53% survival rate. Scores 3 and risk, 5% for high–intermediate, and 0% for high-risk
4 corresponded with 32% and 18% survival rates, patients.
respectively. Gallamini et al. found this model to be Recently, a comparison of these four prognostic
more predictive compared with the IPI. scores was completed by Garcia et al. Concordance
The marker expression in PTCL (mPIT) is a prog- among IPI, PIT, and IPTCLP was 52% for the low-risk
nostic model proposed by Went et al. that integrates group, 27% for the low–intermediate group, 20% for
biologic factors into the model when assigning risk for the high–intermediate group, and 14% for the high-
patients with PTCL-NOS and AITL. The model is risk group. They found that all the scores demonstra-
composed of PS > 2, age >60 years, elevated LDH, ted usefulness in assessing the outcomes of patients
and Ki67 ≥80%. Group 1 has zero or one of the with PTCL, with IPTCLP being most significant for
above factors, group 2 are those patients with two prediction of OS in multivariate analysis. 229
Chapter 14: T-cell non-Hodgkin lymphoma
Specific disease types characteristically positive for CD25. Some cases may
show expression of CD30, particularly in the large
Specific disease types are summarized in Table 14.3.
cells. Stains for TIA-1 and granzyme are usually
negative.
Adult T-cell leukemia/lymphoma (ATL)
The retrovirus HTLVI is critical to the development of Molecular pathology and cytogenetics
ATL. In endemic areas in Japan, approximately 6–37% The prerequisite for the diagnosis is the demonstration
of the population is infected with HTLVI. The USA and of integrated HTLVI genomes in virtually all cases.
Europe are considered low-risk areas, with less than 1% TCR genes are clonally rearranged. Some recurring
of the population who are seropositive. Approximately structural and numerical chromosome aberrations
2–4% of HTLVI carriers develop ATL. HTLVI is trans- have been reported, among them trisomy for chromo-
mitted through sexual intercourse, blood products con- somes 3p, 7q, and 14q, and losses in 6q. These changes
taining white blood cells, shared needles, breast milk, are predominantly encountered in aggressive variants.
and vertically. Transfusion of HTLVI-contaminated
blood products results in seroconversion in approxi- Subtypes
mately 30–50% of patients at a median of 51 days. The ATL is classified into four subtypes based on clinicopa-
median age at presentation is 55 years. Patients present thologic features and prognosis: acute, lymphoma,
with lymphadenopathy (72%), skin lesions (53%), hep- chronic and smoldering. In 1991, Shimoyama and col-
atomegaly (47%), splenomegaly (25%), and hypercalce- leagues reported on the characteristics of 818 ATL
mia (28%). Cellular immune suppression is common; a patients. Patients with the acute type present with hyper-
significant minority of patients may have concurrent calcemia, leukemic manifestations, and tumor lesions,
Strongyloides infection. and had the worst prognosis, with a median survival of
approximately 6 months. Lymphoma-type patients
Pathology present with low circulating abnormal lymphocytes
The neoplastic T-cells are characteristically medium to (<1%) and nodal, liver, splenic, central nervous system,
large, with pleomorphic nuclei that show convolutions bone, and gastrointestinal disease, and had a median
or lobulations that may be pronounced (flower cells) survival of 10 months. Chronic-type patients present
(Figure 14.1). In the most common acute variant, these with >5% abnormal circulating lymphocytes and had a
abnormal lymphocytes are seen in the peripheral blood median survival of 24 months, while the median survival
smear. In chronic and smouldering variants the cells of smoldering type had not yet been reached.
may be predominantly small, with round nuclei. The
cytoplasm is basophilic. Giant cells may be seen. The Treatment
skin shows nodular, diffuse, or perivascular dermal ATL, acute/leukemic, and lymphoma subtypes are dif-
infiltration, frequently with extension of clusters of ficult malignancies to treat. Patients may initially
atypical cells into the epidermis. Bone marrow infiltra- respond to combination chemotherapy, but OS is
tion is often patchy. Increased osteoclastic activity may poor, with a median of 8 months. Response rates of
be seen in bone marrow biopsies even in the absence of 70–90% to combination IFN-α and zidovudine therapy
marrow infiltration. In some cases large cells resem- have been demonstrated in both the leukemic and lym-
bling the Reed–Sternberg cells of Hodgkin lymphoma phoma subtypes of ATL, with associated median sur-
may be seen in involved lymph nodes. In most cases vival rates of 11–18 months. A clinical trial that
these are CD30-positive B-cells that may also express investigated initial cytoreductive therapy with cyclo-
CD15 and they show evidence of EBV infection. These phosphamide, doxorubicin, vincristine, and prednisone
are likely to be reactive and a reflection of the immu- (CHOP) followed by antinucleoside (zidovudine or zal-
nosuppression associated with the lymphoma. citabine), IFN-α, and oral etoposide therapy demon-
Immunophenotypically the cells are mature T-cells strated encouraging results. A Japanese phase III trial
with no expression of terminal deoxynucleotidyl trans- of 118 patients compared biweekly CHOP versus dose-
ferase (TdT) or CD1a. There is usually expression of intensified multiagent chemotherapy: VCAP (vincris-
CD2, CD3, and CD5, but lack of CD7 is common. Most tine, cyclophosphamide, doxorubicin, prednisolone),
230 cases are CD4+/CD8– but CD4+/CD8+ and CD4–/ AMP (doxorubicin, ranimustine, prednisolone), and
CD8+ cases are rarely seen. The cells are VECP (vindesine, etoposide, carboplatin, prednisolone)
Table 14.3 Immunophenotype, EBV status, and genetic features of peripheral T-cell NHL.
Neoplasm CD3 CD5 CD7 CD4 CD8 CD25 CD30 CD52 CD103 TCR EBV Genetic T-receptor
s; c abnormality genes
Adult T-cell leukemia/ + + − + − + −/+ + − NA − Multiple Rearranged
lymphoma
Peripheral T-cell +/− +/− +/− +/− −/+ NA −/+ −/+ − αβ>γδ − Often Rearranged
lymphoma, NOS complex
Angioimmunoblastic + + + +/− −/+ NA −* −/+ − αβ Present Trisomy 3 and Rearranged
T-cell lymphoma in 5; additional X
lymph
nodes
Anaplastic large-cell +/− +/− NA −/+ −/+ + + − − αβ − Table 14.5 Rearranged
lymphoma, primary
systemic type (ALK+
and ALK–)
Subcutaneous + + + − + NA − NA NA αβ – − Rearranged
panniculitis-like T-cell
lymphoma
Cutaneous γδ T-cell + − +/− − − NA −/+ NA NA γδ − − Rearranged
lymphomas
Hepatosplenic γδ T- + − + − − NA − NA NA γδ>αβ − i(7q) and tris- Rearranged
cell lymphoma omy 8
Extranodal NK-/T-cell −;+ − −/+ − − NA − NA NA − + del 6 and 13 −
lymphoma, nasal type
Enteropathy-type T- + + + − +/− NA +/− NA +/− αβ>γδ − *LOH 9p21 Rearranged
cell lymphoma
* T-cells are CD30-negative; however, B-cells are typically CD30+.
NOS, not otherwise specified; ALK, anaplastic lymphoma kinase; NK, natural killer; EBV, Epstein–Barr virus; αβ, alpha-beta; γδ, gamma-delta; LOH, loss of heterozygosity; del, deletion; I, isochrome;
NA, not available; +, ≥90%; −, ≤90%; +/−, ≥50% positive; −/+, ≤50% negative; s, surface; c, cytoplasmic.
Chapter 14: T-cell non-Hodgkin lymphoma
Figure 14.1 (a) Adult T-cell lymphoma/leukemia cells in the blood. The cells have condensed chromatin with folded nuclei that evokes the
term “flower cells.” Both cells have an ample amount of vacuolated cytoplasm. (b) Adult T-cell lymphoma/leukemia involving lymph node. The
large lymphoma cells have several visible nucleoli and abundant cytoplasm. The lymphoma shows a high proliferative rate, as evidenced by
many mitoses and histiocytes that impart a “starry sky” appearance.
and intrathecal prophylaxis. The complete response denileukin diftitox (target CD25), all of which have
(CR) rate was higher in the VCAP–AMP–VECP arm been shown to have an effect on ATL cell cycle
compared with biweekly CHOP arm, with 40% versus progression.
25%, respectively (P = 0.02). The OS at 3 years was 24%
in the VCAP–AMP–VECP arm and 13% in the CHOP Peripheral T-cell lymphoma, not
arm. Subgroup analysis suggested that younger patients
(age <56 years) and patients with a poorer PS (1–3) had otherwise specified (PTCL-NOS)
more benefit with VCAP–AMP–VECP therapy, PTCL-NOS is predominantly a nodal lymphoma that
although small numbers precluded a definitive represents the most common T-cell lymphoma subtype
conclusion. Toxicity was significantly increased with in western countries, comprising approximately 50–60%
VCAP–AMP–VECP compared with biweekly CHOP, of T-cell lymphomas and 5–7% of all NHL. PTCL-NOS
with rates of grade 4 neutropenia, grade 4 thrombocy- usually affects male adults (1.5:1 male:female ratio), with
topenia, and grade 3 or 4 infection of 98% versus 83%, a median age of 61 years (range 17–90) in a large study
74% versus 17%, and 32% versus 15%, respectively. In with 25% of patients presenting in stage I or IIE, 12%
addition, three treatment-related deaths were reported stage III, and 63% stage IV. PTCL-NOS patients from
in the VCAP–AMP–VECP arm. this study commonly presented with unfavorable char-
There is also a role for combination arsenic triox- acteristics, including B-symptoms (40%), elevated LDH
ide (As2O3) with IFN-α therapy, which induces cell (66%), bulky tumor ≥10 cm (11%), non-ambulatory PS
cycle arrest and apoptosis both in HTLVI transformed (29%), and extranodal disease (56%), leading to the
cell lines and primary ATL cells in culture. Arsenic majority of patients (53%) falling into the unfavorable
trioxide and IFN-α in combination with zidovudine IPI category (score of 3–5). Other smaller studies of
(AZT) was studied in a phase II trial with ten patients PTCL-NOS patients have corroborated male preponder-
with newly diagnosed chronic ATL with response seen ance, B-symptoms, and extranodal disease (bone mar-
in all patients, including seven CRs. A retrospective row > liver > skin > lung and bone).
study examined CD52 expression rates in various TCL
subtypes and found that approximately 35% expressed Pathology
CD52, which could have clinical implications for use This is a polymorphous group of tumors that includes
of alemtuzumab, a humanized anti-CD52 antibody. many subtypes that were previously identified as spe-
Future research should continue to explore novel cific entities in the Kiel classification of lymphomas.
232 agents, including proteasome inhibitors, all-trans ret- The morphology is varied, but usually consists of a
inoic acid (ATRA), and recombinant toxins, including pleomorphic mix of medium and large cells that
Chapter 14: T-cell non-Hodgkin lymphoma
A B
Figure 14.2 (a) Peripheral T-cell lymphoma, unspecified. The lymphoma cells are small with condensed chromatin and are associated with
many small blood vessels and eosinophils. (b) Peripheral T-cell lymphoma, unspecified, CD3 stain. Numerous lymphoma cells are highlighted.
diffusely efface the lymph node architecture. High Recurring deletions have been described in the short
endothelial venules are prominent and may have an arm of chromosome 1, the long arm of chromosome 6,
arborizing pattern. The background population and the short arm of chromosome 17. Gains are com-
includes a mixture of small lymphocytes, plasma mon in chromosomes 3, 7, and X. PTCL-NOS of the
cells, histiocytes, and eosinophils (Figure 14.2). large-cell type are frequently characterized by a tetra-
In some cases, the infiltrate is perifollicular and the ploid karyotype. Only one recurring translocation, t
neoplastic cells are often small with little nuclear pleo- (5;9)(q33;q22), has been identified so far, fusing ITK
morphism (T-zone variant). In the lymphoepithelioid and SYK genes. Interestingly, this translocation may be
variant (Lennert lymphoma), the neoplastic cells are indicative of a particular PTCL-NOS subtype morpho-
usually small with clusters of epithelioid macrophages logically characterized by a perifollicular and intrafol-
within the background infiltrate. Scattered EBV- licular growth pattern. Studies using conventional CGH
positive Reed–Sternberg-like cells may also be as a platform resulted in the revelation of a plethora of
seen. The recently described follicular variant is char- complex chromosomal alterations resulting in recur-
acterized by intrafollicular/perifollicular clusters of ring imbalances and in the definition of clearly non-
atypical clear cells. Although a t(5;9) involving ITK random minimal overlapping target regions. The most
and SYK genes has been described in a subset of frequent recurrent genetic losses in PTCL-NOS were
cases, this variant is not well defined and shows over- encountered in 13q, 6q, 9p, 10q, 12q, and 5q. Gains were
lapping features with AITL. observed in 7q, 17q, 16p, 8q, 9q, and others. Some of
Immunophenotypically the cells are mature T-cells these imbalances seem to occur simultaneously and
that do not express TdT or CD1a but are positive for may define particular PTCL-NOS subgroups.
CD2, CD3, CD5, and CD7, although partial loss of T-cell
antigen expression is frequent. The majority of cases are Treatment
CD4-positive and negative for CD8, but other combina- A standard therapeutic regimen does not exist for
tions can be seen. In some cases the cells express CD30 PTCL-NOS. Most adult treatment series have been
and these must be distinguished from ALCL. small, retrospective, and/or were reported in an era
where patients may now be classified as a different
Molecular pathology and cytogenetics histology (e.g. Hodgkin disease or DLBCL).
The cytogenetic and molecular genetic constitution of Traditionally, owing in part to the lack of clinical
peripheral T-cell lymphoma, unspecified (PTCL-NOS), trial data and the heterogeneity in existing reports,
is still poorly defined and – in keeping with its variable chemotherapy for T-cell lymphomas has been
morphologic features – no unifying cytogenetic aberra- extrapolated from aggressive B-cell lymphoma data/
tion exists. Classical banding studies are rare, but trials. In many of the prognostic studies discussed 233
have revealed recurring chromosomal aberrations. above, T-cell NHL patients were typically treated in
Chapter 14: T-cell non-Hodgkin lymphoma
the same manner as intermediate-grade B-cell patients salvage non-cross-resistant combination chemotherapy
with anthracycline-based combination chemotherapy, is a valid option (e.g. ESHAP, ICE, DHAP, etc).
including regimens such as CHOP alone or CHOP Furthermore, patients who have chemotherapy-
alternating with DHAP (dexamethasone, cisplatin, sensitive disease to salvage therapy may benefit from
and cytarabine), mBACOD (adriamycin, cyclophos- subsequent high-dose chemotherapy and autologous
phamide, vincristine, bleomycin, methotrexate, and stem cell transplant (SCT), but long-term survival
dexamethasone) or ACVB (adriamycin, cyclophos- rates for this strategy remain modest, as reviewed later.
phamide, vindesine, bleomycin, and prednisone) or There are small series that have showed the feasibility of
VIM-3 (mitoxantrone, ifosfamide, methyl-GAG, teni- allogeneic SCT for relapsed/refractory PTCL-NOS.
poside, prednisone, and methotrexate) or CVP (cyclo- The pyrimidine antimetabolite, gemcitabine, has
phosphamide, vincristine, and prednisone), CHOP or activity in relapsed/refractory T-cell NHL as a single
CHOP with etoposide or COPBLAM (cyclophospha- agent, with reported response rates of 60% in small
mide, vincristine, prednisone, bleomycin, adriamycin, single institution studies. Gemcitabine has also been
and procarbazine). studied in combination with cisplatin and methylpred-
With conventional-dose systemic anthracycline- nisone; among studies using this combination, ORR
containing chemotherapy, the overall response rate (both CR and partial response [PR]) is approximately
(ORR) is approximately 60%, but relapses are frequent 70–75%. Newer and novel therapeutic agents, such as
and the 5-year OS is only approximately 20–30%. pralatrexate, are discussed later.
Immediately prior to the rituximab era, randomized
clinical trials in “aggressive NHL,” including aggres- Angioimmunoblastic T-cell
sive B-cell and T-cell lymphomas, showed that inten-
sified CHOP was associated with superior survival lymphoma (AITL)
compared with classic CHOP-21. Schmitz et al. ana- AITL, also known as angioimmunoblastic lymph-
lyzed 343 patients treated within the German High- adenopathy with dysproteinemia, is one of the more
Grade Non-Hodgkins Lymphoma Study Group common T-cell lymphomas, accounting for 15–20%
(DSHNHL). The histologic subtypes included ALCL, of cases and 2–3% of all lymphomas. Mean age of
ALK+ (n = 78), ALCL, ALK– (n = 113), PTCL-NOS presentation is 57–65, with a slight male predomi-
(n = 70), ATCL (n = 28), and other subtypes (n = 54). nance, and the majority of patients present with stage
Patients were treated with 6–8 courses of CHOP every III or IV disease. AITL is commonly a systemic dis-
14 and every 21 days or CHOEP (cyclophosphamide, ease with nodal involvement with various associated
doxorubicin, vincristine, etoposide, and prednisone) disease features such as organomegaly, B-symptoms
every 14 and every 21 days. Three-year event-free (50–70%), skin rash, pruritis, pleural effusions,
survival (EFS) and OS were 76% and 90% (ALK+ arthritis, eosinophilia, and varied immunologic
ALCL), 50% and 68% (AITL), 46% and 62% (ALK– abnormalities (positive Coombs’ test, cold agglutins,
ALCL), and 41% and 54% (PTCL-NOS), respectively. hemolytic anemia, antinuclear antibodies, rheuma-
For patients <60 years old with normal LDH levels, toid factors, cryoglobulins, and polyclonal hyper-
etoposide improved 3-year EFS to 75% versus 51% gammaglobulinemia). In a retrospective study by
(P = 0.003). In patients over 60 years old, however, Mourad et al., 157 AITL patients were retrieved
there were no significant differences between any from the GELA LNH87–LNH93 randomized clinical
treatment group. Thus, for this patient population, trials. Of these, 147 patients received a CHOP-like
standard CHOP21 remains the standard of care. In a regimen with intensified courses in half of them.
retrospective study evaluating clinical outcome in Overall 7-year survival was 30%. In multivariate anal-
T-cell NHLs, intensive treatments such as hyper- ysis, only male sex (P <0.004), mediastinal lymph-
CVAD/MA (fractionated cyclophosphamide, doxoru- adenopathy (P <0.041), and anemia (P <0.042)
bicin, vincristine, dexamethasone/methotrexate and adversely affected OS. IPI was not found to be a
cytarabine) and/or early transplant produced no better predictor of survival in this study.
results than CHOP; however, numbers were small and
the patient population was mixed. Pathology
In relapsed or refractory disease, there is no standard The pathology of AITL is defined by a polymorphous
234 of care. For younger, good PS patients, therapy with infiltrate involving lymph nodes that show prominent
Chapter 14: T-cell non-Hodgkin lymphoma
A B
Figure 14.3 (a) Angioimmunoblastic T-cell lymphoma involving lymph node. The pale staining lymphoma cells expand the interfollicular area,
and replace a significant portion of the lymph node. (b) Angioimmunoblastic T-cell lymphoma involving lymph node. The lymphoma cells range
from small with condensed chromatin to large cells with distinct visible nucleoli. Histiocytes, eosinophils, and blood vessels are present in the
infiltrate. (c) Angioimmunoblastic T-cell lymphoma involving lymph node; CD20 stain. Clusters of CD20+ B-cells, including some in residual
follicles, are evident. Many of the large lymphocytes are CD20+, indicating that they are B-cells.
proliferation of high endothelial venules and follicular absent. With more extensive infiltration, there is para-
dendritic cell (FDC) meshworks. In established cases cortical expansion with scattered residual follicles
lymph node architecture is effaced, with perinodal that are atrophic or depleted, resembling follicles of
infiltration, but the peripheral sinus is often patent. Castleman disease. In established cases the FDC
Clusters of clear cells previously thought to be essential proliferation is prominent and tends to encircle
for diagnosis are present in less than 50% of cases. The vessels. Although best appreciated by immunohisto-
polymorphous infiltrate comprises small, medium, chemistry, when pronounced the FDC proliferation
and large lymphoid cells, eosinophils, plasma cells, may be seen on conventional H&E stain.
and histiocytes (Figure 14.3). In early cases, there is The neoplastic cells of AITL are mature follicular
preservation of lymph node architecture with infiltra- helper T-cells that do not express CD1a or TdT.
tion of the perifollicular area surrounding hyperplastic They are positive for pan-T-cell antigens (CD2,
follicles. Unless clusters of clear cells are present, CD3, CD5, and CD7) with rare antigen loss. In keep-
this early nodal involvement by AITL is often misdiag- ing with the follicular T-helper cell phenotype they
nosed as reactive and only appreciated retrospectively are CD4-positive and negative for CD8, usually
when the disease relapses with more typical histology. express PD1 and inducible costimulator (ICOS),
Furthermore, in the early cases, the FDC proliferation, and, in the majority of cases, also express CXCL13 235
a hallmark feature of the disease, is subtle or and CD10.
Chapter 14: T-cell non-Hodgkin lymphoma
The background population contains a mixed pop- denileukin diftitox. A French study evaluated 25 eld-
ulation that includes many small CD8-positive T-cells erly AITL patients treated with CHOP plus rituximab.
and plasma cells. The expanded proliferation of FDC The ORR was 80%, with 44% achieving a CR and 36%
can be highlighted, with antibodies against antigens a PR. With a median follow-up of 24 months, the
expressed on these cells (CD21, CD23, and CD35). 2-year progression-free survival (PFS) was 42%, and
Residual mantle zone B-cells may be present in rela- the 2-year OS was 62%. This study demonstrated that
tion to expanded FDC meshworks. The immunosup- the addition of rituximab to CHOP did not seem to
pression that is associated with the disease results in improve outcomes over CHOP alone.
increased numbers of large EBV-positive B-cells in the
vast majority of cases, and in approximately 25% of
cases this may lead to clonal expansion, at times result-
Anaplastic large-cell lymphoma (ALCL),
ing in an EBV-positive DLBCL that may obscure the systemic type, ALK+ and ALK–
underlying T-cell neoplasm. ALCL, primary systemic type, accounts for approx-
imately 1–3% of all NHL. This disease mainly
Molecular pathology and cytogenetics involves lymph nodes, although extranodal sites
Angioimmunoblastic T-cell lymphoma (AITL) rep- may be involved. Additionally, it is critical to separate
resents an entity in which recurring chromosome primary cutaneous ALCL, which is associated with
abnormalities have been defined. TCR genes are excellent long-term survival rates (5-year DFS >90–
rearranged in the majority of cases, but clonal rear- 95%), from systemic ALCL (Table 14.4). Systemic
rangements of IgH genes may also be detected in ALCL may be divided in part based on the expression
20–30% of tumors. By in situ hybridization, EBV of the tyrosine kinase anaplastic lymphoma kinase
early repeat (EBER) transcripts have been noted in (ALK). When heterogeneous patient populations
varying numbers of infected cells (60–95% of cases). are analyzed, the prevalence of ALK-positivity in
Recurring chromosomal aberrations in AITL are primary systemic ALCL cases is 50–60%. ALK-
trisomies 3 and 5 and, less frequently, an additional positive ALCL is typically diagnosed in men prior to
X chromosome. Some of these aberrations may be of age 35 (male:female ratio 3:0) with frequent systemic
prognostic importance. symptoms, extranodal, and advanced-stage disease.
ALK-negative patients are usually older (median
Treatment age, 61 years) with a male:female ratio of 0.9, and
AITL typically follows an aggressive clinical course, are less likely to present with extranodal disease. In a
although spontaneous regression occurs on rare occa- more recent series, Savage et al. confirmed the
sions. Treatment with anthracycline-based combina- younger median age for ALK-positive versus ALK-
tion chemotherapy results in complete remission rates negative ALCL (34 years versus 58 years, respectively;
of 50–70%; however, only 10–30% of patients are long- P < 0.001), while the male:female ratio was similar
term survivors. In one prospective, non-randomized between ALK groups.
multicenter study, newly diagnosed “stable” AITL The determination of ALK-positivity is important
patients were treated with single agent prednisone as it connotes a significant favorable prognosis, with
whereas patients presenting with “life-threatening” reported 5-year OS rates of 70% versus 49% for ALK-
disease or relapsed/refractory disease received combi- negative ALCL cases. Moreover, the prognosis for
nation chemotherapy. The CR rate was 29% with ALK-positive and ALK-negative ALCL groups may
single agent prednisone while CR rates for relapsed/ be further divided based on CD56 positivity (neural
refractory patients or patients treated initially with cell-adhesion molecule), which portends a signifi-
combination chemotherapy were 56% and 64%, cantly inferior outcome when it is expressed in either
respectively. With a median follow-up of 28 months, ALCL subgroup (Table 14.5).
the OS and disease-free survival (DFS) was 40.5% and
32.3%, respectively, although median OS was 15 Pathology – ALK-positive
months. There are anecdotal reports of relapsed Although the mix of cells may be highly variable, all
AITL patients who have responded to immunosup- cases contain “hallmark” cells that have the character-
pressive therapy, such as low-dose methotrexate/pred- istic appearance of abundant cytoplasm with large
236 nisone and cyclosporine, purine analogs, and eccentric, horse-shoe or kidney-shaped nuclei, with a
Chapter 14: T-cell non-Hodgkin lymphoma
Table 14.4 Comparison of clinicopathologic features of primary cutaneous ALCL and systemic ALCL.
237
Chapter 14: T-cell non-Hodgkin lymphoma
A B
C D
Figure 14.4 (a) Anaplastic large-cell lymphoma involving lymph node; H&E stain. Clusters of lymphoma cells are visible within the subcapsular
sinuses. Less visible lymphoma cells extend into the parenchyma and into capsule which is thickened. (b) ALCL involving lymph node; ALK-1
stain. The ALK+ lymphoma cells within the sinuses and throughout the parenchyma are highlighted. The staining pattern is both nuclear and
cytoplasmic associated with t(2;5). (c) ALCL involving lymph node; H&E stain. The lymphoma cells are large with an abundant amount of
cytoplasm. (d) ALCL involving lymph node; CD30 stain. The ALCL cells show characteristic bright staining for CD30 with a membranous and Golgi
pattern.
perinuclear area of eosinophilia in the cytoplasm. with sheets of more monotonous large cells, either as
Some cells show striking cytoplasmic nuclear pseu- the sole finding or admixed with polymorphic areas,
doinclusions (doughnut cells). The nuclei usually has also been described.
have clumped or dispersed chromatin with basophilic
nucleoli. While infiltration of the lymph node may be Lymphohistiocytic variant (10%)
diffuse in cases with partial infiltration of the node, the In this variant the lymph node is effaced by a population
cells are frequently seen packing sinusoids mimicking of histiocytes with pale cytoplasm that may show
metastatic tumor (Figure 14.4). erythrophagocytosis. There is usually an associated
Several morphological variants have been proliferation of small lymphocytes. The neoplastic cells
described. A proportion of cases show mixed patterns may be scanty and usually have a perivascular
and, in cases that relapse, the second lesion may be a distribution. They may be smaller than those of the com-
morphological variant that defers from the first. mon type and are best highlighted by staining for CD30.
Large cells with a classical morphology are always cytotoxic granules. Unlike ALCL ALK+, EMA is pos-
present but may be scanty with a predominantly peri- itive in only a proportion of cases.
vascular location.
this approach warrants prospective validation (see autol- were treated on these relapsed/refractory CD30+ two
ogous SCT and allogeneic SCT sections below). SGN-35 phase I trials (one q 3 week and one trial weekly
Several novel therapies have been studied as single dosing); 86% of patients (6/7) had documented CR, all
agents for relapsed/refractory disease or in combina- of which were durable. Preliminary results were
tion with CHOP for newly diagnosed patients. Agents recently presented from the pivotal phase II systemic
that have been combined with CHOP include denileu- ALCL trial. Shustov et al. recently presented interim
kin diftitox (Ontak), alemtuzumab (Campath), beva- results of a phase II, single arm, multicenter study of the
cizumab (Avastin), and bortezomib (Velcade). In a antibody–drug conjugate (ADC) in 30 patients with
phase II multicenter trial, 24 patients with T-cell relapsed or refractory ALCL. Brentuximab vedotin
NHL (12% ALCL) received alemtuzumab with 1.8 mg/kg was administered every 3 weeks as a 30-
CHOP as first line therapy. The ORR was 75% (CR minute outpatient IV infusion for up to 16 cycles of
68%) and all ALCL patients (3/3) achieved a CR that treatment. The objective response rate was 87%, with
appeared to be durable. 57% of patients achieving a CR (n = 17) and 30% of
For relapsed/refractory disease, a targeted therapeu- patients achieving a partial remission (PR; n = 9). The
tic agent with impressive single agent activity is brentux- remaining patients had stable disease (n = 3) or were
imab vedotin (SGN-35). As described before, CD30 is not evaluable for response (n = 1). Similar proportions
uniformly expressed in cutaneous and systemic ALCL. of ALK-1-negative and ALK-1-positive patients
Several “cold” anti-CD30 antibodies were studied for the achieved CR and PR. Reduction in tumor burden was
treatment of relapsed/refractory Hodgkin disease and observed in 97% of patients. In August 2001, brentux-
systemic ALCL. MDX-060 (Medarex) is a fully human imab vedotin was FDAA approved for patients whose
anti-CD30 IgG1k monoclonal antibody that has been disease has progressed after a prior chemotherapy treat-
shown to inhibit growth of CD-30-positive tumor cells ment. Please see the Emerging and novel therapies
in vitro and tumor xenograft models. In a phase I/II trial, section below for additional new therapies being
MDX-060 was well tolerated, but clinical activity was studied in relapsed/refractory ALCL.
modest with an 8% response rate (6/72), with 35%
maintaining stable disease. MDX-1401 is a second gen- Subcutaneous panniculitis-like T-cell
eration antibody with an absent fucose, thereby increas-
ing the antibody’s antibody-dependent cellular lymphoma (SCPTL)
cytotoxicity. In preclinical studies, MDX-1401 had SCPTL is a rare T-cell lymphoma that primarily infil-
improved antibody effector function over the fucose- trates the subcutaneous fat without dermal and
containing parental anti-CD30 antibody (MDX-060). epidermal involvement, causing erythematous to vio-
Early phase MDX-1401 trials are ongoing. laceous nodules and/or plaques. It may be associated
SGN-30 (Seattle Genetics) is a chimeric anti-CD30 with a systemic hemophagocytic syndrome – charac-
monoclonal antibody that was evaluated in a phase II terized by high fever, skin lesions, lung infiltrates,
study of 38 relapsed/refractory Hodgkin disease jaundice, hepatosplenomegaly, liver dysfunction,
patients and 41 systemic ALCL. In the ALCL group, coagulation abnormalities, pancytopenia, and a
two patients achieved CR and five additional patients benign prominent histiocytic proliferation with hemo-
achieved a PR (ORR 17%), with response durations phagocytosis. The hemophagocytic syndrome (HPS)
ranging from 27 to 1460+ days. Modifications to this associated with T-cell neoplasms can occur preceding,
antibody have also been made; SGN-35 is an antibody– during, or while the disease is in remission. Moreover,
drug conjugate, which was formed by coupling the anti- a controversial entity known as cytophagic histiocytic
CD30 antibody, cAC10, to monomethyl auristatin E panniculitis (CHP) has been described as an inflam-
(MMAE), an antitubulin agent. In preclinical mouse matory disease with a possible association to SCPTL.
xenograft models, SGN-35 also induced durable dose- CHP is a disease that has been recognized to have
dependent tumor regression compared to either diverse outcomes, ranging from indolent to aggres-
untreated mice or another control group receiving sive/fatal clinical courses. There is increasing evidence
IgG-vcMMAE. In two phase I studies, SGN-35 that, within the group of SCPTL, there is a distinction
appeared to have significant clinical activity in between cases with an αβ T-cell phenotype and cases
relapsed/refractory systemic ALCL and relapsed/refrac- with a γδ phenotype. In the recent WHO–EORTC
240 tory Hodgkin disease. Seven systemic ALCL patients classification, only the cases with an αβ phenotype
Chapter 14: T-cell non-Hodgkin lymphoma
are classified as SCPTL. Cases previously classified as Cutaneous gamma delta (γδ) T-cell
SCPTL with a γδ phenotype, which comprised 25% of
all cases, are now classified as cutaneous γδ T-cell lymphoma (CGD-TCL)
lymphomas. SCPTL has a more indolent course and Cutaneous γδ T-cell lymphomas (CGD-TCL) include
is less likely to be associated with the HPS, as com- lymphomas previously classified as SCPTL with a γδ
pared to cutaneous γδ T-cell lymphomas. phenotype. CGD-TCL have a more aggressive course
To better define the features and prognosis the αβ as compared with SCPTL. Patients present with skin
phenotype and γδ phenotype, 63 patients with the αβ lesions similar to SCPTL; however, they may also have
phenotype and 20 patients with the γδ phenotype were epidermal and dermal involvement. Patients with sub-
analyzed at a workshop of the EORTC Cutaneous cutaneous disease do worse as compared to those with
Lymphoma Group. SCPTL αβ had a CD4–, CD8+, only epidermotropic or dermal disease. Patients with
CD56–, βF1+ phenotype and had a favorable prognosis, CGD-TCL may also have mucosal involvement. There
with 5-year OS 82%. SCPTL αβ was associated with HPS is still debate as to whether the cutaneous and mucosal
in only 17%. SCPTL αβ patients without HPS had a γδ lymphomas are two different diseases or different
significantly better survival than patients with HPS, with presentations of the same disease. HPS can occur with
5-year OS of 91% versus 46% (P < 0.001). SCPTL γδ had CGD-TCL but generally patients do not have evidence
a CD4–, CD8–, CD56+/–, βF1– phenotype and carried of systemic disease on presentation.
a poor prognosis, with 5-year OS of 11%. This poor
outcome was not affected by the presence of HPS or the Pathology
type of treatment. The results of this study indicate that The fourth edition of the WHO Classification of
SCPTL αβ and SCPTL γδ are distinct entities. Tumors of Hematopoietic and Lymphoid Tissues sepa-
rates primary cutaneous T-cell lymphomas that
Pathology express the γδ T-cell receptor from SCPTL. Unlike
The neoplastic T-cell infiltrate involves fat lobules and the latter, primary cutaneous γδ T-cell lymphoma is
shows characteristic rimming of individual fat cells. not confined to the subcutis and can show three major
Septa are typically spared. Unlike cutaneous γδ T-cell overlapping histological patterns: epidermotropic,
lymphoma, SCPTL does not usually involve the der- dermal, and subcutaneous. There may be necrosis
mis and epidermis. The neoplastic cells may vary in and vascular invasion, and the subcutaneous pattern
size from case to case, but are monomorphic in a given shows fat rimming by neoplastic T-cells, but, unlike
case and usually have irregular hyperchromatic nuclei. SCPTL, it usually also shows dermal and/or epidermal
Rimming of individual fat spaces is a characteristic involvement.
feature. Necrosis, apoptosis, and associated fat
necrosis are common features. The neoplastic T-cells Treatment
which are of αβ type are usually CD8-positive and Patients are usually treated with aggressive regimens.
express cytotoxic granules. Toro et al. treated patients with local therapies, such as
topical steroids, psoralen and PUVA radiation, as well
as systemic therapies like IFN-α, IFN-γ, CHOP, radi-
Treatment
ation therapy, and bone marrow transplant. Patients
The clinical course of SCPTL is variable, ranging from with dermal and epidermotropic involvement had a
indolent disease to rapidly fatal fulminant hemophago- median survival of 29 months. Patients with subcuta-
cytosis. When treatment is warranted, treatment varies neous disease had a less favorable median survival of
from only surgery or radiotherapy to doxorubicin- 13 months. In those with aggressive disease, there may
based chemotherapy or high-dose chemotherapy be a potential role for early allogeneic SCT.
followed by autologous SCT. Because of the data sup-
porting the excellent prognosis of SCPTL αβ without
associated HPS, some authors question the use of Hepatosplenic T-cell lymphoma (HSTCL)
aggressive multiagent chemotherapy. A small case series HSTCL is a rare form of PTCL that represents only 1.4%
has been reported where patients have achieved durable of cases of peripheral T-cell and NK-cell lymphomas
responses with the combination of cortosteroids and worldwide. Regionally, it represents a slightly higher
methotrexate. There have also been anecdotal reports percentage of all PTCLs in North America (3.0%) and 241
of denileukin diftitox having activity in this disease. Europe (2.3%) but is rare in Asia (0.2%). Median age at
Chapter 14: T-cell non-Hodgkin lymphoma
time of diagnosis is 34 years, and 68% of cases are in seen in concert with trisomy 8 in a subset of cases.
male patients. Patients present with B-symptoms, Interestingly, rare TCR αβ variants of the disease have
prominent hepatosplenomegaly, anemia, neutropenia, been described, with demonstration of identical chro-
thrombocytopenia (commonly severe), peripheral mosome aberrations.
blood lymphocytosis, and lymphadenopathy. It is
often associated with an aggressive clinical course Treatment
(median survival 16 months). The clinical course of HSTCL is commonly aggressive
despite multiagent chemotherapy, and median survival
Pathology is approximately 1 year. Case reports have described
The enlarged spleen shows involvement of the cords significant clinical activity with the purine analog, pen-
and sinuses of the red pulp with atrophy of the white tostatin, in relapsed patients. Alemtuzumab, both as a
pulp. The liver and bone marrow show a predomi- single agent as well as in combination with cladribine
nantly intrasinusoidal pattern of infiltration. In the and fludarabine, has also been reported anecdotally to
marrow the intrasinusoidal infiltrate may not be have responses. A phase II study by Revandi et al.
appreciated without immunohistochemistry. The studied 24 patients with a variety of T-cell leukemias
neoplastic cells are of medium size and are usually and lymphomas (three of whom had HSTCL) with a
of γδ T-cell receptor type, but may less commonly combination of alemtuzumab 30 mg intravenously
be of αβ type. They are CD3+, but are negative for three times weekly for up to 3 months and pentostatin
CD5. They may express CD56 and CD8, but are 4 mg/m2 IV weekly for 4 weeks followed by alternate
negative for CD4. The pattern of expression of cyto- weekly administration for up to 6 months. Prophylactic
toxic granules shows a non-activated cytotoxic T-cell antibiotics, including antiviral, antifungal, and antibac-
phenotype as TIA-1 and granzyme M are expressed, terial agents, were administered during the treatment
but perforin and granzyme B are usually negative and for 2 months after its completion. Two of the three
(Figure 14.5). patients with HSTCL achieved a CR with this regimen,
and one of these patients was subsequently able to
Molecular pathology and cytogenetics receive an allogeneic SCT. The main toxicity was
Hepatosplenic γδ T-cell lymphoma is associated with a grade 3/4 infections, most frequently cytomegalovirus
recurring chromosomal abnormality, i(7)(q10), that is (CMV) reactivation.
242 C D
Chapter 14: T-cell non-Hodgkin lymphoma
retrospectively compared stage IE patients who received Etoposide has shown in vitro and in vivo efficacy for
radiation alone versus combined chemotherapy with NK-cell neoplasms, pediatric EBV-related HPS, and
radiotherapy. Of note, this group divided IE patients pediatric EBV-associated lymphoproliferative dis-
into limited (confined to nasal cavity) and extensive ease. L-Asparaginase induces the selective apoptosis
(presenting with extension beyond the nasal cavity) of NK lymphoma cells in vitro. Successful results
stage IE disease. Limited-stage IE patients survived lon- in NK-cell lymphoma have been reported for
ger than extensive IE patients overall (5-year OS 90% L-asparaginase, either alone or in combination with
versus 57%, respectively; P < 0.001). Moreover, compar- other chemotherapy. In a 2012 Blood report, this
ing radiation alone to combined modality therapy, the 5- group published a follow-up phase II study confirm-
year OS was not significantly different for limited-stage ing the efficacy & SMILE in advanced-stage NK-/T-
IE, at 89% and 92%, respectively, as well as extensive IE cell lymphoma patients. Toxicities were significant,
disease at 54–58%, respectively. especially hematologic and real, while the estimated
Because of the risk of systemic failure in localized 5-year OS was 50% and 4-year DFS was 64%.
disease, Kim et al. reported a phase II trial of 30 Lee et al. reported on 26 patients with extranodal
patients with newly diagnosed stage IE to IIE, nasal NK-/T-cell lymphoma who received ifosfamide,
extranodal NK-/T-cell lymphoma who received con- methotrexate, etoposide, and prednisolone (IMEP)
current chemoradiotherapy with radiation 40–52.8 Gy chemotherapy as first line treatment [ifosfamide 1.5 g/
and cisplatin 30 mg/m2 weekly. This was followed by m2 (days 1–3), methotrexate 30 mg/m2 (days 3 and 10),
three cycles of VIPD (etoposide 100 mg/m2 day 1–3, etoposide 100 mg/m2 (days 1–3), and prednisolone
ifosfamide 1200 mg/m2 days 1–3, cisplatin 33 mg/m2 120 mg (days 1–5)]. Radiotherapy was administered
days 1–3, and dexamethasone 40 mg days 1–4). All of only to patients with Ann Arbor stage I/II that had
the patients completed chemoradiation and had a not achieved CR or for those who developed local fail-
100% response rate, which included 22 CRs and 8 ure after completing chemotherapy. Sixteen patients
PRs. Twenty-six of 30 completed the subsequent (group A) had nasal or upper aerodigestive tract local-
three cycles of VIPD, and ORR and CR rate were ization (stage I/II) and 10 (group B) had extranasal or
83% and 80%, respectively. The estimated 3-year PFS disseminated disease. Of the 14 evaluable patients in
and OS rates were 85% and 86%, respectively. The group A, 11 (79%) achieved CR after IMEP alone and
results of this study support that nasal extranodal 13 (93%) after chemotherapy plus additional radiother-
NK-/T-cell lymphomas have favorable outcomes apy. Of the 11 patients who achieved CR with chemo-
with frontline chemoradiation. therapy alone, seven developed recurrence. All of the
Patients with systemic disease have poor long-term recurrences were local failure and were successfully
survival (5-year OS 20–25%) with high locoregional treated by additional curative radiotherapy. However,
(over 50%) and systemic failure rates (over 70%). The patients in group B responded poorly (CR 13%).
traditional approach for stage III and IV extranodal
NK-/T-cell lymphoma, nasal is combined modality Enteropathy-type intestinal T-cell
therapy with adriamycin-based chemotherapy and radi-
ation therapy. L-asparaginase has promising outcomes lymphoma (EITCL)
in advanced, relapsed, and refractory disease, and has EITCL (also known as intestinal T-cell lymphoma) is a
been combined with chemotherapy. A retrospective rare T-cell lymphoma of intraepithelial lymphocytes
study of an L-asparaginase-based regimen demonstra- that commonly presents with multiple circumferential
ted an ORR of 82% (CR 56%; 5-year OS 67%). In jejunal ulcers in adults. Patients usually have a prior
patients with advanced disease, there is evidence that brief history of gluten-sensitive enteropathy. EITCL
salvage treatment significantly prolonged the 5-year OS accounts for less than 1% of NHLs and has been recog-
compared with best supportive care alone (38% versus nized to have a poor prognosis, with reported 5-year OS
0%, respectively; P < 0.001). and DFS rates of 20% and 3%, respectively. This is in
Yamaguchi et al. reported a phase I study of a part related to many patients presenting with malnu-
new chemotherapeutic regimen, SMILE, which con- trition and poor PS.
sists of dexamethasone, methotrexate, ifosfamide, EITCL may present without antecedent celiac dis-
L-asparaginase, and etoposide. The rationale for this ease history, but most patients have abdominal pain and
244 regimen is based on both in vivo and in vitro data. weight loss. Evidence of celiac disease serologic markers
Chapter 14: T-cell non-Hodgkin lymphoma
Pathology
The ulcerating tumor contains medium to large atyp-
Autologous stem cell transplantation
ical lymphoid cells that are associated with an inflam- Because of the poor prognosis with conventional ther-
matory response that often includes many histiocytes apy in T-cell lymphoma, there has been interest in the
and eosinophils. If pronounced, the inflammation role of more aggressive treatment approaches, such as
may obscure the neoplastic cells. The adjacent high-dose therapy with autologous stem cell trans-
mucosa shows features of enteropathy with an intra- plantation (autoSCT), which has been considered
epithelial lymphocytosis. In the less common type II standard therapy for many patients with relapsed
enteropathy-type T-cell lymphoma (EATCL), the aggressive B-cell lymphomas, such as DLBCL.
neoplastic T-cells are monomorphic and of medium
size. The tumor is less necrotic than in classical Frontline consolidative therapy
EATCL. The adjacent mucosa shows a prominent There have been many retrospective studies with PTCL
intraepithelial lymphocytosis. In classic EATCL, the patients that have examined upfront (consolidative)
neoplastic T-cells express CD3, CD7, and CD103, high-dose therapy with autoSCT (Table 14.6). OS
show loss of CD5 and often loss of CD8. They are rates in these small studies ranged from 73% at 3 years
CD4-negative. CD30 expression is variable. The to 60% at 5 years. Rodriguez et al. reported on 74
intraepithelial lymphocytosis adjacent to the tumor patients with a 5-year OS rate of 68% after 67 months’
may show loss of CD8 and the presence of an identi- follow-up. Feyler et al. reported on 64 patients, with a
cal T-cell clone to that of the overt lymphoma. In 3-year OS of 53% after 37 months’ follow-up. However,
EATCL type II, the monomorphic neoplastic cells are in both of these reports ALK+ ALCL patients were
CD8-positive and express CD56, which is a similar included, likely biasing the results because of their
immunoprofile to that of the adjacent intraepithelial known superior outcomes compared to other subtypes
lymphocytes. of T-cell lymphoma. The largest retrospective study was
by the European Bone Marrow Transplantation Group,
Molecular pathology and cytogenetics who examined 146 patients with AITL. The OS rate was
EATCL has been shown to harbor distinct genetic 67% at 24 months and 59% at 48 months. Nickelson
alterations as detected by CGH and fluorescent et al. attempted to use an intensified regimen of
in situ hybridization (FISH). Recurring genomic losses MegaCHOEP (cyclophosphamide, adriamycin, vincris-
were observed in 8p, 9p, and 13q, and gains of chro- tine, etoposide, and prednisone) followed by autoSCT
mosomal material were found in 1q, 5q, and 7q. Of in patients <60 years old as that strategy had been
major importance, EATCL harbors a characteristic successfully used in younger patients with aggressive
chromosomal alteration, a gain in 9q with a minimally B-cell lymphoma. Compared with 84% of B-cell NHL
overlapping region at 9q33–34 in approximately 60% patients, only 22/33 (67%) patients were able to receive
of cases. Interestingly, this chromosomal gain was also all treatments, including autoSCT. Sixteen of 33
observed in other types of extranodal aggressive PTCL, patients (49%) achieved a CR or CRu, two patients
namely in four primary gastric and one colonic PTCL, (6%) achieved a PR, three patients (9%) had stable
leading to speculation about a possible association of disease (SD), nine patients (27%) progressed within 2
9q gains with gastrointestinal PTCL as a whole. In months after final restaging or during therapy, one
addition, 9q gains are an infrequent finding in nodal patient had an unknown response at the end of treat-
PTCL-NOS, suggesting profound differences in the ment, and two therapy-related deaths were observed.
overall genetic profile of EATCL as compared to pri- The 3-year EFS and OS were only 25.9% and 44.5%, 245
mary nodal T-cell lymphomas. respectively. The group compared outcomes to a
Table 14.6 Studies of autologous stem cell transplantation for peripheral T-cell lymphoma.
Study Study No. Regimen Disease status at time of Response Overall Comments
transplant survival
Consolidative (first remission)
Gisselbrecht et al., 2002 Prospective 76 BEAM NA No data 32% at 5 years Included ALCL and precur-
(n = 43), (for BEAM) sor T-cell lymphoma
diverse
Corradini et al., 2006 Prospective 62 Mito/Mel CR 56%, PR 16%, PD 24% CR 66%, PR 34% at 12 years Included ALK+ ALCL; 74%
or BEAM 18% received autoSCT
D’Amore et al., 2006 Prospective 77 BEAM CR 50%, PR 35% 71% CR/PR 30 of 39 No ALK+ patients; 75%
patients in CR 1 received autoSCT
year post tx
Rodriguez et al., 2007 Retrospective 19 BEAM/ CR1 42%, CR2 15%, PR1 26%, PR2 79% CR, 5% 60% at 5 years Only AITL
BEAC/CVB/ 5%, PD 10% PR
CY + TBI
Rodriguez et al., 2007 Retrospective 74 BEAM/ CR1 100% NA 68% at 5 years Included ALCL
BEAC/CVB/
CY + TBI
Feyler et al., 2007 Retrospective 64 Diverse CR1 48%, CR2 6%, PR 23%, PD 22% NA 53% at 3 years Included ALK+ ALCL
Kyriakou et al., 2008 Retrospective 146 Diverse CR1 34%, CR2 14%, 70% CR, 7% 59% at 4 years Only AITL
chemotherapy-sensitive 38%, PR
chemotherapy-refractory 14%
Reimer et al., 2008 Prospective 83 Cy/TBI CR 39%, PR 40%, 21% PD 58% CR, 8% 48% at 3 years No ALK+ patients, 66%
PR received autoSCT
Relapsed/refractory disease
Kim et al., 2007 Retrospective 29 Diverse CR2 27%, PR 58%, refractory 14% NA 7.4 months CR prior to autoSCT prog-
median nostic: EFS (P ≤ 0.025) and
OS (P ≤ 0.027)
Rodriguez et al., 2007 Retrospective 123 Diverse CR2 36%, PR 16%, RD 9%, non- CR 73%, PR Pre-autoSCT No difference in OS
treated relapse 3% 11%, SD or status: between CR1 or CR2.
PD 16% CR: 5-year Poorest outcomes for PR2
OS 57%, or RD
PR: 5-year
OS 33%
Smith et al., 2007 Retrospective 25 Bu/Cy/ CR2/PR2 68%, RD 32% No data At 5-years: 34%, No difference in OS for CR1
VP16 RFS 18% or CR2/PR/PD
Yang et al., 2009 Retrospective 30 Diverse CR2 17%, PR2 83% No data At 5-years: CR2 CR prior to autoSCT prog-
40%, PR2 12% nostic for OS
NA, not available; autoSCT, autologous stem cell transplantation; OS, overall survival; CR, complete response; AITL, angioimmunoblastic T-cell lymphoma; BEAM, carmustine, etoposide,
cytarabine, melphalan; BEAC, carmustine, etoposide, cytarabine, cyclophosphamide; CVB, cyclophosphamide, etoposide, methotrexate; ALCL, anaplastic large-cell lymphoma; ALK, anaplastic
lymphoma kinase; Mito, mitoxantrone; Mel, melphalan; Bu, busulfan; Cy, cyclophosphamide; VP16, etoposide; TBI, total-body irradiation; PR, partial response; CR1, first complete response;
CR2, second complete response; PD, progressive disease; PR1, first partial response; PR2, second partial response; RD, refractory disease.
Chapter 14: T-cell non-Hodgkin lymphoma
Yang et al. retrospectively evaluated 64 patients who AITL (n = 11), hepatosplenic γδ lymphoma (n = 3),
underwent autoSCT, 30 of whom were in the relapsed T-cell granular lymphocytic leukemia (n = 1), enter-
or refractory setting. The 5-year PFS and OS rates for opathy T-cell lymphoma (n = 1), and ATLL (n = 2). All
patients in CR1/PR1 groups were 51% and 76%, respec- patients had received at least one line of therapy (mean
tively, decreasing to 12% and 40%, respectively, for 2, range 1–5) before alloSCT, with 25% receiving a
patients in CR2/PR2. The 3-year OS rate for patients previous autoSCT. Fifty-seven patients (74%) received
who achieved CR (n = 21), regardless of transplant a myeloablative conditioning regimen. Only 31 of 77
timing, was 71.8%. They found, however, that 3-year patients were in CR at the time of alloSCT, whereas 26
OS rates differed significantly in patients undergoing of 77 were in PR. With mean follow-up of 43 months,
transplantation in CR1/PR1 (60%) and those under- the 5-year OS and EFS rates were 57% and 53%, respec-
going transplantation in a salvage setting (37.7%). tively. In multivariate analysis, chemoresistant disease
However, the 3-year OS rate for patients in CR2 was at the time of alloSCT and the occurrence of severe
70.9%, compared with 50% for those in PR1. grade 3–4 acute graft-versus-host disease were the
Consequently, the achievement of CR at the time of strongest adverse prognostic factors for OS (P = 0.03).
transplantation was a more significant factor for pre- Kyriakou et al. retrospectively analyzed 45 patients
dicting survival than was transplant timing. Although with AITL who underwent alloSCT. The majority of
this data is retrospective, it may indicate that there may patients had received over two lines of previous che-
be a role for autoSCT even in the relapse/refractory motherapy, and 11 of the 45 patients had treatment
setting, especially if a CR2 can be achieved. failure to a prior autoSCT. Twenty-five patients under-
went a myeloablative alloSCT and 20 underwent
reduced-intensity alloSCT. PFS and OS rates were
Allogeneic stem cell transplantation 62% and 53% and 66% and 64% at 1 and 3 years,
The relapse rate among relapsed/refractory T-cell lym- respectively. Those with chemotherapy-sensitive dis-
phoma patients is high following standard chemother- ease had significantly improved outcomes. In this
apy as well as autoSCT, therefore there has been interest small series, the results of the reduced-intensity
in evaluating the use of allogeneic SCT (alloSCT) alloSCT were similar to those who received a myelo-
(Table 14.7). The Societe Francaise de Greffe de ablative alloSCT.
Moelle et de Therapie Cellulaire conducted a retrospec- Shustov et al. retrospectively analyzed 17 patients
tive analysis of 77 patients with T-cell lymphoma; his- with relapsed/refractory (n = 14) or poor-risk newly
tologies included ALCL (n = 27), PTCL-NOS (n = 27), diagnosed (n = 3) T-cell and NK-cell lymphomas who
Table 14.7 Studies of allogeneic stem cell transplantation for T-cell lymphoma (non-cutaneous).
underwent non-myeloablative alloSCT. These often obviating the need for the addition of
patients were highly pre-treated with median number radiation. Patients with extranodal NK-/T-cell lym-
of prior treatment three (range 1–7), and seven phoma can frequently present with localized (stage I
patients (41%) received prior autoSCT. Eight (47%) or II) disease that responds well to treatment with
patients were in CR and four (24%) patients were in radiation with or without chemotherapy. Studies spe-
PR at the time of alloSCT. Five (29%) patients had cifically addressing the benefits of radiation therapy for
refractory disease prior to alloSCT. With a median stage I/II non-extranodal NK-/T-cell NHL or bulky sites
follow-up of 3.3 years, 9 of 17 patients were still alive. with advanced disease have not been done. Despite
Estimated 3-year OS and PFS were 59% and 53%, lacking data, common practice has been to apply radio-
respectively. Interestingly, five out of nine patients therapy similar to aggressive B-cell lymphoma.
who had evidence of disease prior to stem cell infu-
sion achieved CR after transplantation, suggesting
the role for graft-versus-lymphoma effect. Similar
conversions from PR to CR after non-myeloablative
Emerging and novel therapies
conditioning and alloSCT were observed in other Identification of relevant proto-oncogenes and tumor
studies in patients with PTCL. suppressor genes involved in the pathogenesis of T-cell
Smith et al. recently published data comparing NHL represents potential candidates for molecular-
post-transplant outcomes among 241 recipients (≤60 based therapy. There are many drugs being examined
years) of autoSCT (n = 115) and alloSCT (n = 126, 76 in patients with T-cell NHL, some of which have been
matched siblings) for T-cell lymphoma between 1996 proven effective in other malignancies (Table 14.8).
and 2006, reported to the CIBMTR (Center for Proteasome inhibitors, including bortezomib, rely
International Blood and Marrow Transplant on inhibition of the ubiquitin-proteasome pathway,
Research). The median age was 43 years for autoSCT and reports of phase II trials with single agent borte-
and 38 years for alloSCT. In the alloSCT cohort, 59% zomib as well as in combination with CHOP show
received myeloablative conditioning and 60% had promising ORRs of approximately 60%. Histone
matched sibling donors. AutoSCT, compared to deacetylase (HDAC) inhibitors trigger both caspase-
alloSCT, was used more frequently in ALCL (53% dependent and caspase-independent apoptosis; romi-
versus 40%; P = 0.04) and in those with less advanced depsin, a pan-HDAC inhibitor, has shown activity in
and more sensitive disease, including those in first CR both CTCL and PTCL-NOS with ORR 41%. These
(35% versus 14%; P = 0.001), chemosensitive disease responses were durable, with a median of 14.9 (range
(86% versus 60%; P < 0.0001), and ≤2 lines prior 1–66) months, and many of these patients were heavily
therapy (65% versus 44%; P < 0.001). Median follow- treated prior to receiving romidepsin. Romidepsin
up was 71 months for autoSCT and 49 months for received accelerated FDA approval in June 2011 for
alloSCT. Treatment-related mortality at 100 days in the treatment of PTCL patients who have received at
the autoSCT and alloSCT cohorts was 2% and 17%, least one prior therapy.
respectively. For those beyond CR1 in the autoSCT Denileukin diftitox (DAB389IL-2, Ontak) is a novel
and alloSCT cohorts, OS at 1 year was 62% versus 52% recombinant fusion protein consisting of peptide
and at 3 years 53% versus 41%, respectively. Relapsed sequences for the enzymatically active and membrane
T-cell NHL was the cause of death in 73% of autoSCT active and membrane translocation domains of diph-
and 44% of alloSCT patients. theria toxin with recombinant interleukin-2. It has
been FDA-approved for cutaneous T-cell NHL,
although clinical benefit has been reported in other
Radiation therapy T-cell NHL patients. As a single agent, denileukin
Because of the paucity of clinical trials, the integration diftitox has a 48% ORR in relapsed/refractory PTCL.
of radiation therapy into the treatment plan of most Recently presented phase II data combining denileu-
T-cell NHL patients often models that of B-cell NHL. kin diftitox and CHOP in relapsed/refractory PTCL
Compared to DLBCL, T-cell NHL more commonly showed a promising 67% ORR and 58% CR rate, and
presents with stage III/IV disease (48% and 74%, respec- there is a larger multicenter trial planned.
tively) and T-cell patients frequently present with com- Alemtuzumab is a monoclonal anti-CD52 anti-
bined nodal and extranodal disease at diagnosis body that has shown activity in PTCL. An Italian 249
(especially PTCL-NOS, ALCL, and AILT), thereby group combined CHOP with alemtuzumab in 24
Chapter 14: T-cell non-Hodgkin lymphoma
patients, 14 of whom were classified as unspecified non-randomized, open-label study, PROPEL, treated
(PTCL-u). Of these patients, nine had either a CR or 117 heavily pre-treated patients with relapsed or
PR, with five having disease progression. At a mean refractory PTCL (n = 109 evaluable for efficacy). The
follow-up of 495 days, 5/14 patients with PTCL-u were ORR was 27%, with eight patients having CR and
alive and all were in CR. Toxicity due to infections can seven with PR. Of note, only one (5%) of 20 patients
be prohibitive with alemtuzumab and more trials of with B-cell lymphoma responded, while 54% of
this combination are ongoing. In addition, given that patients with T-cell lymphoma obtained
about 30–40% of PTCL-u are CD52+, determining remission. Among the T-cell lymphoma patients who
antigen status is important. Huang and colleagues responded, eight (31%) had CR, and six (23%) had PR.
studied 17 patients with relapsed T-cell NHL using Monitoring for survival is ongoing, but responses lon-
13-cis-retinoic acid with IFN-α. A response rate of ger than 1 year have been observed. Of note, in this
31% was documented in four of six ALCL patients study 70% of responses were achieved after a single
and one of seven PTCL-u, although median survival dose, and five patients were able to undergo hemato-
for the entire group of patients was 3.6 months. poietic stem cell transplant (HSCT). Pralatrexate
Pralatrexate (10-propargyl-10-deazaaminopterin) was overall well tolerated, with 77 patients (69%)
is an analog of methotrexate and, when compared to receiving full-dose therapy, and 85% of all scheduled
methotrexate, has improved cellular transport via doses were administered. The most frequent grade
RFC-1, leading to a greater intracellular concentration 3 and 4 adverse effects were mucositis (17% and
250 of drug and more effective inhibition of dihydrofolate 4%, respectively) and thrombocytopenia (14% and
reductase (DHFR). An international, phase II, 19%, respectively). Nausea and fatigue were also
Chapter 14: T-cell non-Hodgkin lymphoma
Evens AM, Rosen ST, Helenowski I et al. A phase I/II trial of Jaffe ES, Chan JK, Su IJ, et al. Report of the workshop on
Bortezomib combined concurrently with Gemcitabine nasal and related extranodal angiocentric T/natural
for relapsed or refractory DLBCL and peripheral T-cell killer cell lymphomas. Definitions, differential
lymphomas. Br J Haematol, 2013;163:55–56. diagnosis, and epidemiology. Am J Surg Pathol,
Foroni L, Foldi J, Matutes E, et al. Alpha, beta and gamma 1996;20:103–111.
T-cell receptor genes: rearrangements correlate with Jaffe ES, Gaulard P, Ralfkiaer, Cerroni L, Meijer CJLM.
haematological phenotype in T cell leukaemias. Br J Subcutaneous panniculitis-like T-cell lymphoma. In:
Haematol, 1987;67:307–318. Swerdlow SH, Campo E, Harris NL, et al. (eds). WHO
Gale J, Simmonds PD, Mead GM, Sweetenham JW, Wright Classification of Tumours of Haematopoietic and
DH. Enteropathy-type intestinal T-cell lymphoma: Lymphoid Tissues, 4th ed. Lyon: IARC; 2008: 294.
clinical features and treatment of 31 patients in a single Kanavaros P, Lescs MC, Briere J, et al. Nasal T-cell
center. J Clin Oncol, 2000;18:795–803. lymphoma: a clinicopathologic entity associated with
Gallamini A, Stelitano C, Calvi R, et al. Peripheral T-cell peculiar phenotype and with Epstein–Barr virus. Blood,
lymphoma unspecified (PTCL-U): a new prognostic 1993;81:2688–2695.
model from a retrospective multicentric clinical study. Kim SJ, Kim K, Kim BS, et al. Phase II trial of concurrent
Blood, 2004;103:2474–2479. radiation and weekly cisplatin followed by VIPD
Gallamini A, Zaja F, Patti C, et al. Alemtuzumab (Campath- chemotherapy in newly diagnosed, stage IE to IIE, nasal,
1H) and CHOP chemotherapy as first-line treatment of extranodal NK/T-cell lymphoma: Consortium for
peripheral T-cell lymphoma: results of a GITIL (Gruppo Improving Survival of Lymphoma study. J Clin Oncol,
Italiano Terapie Innovative nei Linfomi) prospective 2009;27:6027–6032.
multicenter trial. Blood, 2007;110:2316–2323. Knudson AG. Antioncogenes and human cancer. Proc Natl
Gascoyne RD, Aoun P, Wu D, et al. Prognostic significance Acad Sci U S A, 1993;90:10914–10921.
of anaplastic lymphoma kinase (ALK) protein expression Kyriakou C, Canals C, Finke J, et al. Allogeneic stem cell
in adults with anaplastic large cell lymphoma. Blood, transplantation is able to induce long-term remissions in
1999;93:3913–3921. angioimmunoblastic T-cell lymphoma: a retrospective
Gisselbrecht C, Gaulard P, Lepage E, et al. Prognostic study from the Lymphoma Working Party of the
significance of T-cell phenotype in aggressive non- European Group for Blood and Marrow Transplantation.
Hodgkin’s lymphomas. Groupe d’Etudes des J Clin Oncol, 2009;27:3951–3958.
Lymphomes de l’Adulte (GELA). Blood, 1998;92:76–82. Marafioti T, Paterson JC, Ballabio E, et al. The inducible T-
Glass B, Kloess M, Bentz M, et al. Dose-escalated CHOP plus cell co-stimulator molecule is expressed on subsets of T
etoposide (MegaCHOEP) followed by repeated stem cell cells and is a new marker of lymphomas of T follicular
transplantation for primary treatment of aggressive high- helper cell-derivation. Haematologica, 2010;95
risk non-Hodgkin lymphoma. Blood, 2006;107:3058–3064. (3):432–439.
Grogg KL, Attygalle AD, Macon WR, et al. Expression of Melnyk A, Rodriguez A, Pugh WC, Cabannillas F.
CXCL13, a chemokine highly upregulated in germinal Evaluation of the Revised European–American
center T-helper cells, distinguishes angioimmunoblastic Lymphoma classification confirms the clinical
T-cell lymphoma from peripheral T-cell lymphoma, relevance of immunophenotype in 560 cases of
unspecified. Mod Pathol, 2006;19(8):1101–1107. aggressive non-Hodgkin’s lymphoma. Blood,
1997;89:4514–4520.
Gutierrez-Garcia G, Garcia-Herrera A, Cardesa T, et al.
Comparison of four prognostic scores in peripheral T-cell Mourad N, Mounier N, Briere J, et al. Clinical, biologic, and
lymphoma. Ann Oncol, pathologic features in 157 patients with
angioimmunoblastic T-cell lymphoma treated within the
Hasserjian RP, Harris NL. NK-cell lymphomas and Groupe d’Etude des Lymphomes de l’Adulte (GELA)
leukemias: a spectrum of tumors with variable trials. Blood, 2008;111:4463–4470.
manifestations and immunophenotype. Am J Clin Pathol,
2007;127(6):860–868. O’Connor OA, Horwitz S, Hamlin P, et al. Phase II-I-II study
of two different doses and schedules of pralatrexate, a
Huang Y, de Reynies A, de Leval L, et al. Gene expression high-affinity substrate for the reduced folate carrier, in
profiling identifies emerging oncogenic pathways patients with relapsed or refractory lymphoma reveals
operating in extranodal NK/T-cell lymphoma, nasal type. marked activity in T-cell malignancies. J Clin Oncol,
Blood, 115:1226–1237. 2009;27:4357–4364.
Iqbal J, Weisenburger DD, Greiner TC, et al. Molecular Pellatt J, Sweetenham J, Pickering RM, Brown L, Wilkins B.
signatures to improve diagnosis in peripheral T-cell A single-centre study of treatment outcomes and survival
lymphoma and prognostication in angioimmunoblastic in 120 patients with peripheral T-cell non-Hodgkin’s
252 T-cell lymphoma. Blood, 115:1026–1036. lymphoma. Ann Hematol, 2002;81:267–272.
Chapter 14: T-cell non-Hodgkin lymphoma
Pileri SA, Ralfkiaer E, Weisenberger DD, et al. Peripheral Sonnen R, Schmidt WP, Muller-Hermelink HK, Schmitz N.
T-cell lymphoma, not otherwise specified. In: Swerdlow The International Prognostic Index determines
SH, Campo E, Harris NL, et al. (eds). WHO Classification the outcome of patients with nodal mature T-cell
of Tumours of Haematopoietic and Lymphoid Tissues, 4th lymphomas. Br J Haematol, 2005;129:366–372.
ed. Lyon: IARC; 2008: 306–308. Stein H, Foss HD, Dürkop H, et al. CD30(+) anaplastic
Rabbitts TH. Chromosomal translocation master genes, large cell lymphoma: a review of its histopathologic,
mouse models and experimental therapeutics. Oncogene, genetic, and clinical features. Blood,
2001;20:5763–5777. 2000;96(12):3681–3695.
Rao PH, Houldsworth J, Dyomina K, et al. Chromosomal Swerdlow SH, Harris CE. WHO Classification of Tumours of
and gene amplification in diffuse large B-cell lymphoma. Haematopoietic and Lymphoid Tissues, 4th ed. Lyon:
Blood, 1998;92:234–240. IARC; 2008.
Reimer P, Rudiger T, Geissinger E, et al. Autologous stem- The International Non-Hodgkin’s Lymphoma Prognostic
cell transplantation as first-line therapy in peripheral Factors Project. A predictive model for aggressive
T-cell lymphomas: results of a prospective multicenter non-Hodgkin’s lymphoma. N Engl J Med,
study. J Clin Oncol, 2009;27:106–113. 1993;329:987–994.
Rodriguez J, Conde E, Gutierrez A, et al. Prolonged survival The Non-Hodgkin’s Lymphoma Classification Project. A
of patients with angioimmunoblastic T-cell lymphoma clinical evaluation of the International Lymphoma Study
after high-dose chemotherapy and autologous stem cell Group classification of non-Hodgkin’s lymphoma. Blood,
transplantation: the GELTAMO experience. Eur J 1997;89:3909–3918.
Haematol, 2007;78:290–296. Theodorou I, Bigorgne C, Delfau MH, et al. VJ
Rodriguez J, Conde E, Gutierrez A, et al. The results of rearrangements of the TCR gamma locus in peripheral
consolidation with autologous stem-cell transplantation T-cell lymphomas: analysis by polymerase chain reaction
in patients with peripheral T-cell lymphoma (PTCL) in and denaturing gradient gel electrophoresis. J Pathol,
first complete remission: the Spanish Lymphoma and 1996;178:303–310.
Autologous Transplantation Group experience. Ann Toro JR, Liewehr DJ, Pabby N, et al. Gamma-delta T-cell
Oncol, 2007;18:652–657. phenotype is associated with significantly decreased
Rudiger T, Weisenburger DD, Anderson JR, et al. survival in cutaneous T-cell lymphoma. Blood,
Peripheral T-cell lymphoma (excluding anaplastic large- 2003;101(9):3407–3412.
cell lymphoma): results from the Non-Hodgkin’s Tsukasaki K, Krebs J, Nagai K, et al. Comparative genomic
Lymphoma Classification Project. Ann Oncol, hybridization analysis in adult T-cell leukemia/
2002;13:140–149. lymphoma: correlation with clinical course. Blood,
Schmitz N, Trumper L, Ziepert M, et al. Treatment and 2001;97:3875–3881.
prognosis of mature T-cell and NK-cell lymphoma: an Vose J, Armitage J, Weisenburger D. International
analysis of patients with T-cell lymphoma treated in peripheral T-cell and natural killer/T-cell lymphoma
studies of the German High-Grade Non-Hodgkin study: pathology findings and clinical outcomes. J Clin
Lymphoma Study Group. Blood, 116:3418–3425. Oncol, 2008;26:4124–4130.
Shimoyama M. Diagnostic criteria and classification of Went P, Agostinelli C, Gallamini A, et al. Marker expression
clinical subtypes of adult T-cell leukaemia-lymphoma. in peripheral T-cell lymphoma: a proposed clinical-
A report from the Lymphoma Study Group (1984–87). pathologic prognostic score. J Clin Oncol,
Br J Haematol, 1991;79:428–437. 2006;24:2472–2479.
Siegert W, Agthe A, Griesser H, et al. Treatment of Willemze R, Jansen PM, Cerroni L, et al. Subcutaneous
angioimmunoblastic lymphadenopathy (AILD)-type panniculitis-like T-cell lymphoma: definition,
T-cell lymphoma using prednisone with or without the classification, and prognostic factors: an EORTC
COPBLAM/IMVP-16 regimen. A multicenter study. Cutaneous Lymphoma Group Study of 83 cases. Blood,
Kiel Lymphoma Study Group. Ann Intern Med, 2008;111:838–845.
1992;117:364–370. Yong W, Zheng W, Zhang Y, et al. L-asparaginase-based
Smith SD, Bolwell BJ, Rybicki LA, et al. Autologous regimen in the treatment of refractory midline nasal/
hematopoietic stem cell transplantation in peripheral nasal-type T/NK-cell lymphoma. Int J Hematol,
T-cell lymphoma using a uniform high-dose regimen. 2003;78:163–167.
Bone Marrow Transplant, 2007;40:239–243. Yong W, Zheng W, Zhu J, et al. L-asparaginase in the
Smith SM, Burns LJ, van Besien K, et al. Hematopoietic cell treatment of refractory and relapsed extranodal NK/
transplantation for systemic mature T-cell non-Hodgkin T-cell lymphoma, nasal type. Ann Hematol,
lymphoma. J Clin Oncol, 2013;31:3100–3109. 2009;88:647–652. 253
Chapter
15
Primary cutaneous lymphoma
Sean Whittaker
Pathology: Andrew Wotherspoon Molecular cytogenetics: Andreas Rosenwald and German Ott
The skin is the second most frequent extranodal site, plaques to extensive thick plaques or tumors and
after the gastrointestinal tract, for lymphoma, with an even erythroderma, and a minority (approximately
annual incidence of 0.5–1 per 100 000, although recent 30%) will present with advanced disease. However,
Scandinavian studies have suggested an incidence of 4 many patients do not develop disease progression. A
per 100 000, possibly because of improved diagnosis number of clinical variants are recognized, including
and registration. hypopigmented and folliculotropic variants.
Primary cutaneous T-cell lymphoma (CTCL) com- Staging is based on cutaneous surface area involved
prises a heterogeneous group of non-Hodgkin’s lym- and type of skin lesion; patches and plaques involving
phoma (NHL), of which mycosis fungoides (MF) is the less than 10% of the body surface area (stage T1/IA;
most common clinicopathologic subtype. MF typically Figure 15.1) and more than 10% (stage T2/IB;
has an indolent course but disease progression may Figure 15.2), tumors (stage T3/IIB; Figure 15.3) and
occur in approximately 35% of patients. Sézary syn- erythrodermic disease (stage T3/III; Figure 15.2). In
drome (SS), a leukemic form of CTCL, is very closely 2007 a revised staging system was proposed, incorp-
related to MF and has a poor prognosis, with a median orating stratification of early-stage skin disease into
survival of less than 3 years. patches only compared to those with patches and pla-
Primary cutaneous B-cell lymphomas (CBCL) are ques as well as a detailed histologic and molecular
less common, comprising approximately 20% of all classification of node and peripheral blood involvement
primary cutaneous lymphomas. They typically present (Table 15.2). In 2010 this proposed staging model was
with cutaneous papules, plaques, or nodules, and can validated on a large cohort of 1502 patients. The prog-
be broadly divided into follicle center cell lymphoma, nosis of those with stage IA and IB disease is excellent,
marginal zone lymphoma, and large B-cell lymphoma. with 5-year disease-specific survival rates of 98% and
The WHO–EORTC classification system 89%, respectively. However, those patients with patches
(Table 15.1) has clarified the classification of primary only (T1a/T2a) have a significantly better prognosis than
cutaneous lymphomas, and the distinction of rare those with patches and plaques (T1b/T2b), with disease-
CTCL variants from MF/SS is critical as the prognosis specific survival rates of 100%/96% compared to 90%/
is poorer and treatment options are different. 86%, respectively. In contrast, the presence of tumors
(IIB) or erythroderma (III) is associated with a worse
prognosis (5-year disease-specific survival rates of 52%/
Primary cutaneous T-cell 48%, respectively) (Table 15.3). Notably, the presence of
lymphomas (CTCL) a peripheral blood T-cell clone identical to the cutane-
ous T-cell clone (B0b) is associated with a significantly
Mycosis fungoides worse prognosis. Histologic involvement of lymph
MF is generally associated with an indolent clinical nodes (5-year disease-specific survival of 25% and
course and is characterized by polymorphic atrophic median survival of 2 years) and visceral disease are
erythematous patches and scaly plaques. Some associated with a poor prognosis (5-year disease-specific
patients progress from having limited patches and survival 18% and median survival 18 months).
254 Lymphoma, Second Edition, ed. Robert Marcus, John W. Sweetenham, and Michael E. Williams. Published by
Cambridge University Press. © Cambridge University Press 2014.
Chapter 15: Primary cutaneous lymphoma
Table 15.1 WHO–EORTC classification of cutaneous lymphomas with primary cutaneous manifestations.
Sézary syndrome
SS is defined by a clinical triad of erythroderma, var-
iable presence of peripheral lymphadenopathy and
atypical mononuclear cells (Sézary cells) comprising
>20% of total lymphocyte count or a total Sézary count
>1000 × 109/l (B2), (Figures 15.4 and 15.5), although
Figure 15.1 Mycosis fungoides, patch stage. the distinction from erythrodermic MF patients with
early blood involvement (5% or more of peripheral
blood lymphocytes (B1)) is questionable, particularly
Multivariate analysis confirmed that stage, age, male as the prognosis for B1 and B2 patients is similar. SS
gender, lactate dehydrogenase (LDH), B0b, and follicu- patients usually have a modest CD4 lymphocytosis
lotropic MF are independent prognostic factors. with a blood T-cell clone as indicated either by aber-
Common sites of visceral involvement include pulmo- rant expression of pan-T-cell antigens, cytogenetics, or
nary, nasopharynx, and CNS. TCR gene analysis helping to distinguish SS from an
All patients should have a full clinical examination inflammatory erythroderma.
and adequate diagnostic biopsies for histology, immu- Patients present with a generalized exfoliative
nophenotypic, and molecular studies. Peripheral erythroderma, ectropion, scalp alopecia, palmoplantar
blood samples should be taken for routine haematol- hyperkeratoses, and nail dystrophy associated with
ogy, biochemistry, serum LDH, Sézary cells, lympho- severe intractable pruritus. The prognosis is poor,
cyte subsets, HTLVI serology, and T-cell receptor with a median survival of 3 years and 5-year disease- 255
(TCR) gene analysis. Any palpable bulky peripheral specific survival of 30%. The majority die of
Chapter 15: Primary cutaneous lymphoma
Skin
T1 Limited patches, *papules, and/or plaques covering <10% of the skin surface
May further stratify into T1a (patch-only) versus T1b. (plaque ± patch)
T2 Patches, papules, or plaques covering ≥10% of the skin surface. May further stratify intoT2a
(patch-only) versus T2b, (plaque ± patch)
T3 One or more tumors‡ (≥1 cm diameter)
T4 Confluence of erythema covering ≥80% body surface area
Node
N0 No clinically abnormal peripheral lymph nodes§; biopsy not required
N1 Clinically abnormal peripheral lymph nodes; histopathology Dutch grade 1 or NCI LN02
N1a Clone-negative#
N1b Clone-positive#
N2 Clinically abnormal peripheral lymph nodes; histopathology Dutch grades or NCI LN3
N2a Clone-negative#
N2b Clone-positive#
N3 Clinically abnormal peripheral lymph nodes; histopathology Dutch grades 3–4 or NCI LN4;
clone-positive or -negative
Nx Clinically abnormal peripheral lymph nodes; no histologic confirmation
Visceral
Mo No visceral organ involvement
M1 Visceral involvement (must have pathology confirmation, and organ involved
should be specified)
Bleed
B0 Absence of significant blood involvement ±5% of peripheral blood lymphocytes
are atypical (Sézary) cells
B0a Clone-negative#
B0b Clone-positive#
B1 Low blood tumor burden: >5% of peripheral blood lymphocytes are atypical (Sézary)
cells but does not meet the criteria of B2
B1a Clone-negative#
B1b Clone-positive#
B1 High blood tumor burden: ≥1000/μL Sézary cells with positive clone#
In pagetoid reticulosis there is an acanthotic epi- lymphocytes that are either scattered individually (NCI
dermis that is infiltrated by halo cells, either individu- LN2) or present in small clusters (NCI LN3). With
ally or in small clusters. more advanced infiltration of the nodes there is partial
Granulomatous slack skin shows a band-like infil- effacement of the node by neoplastic cells with a mainly
trate in early stages, but as the disease progresses there is paracortical distribution or diffuse infiltration and total
infiltration of the dermis by small cells. In all cases there effacement of the node (NCI LN4).
is an associated population of multinucleate giant his- The typical immunophenotype is that of a mature
tiocytic cells that often contains elastic fibers (elastoly- T-cell and shows staining for CD2, CD3, and CD5. The
sis) or lymphocytes within their cytoplasm. cells are usually CD4-positive, with lack of CD8
In SS the features can be similar to classical MF, although CD8-positive cases are described, particularly
with epidermotropism and Pautrier microabscesses in childhood MF and hypopigmented variants. There is
present, but the pathology may also be non-diagnostic usually no expression of CD30 in early stages of disease
in almost one-third of patients with proven SS. but TCRαβ is usually expressed. There is frequent loss
Lymph node involvement by MF is characterized by of CD7 and as the disease progresses there may also be 257
dermatopathic type changes (NCI LN1), with atypical loss of CD2 and CD5, particularly in the intraepidermal
Table 15.3 Prognosis in CTCL and CBCL.
Stage IA IB IIA IIB IIIA/ IVA1/ IVB LYP pcALCL SPTCL pcPTL, Pc Pc small/
MF/SS or IIIB IVA2 unspecified epidermotropic medium
disease CD8+ CTCL pleomorphic
entity CTCL
Disease- 98 89 89 56 54/48 41/23 18 100 90–95 82 16 18 60–80
specific
survival at 5
years (%)
Disease- 95 77 67 42 45 20 –
specific
survival at
10 years (%)
Median 35.5 21.5 15.8 4.7 4.7/3.4 3.8/2.1 1.4
survival
(years)
Risk of 8% 21% 17% 48% 53/82% 62/77% 82%
disease
progression
at 5 years
Prognosis in CBCL.
Disease entity pcMZL pcFCL pcLCL
Disease-specific survival 5 years (%) 99 95 50–55
Chapter 15: Primary cutaneous lymphoma
PUVA therapy is commonly reported and patients preventing relapse and therefore should, if possible, be
often require repeated courses over many years. Many avoided, particularly as this group of patients already
patients will inevitably have a high total cumulative have an increased risk of secondary malignancies.
UVA dose and the risks of non-melanoma and mela- Effort should be made to restrict the total PUVA
noma skin cancer are consequently increased for these dose to less than 200 treatment sessions or a total
262
patients. Maintenance therapy is only rarely effective at cumulative dose of 1200 J/cm2. However, in
Chapter 15: Primary cutaneous lymphoma
exceptional circumstances patients may receive a before being returned to the patient. This regimen is
greater total dose if clinically justified and with the repeated on two successive days and the 2-day cycle
consent of the patient. repeated monthly or fortnightly. In vitro evidence
suggests a proportion of the UVA-exposed leukocytes,
Radiotherapy and total skin electron beam therapy including some tumor lymphocytes, undergo apopto-
MF and other CTCL variants are extremely radiosen- sis and that dendritic cells are activated during the ex
sitive. Individual thick plaques, eroded plaques, or vivo circulation with induction of a host antitumor
tumors can be treated successfully with localized and immune response after the treated cells are returned to
relatively low-dose superficial orthovoltage radiother- the patient. Recent evidence has also shown activation
apy (2 or 3 fractions of 400 cGy at 80–120 kV). Large of dendritic cells during an expanded period of ex vivo
tumors may require a higher energy. Closely adjacent incubation overnight (transimmunization).
and overlapping fields can often be retreated because ECP is FDA-approved for the treatment of CTCL,
of the low doses used. but there are no randomized studies to clarify
Electron beam therapy is also used to treat local whether ECP has an impact on OS. The original
disease and the energy used is dependent on the depth open study of ECP in 29 patients with erythrodermic
of disease. Whole-body total skin electron beam CTCL (stage III/IVA) reported a response rate of
(TSEB) therapy has also been extensively used to 73%, but response rates in patients with earlier stages
treat resistant widespread cutaneous disease. were much lower (38%). Subsequently, a median
A meta-analysis of open uncontrolled and mostly survival of 62 months was reported in the original
retrospective studies of TSEB as monotherapy in 952 cohort of 29 erythrodermic patients, which compares
patients with CTCL established that responses are favorably with historical controls (30 months),
stage-dependent, with CR of 96% in stage IA–IIA although a later study of 33 patients with SS treated
disease, but relapse rates are high, indicating that this with ECP reported a median survival of 39 months. A
approach is not curative in early-stage disease. In stage systematic review of response rates in erythrodermic
IIB disease CRs are less common (36%), but erythro- disease (stage III/IVA) with ECP has shown overall
dermic (stage III) disease shows CR of 60%. Greater responses of 35–71%, with CRs of 14–26%. ECP has
skin surface dose fractionation (32–36 Gy) and higher not been found to be of benefit in stage IB patients
energy (4–6 MeV electrons) are associated with a who have molecular evidence of a peripheral blood
higher rate of CR, and 5-year relapse-free survivals of clone. There have been claims that the CD8 count is
10–23% have been reported. A comparative retrospec- critical in predicting whether patients will respond to
tive study of TSEB versus topical mechlorethamine in ECP, although others have provided evidence that the
early stage MF showed similar response rates and total baseline Sézary count is the best predictor of
duration of response, suggesting that TSEB therapy response.
should be reserved for those who fail first and second Preliminary non-randomized cohort studies suggest
line therapies. that the combination of interferon-α and ECP is more
Adverse effects of TSEB include temporary alope- effective than ECP alone. There are also isolated case
cia, telangiectasia, and skin malignancies. Although reports of combined ECP and interferon-γ which have
TSEB is usually only given once in a lifetime, several induced complete clinical and molecular responses.
reports have documented patients who have received
two or three courses, although the total dose tolerated Immunotherapy
and duration of response have been lower with sub- Immunotherapy may enhance the antitumor host
sequent courses. immune response by promoting the generation of
cytotoxic T-cells and TH1 cytokine responses.
Systemic therapies Interferon-α has overall response rates (ORRs) of
45–74% and CRRs of 10–27%.Various dosage sched-
Extracorporeal photopheresis ules have been employed (3MU × 3/week–36 MU/
Extracorporeal photopheresis (ECP) involves admin- day), and it appears that response rates are higher for
istration of oral psoralen, followed by ex vivo collec- larger doses (overall responses of 78% compared to
tion of an enriched buffy coat preparation using a cell 37% for the lower dose schedule), but dose-limiting 263
separator. These leukocytes are then exposed to UVA toxicities include fever, chills, malaise, leukopenia,
Chapter 15: Primary cutaneous lymphoma
in 25% of patients but can be prevented by concurrent antibody has also shown objective responses in small
steroid therapy. Myelosuppression is rare. cohorts. A variety of other agents have shown efficacy in
small phase II studies, including pralatrexate (folate
Histone deacetylase (HDAC) inhibitors antagonist) and bortezomib (proteasome inhibitor),
Both the HDAC inhibitors oral suberoylanilide but pivotal studies of these agents are awaited.
hydroxamic acid (SAHA; Vorinostat) and intrave-
nous depsipeptide (Romidepsin) are FDA-approved
for advanced MF/SS. In a phase II open label study of
MF clinicopathologic variants
oral SAHA in 74 patients, ORRs of 30% have been
reported.
Pagetoid reticulosis
A phase II open study of Romidepsin in 96 patients Pagetoid reticulosis is a rare localized solitary variant
reported an ORR of 34%, with 6% CRs and a median of CTCL, characterized clinically by an isolated, per-
duration of response of 15 months. The most common sistent scaly plaque, commonly involving an acral site
adverse events were fatigue, diarrhea, nausea, non- and histologically by intense epidermotropism and an
specific and reversible electrocardiogram changes, aberrant T-cell phenotype. A more generalized variant
and thrombocytopenia. with multiple plaques at other sites has also been
described, but it is likely that this represents the
aggressive CD8+ epidermotropic CTCL variant. The
Bone marrow transplantation
natural history of pagetoid reticulosis is of very slow
Small cohort studies assessing the use of TSEB and or local extension with an excellent prognosis.
total body irradiation combined with high-dose con- Treatment: successful remission and cure has been
ditioning chemotherapy prior to autologous stem cell reported with both surgical excision and low-dose
transplantation (ASCT) in patients with stage IIB/ superficial radiotherapy.
IVA disease have shown good clinical responses, but
high relapse rates, and autologous transplants are not
currently considered for MF/SS patients. Folliculotropic and syringotropic MF
Myeloablative allogeneic stem cell or bone marrow Folliculotropic and syringotropic clinicopathologic var-
transplantation has only been used in a few patients, iants of MF are common. Clinically these variants con-
with encouraging results, but the associated mortality sist of multiple small (1–3 mm) red macules and papules
suggests that this approach is difficult to justify. clustered around hair follicles and frequently associated
However, a graft-versus-lymphoma effect may be with alopecia. Histologically the atypical lymphocytes
therapeutically important and results of non- show tropism for the hair follicle or sweat gland epithe-
myeloablative allogeneic transplantation for selected lium but may demonstrate epidermotropism.
patients with advanced stages of MF/SS have shown Follicular mucinosis may occur in association with
very encouraging results. A review of the European MF or as a separate entity. Clinically there is follicular
Group for Blood and Marrow Transplantation plugging with hair loss, and boggy erythematous pla-
(EBMT) experience of allogeneic SCT in 60 patients ques may be present. Histologically, mucinous degen-
with MF/SS has confirmed that RIC allo-SCT has eration can be seen deposited throughout the follicular
lower transplant-related mortality than myeloabla- epithelium. If the hair follicles have been destroyed,
tive allo-SCT and that conditioning, including scarring alopecia may occur. Studies using multivari-
T-cell depletion, increases the risk of early relapse ate analysis suggest that the prognosis of folliculo-
significantly. Reported non-relapse mortality is tropic variants of MF are significantly worse.
approximately 25% at year 1, with OS of 50% at Treatment: because of the perifollicular localization
year 3. of the dermal infiltrates, folliculotropic MF is often less
responsive to skin-targeted therapies, such as PUVA
Emerging novel therapies and topical nitrogen mustard, than classical plaque-
A humanized chimeric anti-CD4 monoclonal antibody stage MF, but localized disease may respond well to
(HuMax-CD4/zanolimumab) has shown response rates radiotherapy. In patients with extensive disease, TSEB
of 56%. Patients with a clinical response develop an irradiation is an effective treatment, but sustained com-
eczema-like reaction and CD4 depletion may be pro- plete remissions are rarely achieved. PUVA combined 265
longed for up to 1 year. A radiolabeled anti-CD5 with retinoids or interferon-α should also be considered.
Chapter 15: Primary cutaneous lymphoma
Granulomatous variants of MF
GSS presents with redundant erythematous folds typ-
ically in flexural sites. The histology is granulomatous,
with extensive areas of elastolysis, and the atypical
lymphocytic infiltrate may be sparse and may not
show epidermotropism. The condition must be distin-
guished from other forms of cutis laxa. While GSS is
considered to be a variant of MF, patients rarely show
typical MF patches and plaques. Granulomatous infil-
tration has also been described in plaques of MF with-
out the typical features of slack skin.
Treatment: GSS responds to radiotherapy. Rapid
recurrences after surgical excision have been reported. Figure 15.8 Lymphomatoid papulosis.
Primary cutaneous (anaplastic) CD30- may have multiple nucleoli. There may be an associ-
ated infiltrate of small lymphocytes but eosinophils
positive large-cell lymphoma and plasma cells are not prominent.
Primary cutaneous CD30-positive anaplastic large-cell Immunophenotypically CD30 is present in over
lymphoma (pcALCL) is usually seen in adults and 75% of large cells. These cells also express CD2, CD3,
presents as large solitary nodules or plaques which CD5, and CD25, although variable loss of CD2, CD3,
may ulcerate. Multiple lesions may occur but, crucially, and CD5 is common. The cells are positive for cyto-
there are no associated typical atrophic patches or pla- toxic granules (TIA-1 and granzyme) and MUM-1 but
ques characteristic of MF. Individual lesions may par- do not express epithelial membrane antigen (EMA)
tially regress spontaneously, and disease may remain and are negative for ALK.
localized to a limb. Progression to extracutaneous sites Treatment: both excision and localized radiother-
is rare but has been reported in 10%. The prognosis for apy are appropriate for isolated lesions. Nodal involve-
patients with primary cutaneous disease is excellent ment may also respond to radiotherapy. Oral
(5-year survival rates of 90–95%; Table 15.3). methotrexate is also effective. Multiagent chemother-
However, staging consisting of bone marrow and CT apy, including CHOP, can be effective, but is not rec-
scans is required to exclude secondary cutaneous ommended unless extracutaneous sites are involved. A
involvement in primary nodal CD30+ anaplastic lym- phase II study of anti-CD30 antibodies in pcCD30+
phoma kinase(ALK)+ lymphomas (Figure 15.9). CTCL variants has shown responses in 16 of 23 patients,
with 10 patients achieving a complete remission, includ-
Pathology ing patients with pcALCL. Therapeutic options are
Morphologically these lesions are characterized by a shown in Table 15.4.
dense dermal infiltrate of large cells with immunoblas-
tic, pleomorphic, or anaplastic morphology. Molecular pathology of CD30+ T-cell
Epidermotropism may be seen. Large bizarre cells lymphoproliferative disorders of the skin (CD30+ LPD)
with irregular nuclei and multiple nucleoli and abun- TCR genes are clonally rearranged in the vast majority of 267
dant pale cytoplasm are present, and some of these cases. Nucleophosmin (NPM)–ALK fusion transcripts
Chapter 15: Primary cutaneous lymphoma
lymphoma
Subcutaneous panniculitis-like T-cell lymphoma Immunophenotypically the cells express a mature
(SPTCL) is a lymphoproliferative disease originating T-cell phenotype and express CD2, CD3, and CD5. The
and presenting in the subcutaneous tissue and predom- cells are characteristically positive for CD8, lack CD4,
inantly affecting adults. It is characterized clinically by and express cytotoxic granule-related proteins TIA-1,
solitary or multiple subcutaneous nodules and plaques, granzyme, and perforin. CD30 is usually negative. The
which mainly involve the legs. SPTCL is an indolent cells express TCRαβ and are EBV-negative. In contrast,
lymphoma derived from alpha/beta T-cells that are primary cutaneous γδ T-cell lymphomas are usually
CD8-positive and show histologic involvement con- negative for CD4 and CD8 and may express CD56.
fined to the subcutis. This CTCL variant should be Treatment: recent studies suggest that most
distinguished from a primary cutaneous gamma/delta patients have an indolent disease that can be success-
T-lymphoma which has a poor prognosis and is usually fully treated with systemic corticosteroids and/or local
CD56-positive with histologic evidence of both epider- radiation therapy. However, in patients with extensive
motropic and subcutaneous involvement. refractory disease, doxorubicin-based chemotherapy
A hemophagocytic syndrome is a recognized com- may be appropriate. Specifically in those with primary
plication of SPTCL and cutaneous γδ T-cell lympho- cutaneous γδ T-cell lymphoma, preconditioning che-
mas in which patients present with fever, motherapy followed by an autologous/allogeneic bone
pancytopenia, and hepatosplenomegaly, and this may marrow transplant may be the only therapeutic option
herald a fulminant clinical course. Dissemination to for improving survival. Therapeutic options are
lymph nodes and other organs is uncommon. shown in Table 15.4.
Further reading
Agar N, Wedgeworth E, Crichton S, et al. Survival
outcomes and prognostic factors in mycosis
fungoides and Sézary syndrome: validation of the
revised International Society for Cutaneous
Lymphomas and the European Organisation for
Treatment and Research staging proposal. J Clin Oncol,
2010;28:4730–4739.
Bekkenk MW, Geelen FAMJ, van Voorst Vader PC, et al.
Primary and secondary cutaneous CD30 positive
lymphoproliferative disorders. Blood, 271
Figure 15.12 Large B-cell lymphoma involving the leg. 2000:95;3653–3661.
Chapter 15: Primary cutaneous lymphoma
Duarte R, Canals C, Onida F, et al. Allogeneic hematopoietic cutaneous T-cell lymphoma. J Clin Oncol,
cell transplantation for patients with mycosis fungoides 2010;28:1870–1877.
and Sézary syndrome: a retrospective analysis of the Quereux G, Marques S, Nguyen J-M, et al. Prospective
Lymphoma Working Party of the European group for multicenter study of pegylated liposomal doxorubicin
Blood and Bone Marrow transplantation. J Clin Oncol, treatment in patients with advanced or refractory mycosis
2010;28:4492–4499. fungoides or Sézary syndrome. Arch Dermatol,
Duvic M, Hymes K, Heald P, et al. Bexarotene is effective and 2008;144:727–733.
safe for treatment of refractory advanced-stage cutaneous Senff N, Hoefnagel J, Jansen P, et al. Reclassification of 300
T-cell lymphoma: multinational phase II–III trial results. primary cutaneous B-cell lymphomas according to the
J Clin Oncol, 2001;19:2456–2471. new WHO-EORTC classification for cutaneous
Jones GW, Kacinski BM, Wilson LD, et al. Total skin electron lymphomas: comparison with previous classifications
radiation in the management of mycosis fungoides: and identification of prognostic markers. J Clin Oncol,
consensus of the European Organization for Research and 2007;25:1581–1587.
Treatment of Cancer (EORTC) Cutaneous Lymphoma Senff N, Hoefnagel J, Nelis K, et al. Results of radiotherapy in
Project Group. J Am Acad Dermatol, 2002;47(3):364–370. 153 primary cutaneous B-cell lymphomas classified
Kaye FJ, Bunn PA Jr, Steinberg SM, et al. A randomized trial according to the WHO-EORTC classification. Arch
comparing combination electron beam radiation and Dermatol, 2007;143:1520–1526.
chemotherapy with topical therapy in the initial Senff N, Noordijk E, Kim Y, et al. European Organization
treatment of mycosis fungoides. N Engl J Med, for Research and Treatment of Cancer and
1989;321:1784–1790. International Society for Cutaneous Lymphoma
Kim Y, Willemze R, Pimpinelli N, et al. TNM consensus recommendations for the management of
classification system for primary cutaneous lymphomas cutaneous B-cell lymphomas. Blood,
other than mycosis fungoides and Sézary syndrome: a 2008;112:1600–1609.
proposal of the International Society for Cutaneous Trautinger F, Knobler R, Willemze R, et al. EORTC
Lymphoma (ISCL) and the Cutaneous Lymphoma Task consensus recommendations for the treatment of
Force of the European Organisation for Research and mycosis fungoides/Sézary syndrome. Eur J Cancer,
Treatment of Cancer (EORTC). Blood, 2006;42:1014–1030.
2007;110:479–484.
Vonderheid EC, Bernengo MG, Burg G, et al. Update on
Lundin J, Hagberg H, Repp R, et al. Phase 2 study of erythrodermic cutaneous T-cell lymphoma: report of the
alemtuzumab (anti-CD52 monoclonal antibody) in International Society for Cutaneous Lymphomas. J Am
patients with advanced mycosis fungoides/Sézary Acad Dermatol, 2002;46(1):95–106.
syndrome. Blood, 2003;101:4267–4272.
Whittaker S, Demierre MF, Kim E, et al. Final results from a
Olsen E, Kim Y, Kuzel T, et al. Phase IIB multicenter trial of
multicentre international pivotal study of romidepsin in
vorinostat in patients with persistent, progressive or
refractory cutaneous T-cell lymphoma. J Clin Oncol,
treatment refractory cutaneous T-cell lymphoma. J Clin
2010;28:4485–4491.
Oncol, 2007;25:1–9.
Willemze R, Jaffe ES, Burg G. WHO-EORTC
Olsen E, Vonderheid E, Pimpinelli N, et al. Revisions to the
classification for cutaneous lymphomas. Blood,
staging and classification of mycosis fungoides and
2005;105(10):3768–3785.
Sézary syndrome: a proposal of the International Society
for Cutaneous Lymphomas (ISCL) and the cutaneous Willemze R, Jansen P, Cerroni L, et al. Subcutaneous
lymphoma task force of the European Organisation of panniculitis-like T-cell lymphoma: definition,
Research and Treatment of Cancer (EORTC). Blood, classification and prognostic factors: an EORTC
2007;110:1713–1722. cutaneous lymphoma group study of 83 cases. Blood,
2008;111:838–845.
Pimipinelli N, Olsen E, Santucci M, et al. Defining early
mycosis fungoides. J Am Acad Dermatol, Yu J, McNiff J, Lund M, Wilson L. Treatment of primary
2005;53:1053–1063. cutaneous CD30+ anaplastic large cell lymphoma with
radiation therapy. Int J Radiat Oncol Biol Phys,
Prince M, Whittaker S, Hoppe R. How we treat 2008;70:1542–1545.
mycosis fungoides and Sézary syndrome. Blood,
2009;114:4337–4353. Zinzani P, Baliva G, Magagnoli M, et al.
Gemcitabine treatment in pretreated cutaneous
Prince M, Duvic M, Martin A, et al. Phase III placebo T-cell lymphoma: experience in 44 patients. J Clin
controlled trial of denileukin diftitox for patients with Oncol, 2000;18:2603–2606.
272
Chapter
16
Lymphoma in the immunosuppressed
Michele Spina, Robert Marcus, Shireen Kassam, and Umberto Tirelli
Pathology: Andrew Wotherspoon Molecular cytogenetics: Andreas Rosenwald and German Ott
Immunodeficient patients are at higher risk for devel- According to the WHO classification, B-cell
oping lymphoproliferative disorders (LPDs), above all PLTDs can be divided into three main groups:
non-Hodgkin lymphomas (NHLs). Even though the
association between primary immunodeficiency dis- 1. Diffuse B-cell hyperplasia, characterized by
eases (i.e. X-linked lymphoproliferative syndrome, differentiated plasma cells and preservation of the
common variable immunodeficiency, ataxia telangiec- normal lymphoid architecture.
tasia, and Wiskott–Aldrich syndrome) and LPDs, on 2. Polymorphic PTLDs characterized by nuclear
the one hand, and between LPDs and autoimmune atypia, tumor necrosis, and destruction of the
diseases, on the other, is well known, the leading causes underlying lymphoid architecture.
of immunosuppression are, at present, organ trans- 3. Monomorphic PTLDs, which includes high-grade
plantation and human immunodeficiency virus (HIV) invasive lymphoma of B- or T-lymphocyte
infection. Lymphomas in the immunocompromised centroblasts.
host have been traditionally regarded as having a Monomorphic B-cell PTLDs can be further divided
more aggressive course, including extranodal involve- into diffuse large cell lymphomas and Burkitt or
ment, a more rapid clinical course, poorer response to Burkitt-like lymphomas that display their character-
conventional therapies, and poorer outcome, but this istic chromosomal translocations. Monomorphic
has been called into question both in post-transplant T-cell PTLDs can be further divided into large cell,
lymphoproliferative disorders (PTLD) following solid anaplastic, or unspecified.
organ transplant (SOT) and HIV-related lymphoma Hodgkin lymphoma (HL) is rare but well described
by more recent publications. in this setting. EBV is closely involved in the patho-
genesis and the majority of cases arise in response to
primary infection with EBV or to reactivation of pre-
Tumors following solid organ viously acquired EBV.
transplantation (PTLD) Other low-grade B-cell lymphomas of B-cell ori-
PTLDs are well recognized and potentially life- gin; marginal zone lymphoma and lymphoplasma-
threatening complications after SOT. PTLD is seen cytoid lymphoma also have an increased incidence
in up to 10% of all SOT recipients. It is the most after SOT and are associated with prior exposure to
common form of post-transplant malignancy in chil- Helicobacter pylori and hepatitis C, respectively.
dren and in adults it is the second most common The incidence of PTLD after SOT varies between
malignancy after skin cancer. In both children and children and adults and also according to the organ
adults it is the most common cause of cancer-related transplanted. In adult recipients, PTLD occurs in
mortality after SOT, with the reported overall mortal- 1–2.3% of kidney transplants, 1–2.8% of liver trans-
ity for PTLD exceeding 50%. Up to 85% of PTLDs plants, 1–6.3% of heart transplants, 2.4–5.8% of heart–
are of B-cell lineage and most of these (over 80%) lung transplants, 4.2–10% of lung transplants, and up
are associated with Epstein–Barr virus (EBV) to 20% of small bowel transplants. The incidence of
infection. Around 10–15% of PTLDs are of T-cell PTLD is significantly higher in pediatric recipients,
lineage, around 30% of which are associated with EBV. where it has been reported in 1.2–10.1% of kidney
Lymphoma, Second Edition, ed. Robert Marcus, John W. Sweetenham, and Michael E. Williams. Published by 273
Cambridge University Press. © Cambridge University Press 2014.
Chapter 16: Lymphoma in the immunosuppressed
transplants, 4–15% of liver transplants, and 6.4–19.5% polymorphic PTLD seen in adults, with a close associ-
of lung, heart, and heart–lung transplants. ation between rising EBV copy number and the devel-
Although PTLD may occur at any time after trans- opment of PTLD. There is also some preliminary
plantation, the risk of developing PTLD is greatest evidence of an association of specific HLA subtypes
within the first year and declines over time. A report in the recipient that may be protective against PTLD
by the Transplant Collaborative Study showed the development.
incidence of PTLD to be 224/100 000 in the first Patients who have compromised immune systems
year, 54/100 000 in the second year, and 31/100 000 prior to transplantation, such as the elderly or those
in the sixth year following transplantation. A more with pre-existing CMV infection or hepatitis C, may be
recent study from the French Transplant Registry sug- particularly vulnerable to over-immunosuppression
gests an overall incidence in renal transplant recipients and hence PTLD. Recently, prognostic indices have
of between 0.2% and 0.4%, with a recent decline in been developed that differ subtly but definitely from
incidence perhaps related to lower intensity immuno- those devised for diffuse large B-cell lymphoma
suppression (IS) being used in the post-transplant (DLBCL) in the non-immunosuppressed context. The
setting and a confirmation of an increased association simplest of these was devised by Choquet and col-
with both antilymphocyte globulin (ATG) usage and leagues, and consists of Eastern Cooperative Oncology
prior EBV exposure. A biphasic distribution was seen Group (ECOG) performance status 0–1 versus 2–4,
in renal transplant recipients, with the incidence rising elevated lactate dehydrogenase (LDH), and age >60 or
in the first year, declining in years 2–4, and rising again <60 years. Patients with no risk factors had a 2-year
thereafter. These findings were also observed in a large survival of 88%, those with a single risk factor 50%, and
US registry series. in those with two or three risk factors there were no
PTLD arises, in large part, as a consequence of the long-term survivors.
potent immunosuppressive agents necessary to pre- A later refinement of this approach in a larger
vent allograft rejection. The risk following administra- series recently published by Caillard and colleagues
tion of specific immunosuppressive agents is still a confirms that age, LDH, and stage of disease are
matter of controversy. The incidence of lymphoma major prognostic factors, together with extranodal
was found to increase after the introduction of cyclo- sites, including CNS, also having a major impact on
sporine but may be associated with other cofactors. In survival, with an overall survival (OS) in this 500-
other large series, treatment with cyclosporine did not patient series of just 50% at 5 years. Superior results
enhance the risk for lymphoma compared to azathio- have been obtained since the introduction of
prine. However, a number of small retrospective stud- rituximab.
ies have suggested that the routine introduction of Diagnosis can only be made on biopsy of affected
cyclosporine as an immunosuppressive agent in the tissues, and suspicion may also be raised by the classi-
1980s was associated with an increased incidence of cal systemic symptoms of lymphoma: fever, sweats,
PTLD. Experience with tacrolimus is more limited, weight loss, or pruritus, or, specifically in the PTLD
although an increased incidence of lymphomas has context, a rise in EBV copy number or deterioration of
been reported both in adults and pediatric patients. It organ function. Staging should include CT and PET
has been speculated that the overall intensity of immu- where available, bone marrow biopsy and MRI of CNS
nosuppression, rather than an individual agent, plays a where indicated. Assessment of comorbidities (specif-
key role in the development of lymphoma, which is ically cardiac, lung, and renal function) will need to be
supported by the observation of a higher incidence of assessed prior to commencement of any therapeutic
lymphoma in non-renal transplant recipients receiv- strategy.
ing higher doses of immunosuppression. Patients Surgical resection and/or local radiotherapy
receiving monoclonal antibodies, such as OKT3 or should be considered in the rare patient with localized
ATG, demonstrate a significantly higher incidence of disease.
PTLD. There is also some evidence that the high early The initial treatment in all patients with PTLD is
incidence of aggressive PTLD in heart–lung recipients to reduce immunosuppression in the hope that rising
relates to the large volume of potentially EBV-infected T-cell numbers will induce antitumor activity. The
lymphoid tissue. EBV plays a major role in the patho- approach to reducing immunosuppressive therapy
274 genesis of PTLD in the pediatric context and in early needs to be carefully individualized and will depend
Chapter 16: Lymphoma in the immunosuppressed
on the nature and extent of disease and the type of proportion of patients; they may pose particular prob-
transplant recipient, i.e. whether they have a life- lems for patients with compromised cardiac function.
supporting graft (e.g. heart or heart–lung, small These observations led to a prospective European
bowel, or liver) or a non-life-supporting graft (e.g. study where all patients with PTLD were treated first
kidney). Cyclosporine or tacrolimus dosage are typi- with rituximab as a single agent; those responding
cally reduced over 4–6 weeks to give whole blood underwent short-course rituximab maintenance ther-
trough levels of around 25–50% that of normal ther- apy, while those that did not received four cycles of
apeutic levels, depending on the extent of disease and R-CHOP chemotherapy. A total of 91 patients have
type of graft, and transplanted organ function carefully now been reported; the trial included patients with
monitored. Myelosuppressive drugs such as azathio- both early and late PTLD, with the majority having
prine or mycophenolate mofetil should be reduced or undergone renal or liver transplantation. Interestingly,
stopped if at all possible. A response to reduction in and contrary to early reports, almost 50% of the
immunosuppression is usually seen within 2–4 weeks, so-called “late” patients had evidence of persistent EBV
with long-term remissions in around 50% of patients. positivity. The overall response rate (ORR) to single
The traditional approach has been to regard poly- agent rituximab was 54%, with 32% of patients entering
clonal PTLD developing during the first year after complete remission (CR). Those patients not entering
transplantation (early disease) as likely to be respon- CR then underwent four cycles of R-CHOP chemother-
sive to reduced immunosuppression, whereas mono- apy. The ORR was >90%, with similar numbers disease-
clonal tumors are generally regarded as being free at 1 year. Long-term follow-up data are awaited.
unresponsive to a reduction in immunosuppression A retrospective study from Evens and colleagues
(RIS) and require more intensive approaches. also provides useful guidance. Eighty patients with
The first systematic study to test this was published PTLD (with a median time from transplant to PTLD
by Tsai and colleagues. In their series, patients treated of 48 months) were reported. All received RIS as initial
with RIS, patients with organ dysfunction, a raised strategy; 59 then received rituximab ± chemotherapy,
LDH, and multisite disease fared poorly, whereas with a 3-year PFS of 57% and OS of 73%. Once again,
those patients with none of these factors had a 90% 22/30 of the “early” patients were EBV-positive and 17/
response rate. RIS then is a prerequisite for subsequent 50 of the “late” patients were EBV-positive in this series.
therapy but may be insufficient in itself. Current UK and National Comprehensive Cancer
Another approach has been to treat PTLD identi- Network (NCCN) guidelines follow these results: after
cally to NHL arising in the non-transplant context. An staging, all patients with multifocal disease are recom-
early series of patients with B-cell PTLD treated with mended to receive RIS as initial therapy; those enter-
the conventional CHOP (cyclophosphamide, doxoru- ing CR receive no further therapy; patients with
bicin, vincristine, prednisolone) regimen showed a residual or progressive disease then receive single
5-year progression-free survival (PFS) of 43%, with agent rituximab, and, for those not entering remission,
the majority of failures due to recurrence. R-CHOP is recommended. Special provision will need
In the rituximab era a number of approaches have to be made for those patients with heart transplants
been used to assess the value of rituximab alone or where anthracyclines may be contraindicated.
with chemotherapy. Such results are, of course, only
applicable to the majority of patients with B-cell
PTLD. The recommended therapies for T-cell and
T-cell post-transplant
Hodgkin PTLD are discussed below. lymphoproliferative disorders
The first of these was published by Choquet and This is a rare complication, with fewer than 15% of
colleagues: 46 patients were treated with rituximab patients having T-cell phenotype in most series. The
alone, with a response rate of 44% and a 5-year survival outlook here is poorer than in B-cell PTLD, perhaps
of 67%. Once again, raised LDH, age, and multisite reflecting the increased chemotherapy resistance of
disease were associated with a poor outcome. T-cell tumors generally and the lack of an effective
In a small series of 18 PTLD patients, Taylor et al. monoclonal antibody strategy in this group. Published
demonstrated a high response with R-CHOP, with a series are rare but the general consensus is that patients
62% PFS at 5 years. Such intensive regimens are not should be treated with RIS followed by CHOP or
without toxicity and may not be necessary in a CHOP-like therapy. 275
Chapter 16: Lymphoma in the immunosuppressed
in combination with chemotherapy in patients with AMC studies mentioned above suggests that the addi-
HIV–NHL. The Italian Cooperative Group on AIDS tion of rituximab to the infusional regimen, DA-
and Tumors (GICAT) performed a study on the admin- EPOCH, results in a better event-free survival and
istration of R-CDE with concomitant HAART in 74 OS than R-CHOP. However, this is a post-hoc analysis
patients, and 70% achieved a CR. Treatment-related of data from two studies in which patients were treated
toxicity was acceptable for this patient setting except over different time periods and thus other factors,
for infectious events. After a median follow-up of 16 including differences in supportive care, may have
months (range 1–57 months), median survival has not confounded the results. On the basis of these results,
been reached, yet, and 2-year OS rate is 62%, DFS is however, DA-EPOCH plus rituximab is considered
89%, and PFS is 86%. Other studies have evaluated the the chemotherapy regimen of choice in the USA,
effectiveness of rituximab in combination with chemo- whereas in Europe R-CHOP remains the standard of
therapy in the treatment of HIV–NHL. The first trial care. A randomized trial of these treatment regimens is
took place in France and used R-CHOP to treat 61 required to truly answer the question of superiority.
patients, with a 77% CR rate and a 2-year OS of 75% An additional controversy, and one that has not
and PFS of 69%. A phase II study was performed in been explored in a randomized study, is whether
Spain on the same treatment regimen. Out of 81 HAART should be continued during chemotherapy
patients the following rates were achieved: CR 69%, or not. In the USA, centers that use the DA-EPOCH
3-year OS 56%. The only randomized phase III study regimen suspend cART because of potential adverse
of CHOP with and without rituximab, conducted by the pharmacokinetics and pharmacodynamics interac-
AIDS-Malignancies Consortium (AMC) in the USA, tions with chemotherapy and the potential for
led to controversy as to the benefit of rituximab. In all, increased toxicity. This strategy does not appear to
150 patients were enrolled and no differences were affect lymphoma outcomes adversely or increase
observed in CR (58% in R-CHOP versus 47% in infectious complications. In Europe, it is usual to
CHOP group) and OS rates. However, there was a continue HAART during chemotherapy, avoiding
significantly increased risk of death from infectious ritonavir-boosted regimens wherever possible, as
complications in the R-CHOP group compared to the they are associated with greater toxicity.
CHOP alone group (14% versus 2%; P = 0.035).
However, death from infection occurred mainly in
those with a CD4 count below 50 cells/mm3, and Burkitt lymphoma (BL)
many patients received maintenance rituximab follow- Even in the HAART era, treatment of BL remains
ing completion of chemotherapy, an approach not cur- challenging since cure requires the use of short-course,
rently used in the HIV-negative setting. Subsequently, dose-intense chemotherapy regimens such as hyper-
the AMC conducted a randomized phase II study of CVAD (cyclophosphamide, vincristine, doxorubicin,
DA-EPOCH with either concurrent or sequential ritux- dexamethasone alternating with methotrexate and
imab. It is worth mentioning that many centers in the cytarabine) and CODOX-M/IVAC. CHOP-based reg-
USA favor the infusional DA-EPOCH regimen over imens, as used for DLBCL, have little efficacy in BL,
CHOP in HIV-related DLBCL as some phase II studies producing survival rates of only a few months. There
demonstrated a better response rate (RR) and OS. The has been reluctance to use more intensive chemother-
addition of rituximab to this regimen did not result in apy regimens for HIV-associated BL because of con-
increased risk of death from infection, and concurrent cerns about toxicity, in particular the increased risk of
administration of rituximab was superior to sequential opportunistic infections. However, evidence is emerg-
administration. A recent meta-analysis of prospective ing to suggest that these regimens can be safely deliv-
studies has confirmed the benefit in RR and OS of the ered, achieving similar response rates and OS as
addition of rituximab to chemotherapy. Therefore, the HIV-negative patients with BL. American and
addition of rituximab to chemotherapy is considered Spanish investigators report a 63–71% CR rate, a
the gold standard for the treatment of patients with HIV 46–60% failure-free survival rate at 2 years, and the
infection and DLBCL. same toxicity as in the general population. A retro-
Whether infusional chemotherapy could be con- spective study addresses concerns regarding pro-
sidered more effective than classical chemotherapy is longed immunosuppression and loss of viral control
still a matter of debate. A pooled analysis of both the following intensive chemotherapy. Excellent immune 277
Chapter 16: Lymphoma in the immunosuppressed
recovery was demonstrated in 30 HIV-positive to the PEL setting. However, the improvement in
patients with BL treated with CODOX-M/IVAC. The patients with PEL is below what has been reported in
viral load was undetectable in 88% and 87% of patients patients with HIV infection and DLBCL. As mentioned
at 6 and 12 months, respectively, following chemo- before, clinical series report the prognostic impact of
therapy. The CD4 count was greater than 200/mL in HAART, in association with chemotherapy, on PEL
58% and 80% of patients at 6 and 12 months, outcome. Complete remissions of PEL have been
respectively. reported after implementation of HAART alone, with-
Given the benefits of rituximab in the treatment of out chemotherapeutic agents. Although this experience
DLBCL and the strong expression of CD20 in Burkitt is based on single case reports only, a role for immune
cells, its use in HIV-negative BL is becoming more reconstitution in control of this aggressive lymphoma
widespread. However, data in the HIV-positive setting has been suggested. Thus, initiating or continuing
is limited. A prospective study of 36 patients with BL HAART as part of supportive therapy is recommended
treated with intensive chemotherapy and rituximab when commencing treatment for HIV-positive patients
included 19 HIV-positive patients. Although HIV- with PEL. High-dose chemotherapy with autologous
positive patients experienced more severe mucositis stem cell transplant has been reported in only two indi-
and a higher incidence of infection, their outcome vidual case reports of PEL and currently there is no
was not significantly different to HIV-negative evidence to support using high-dose chemotherapy as
patients, with a CR rate of 84% and a 2-year OS of consolidation therapy in first remission. Novel
73%. A recent retrospective analysis of 80 patients with approaches for PEL therapy outside traditional chemo-
BL treated with CODOX-M/IVAC with or without therapy have also been suggested. These include the
rituximab included 14 patients who were HIV- addition of antiviral therapy as well as inhibition of
positive, 10 of whom received rituximab. The results specific cellular targets. Antitumor activity of antiviral
demonstrated that there were fewer relapses in therapy directed against Kaposi’s sarcoma-associated
patients treated with rituximab, but only a non- herpesvirus (KSHV)/human herpesvirus 8 (HHV8)
significant trend to improved survival. Importantly, infection has been reported. However, this experience
the outcome for those with HIV infection was com- is based on single case reports. Patients with a diagnosis
parable to the HIV-negative patients. of PEL, related or not to HIV infection, experienced
In the USA, to overcome concerns of treatment prolonged complete remission with cidofovir, an anti-
toxicity in HIV-associated BL, the DA-EPOCH regi- viral agent with a broad activity against multiple DNA
men plus rituximab is favored. Although numbers viruses and one of the most effective agents to inhibit
treated are small, CR rates of 63–100% have been KSHV/HHV8 replication. The inhibition of viral repli-
reported. cation and a direct proapoptotic effect on lymphoma
cells by the high concentration of cidofovir achieved
through the intracavitary administration probably
Primary effusion lymphoma (PEL) explain this relevant clinical effect. Another approach
PEL is a rare lymphoma that has a dismal prognosis, may be to target NF-κB through the use of proteasome
with a median survival of about 6 months. Given its inhibition with drugs such as bortezomib, which induces
rarity, however, there are very few longitudinal observa- apoptosis of PEL cell lines in vitro alone or in combina-
tional series of patients with PEL, and no large prospec- tion with chemotherapy. Other approaches that have
tive trials have ever defined optimal treatment strategies. been proposed include targeting latency phase genes
Furthermore, the vast majority of the observed and such as LANA-1, using small RNA transcripts to silence
published series on HIV–NHL have all been comprised essential viral genes, and augmenting host immunity
of more common histological types, such as BL or against HHV8.
DLBCL, and it is not clear that their findings and obser-
vations apply to PEL, given its unique clinical presenta-
tion and pathogenesis. In cohorts of HIV-positive Plasmablastic lymphoma (PBL)
patients, prior to the introduction of HAART, the ther- PBL accounts for 2.6% of all HIV–NHL, suggesting an
apeutic results were unsatisfactory despite the use of increase in the past years. It is generally associated with
aggressive polychemotherapy regimens, including severe immunodeficiency and predominantly occurs
278 anthracyclines. The significant improvement of AIDS– in males. The most notable feature of PBL is its pred-
NHL prognosis observed in the HAART era also applies ilection for the oral cavity even though one-third of
Chapter 16: Lymphoma in the immunosuppressed
patients have extraoral presentation. Other patients disease. To date, the results achieved by salvage thera-
present with typical B-symptoms of lymphoma (i.e. pies that do not include a standard high-dose chemo-
fever, night sweats, and unintentional weight loss) but therapy regimen with peripheral blood stem cell
no symptoms are particularly sensitive or specific for transplant have been very frustrating (median survival
PBL. Despite the introduction of HAART, the prog- 2–4 months). With the introduction of HAART into
nosis of PBL remains very poor. Three reports have clinical practice, more aggressive treatment protocols,
noted initial regression of PBL after administration of whose effectiveness has already been documented in
HAART only; two of these patients subsequently HIV-negative lymphoma patients, can be taken into
relapsed and control of the tumor was achieved with consideration. Several studies support the feasibility of
both HAART and chemotherapy; the third patient was high-dose chemotherapy in combination with autolo-
also treated with chemotherapy with good response, gous stem cell transplant in patients with HIV–NHL,
but follow-up was only available at 1 month. The most chemosensitive recurrence, or partial remission (PR)
commonly reported lymphoma-specific chemother- after first line chemotherapy, proving that peripheral
apy is CHOP, and two retrospective case series report blood stem cell collections are adequate, anchoring
a CR rate of around 66%. Despite this encouraging rates are similar to those recorded in HIV-negative
response rate, PFS is only a few months, with a median patients, and high-dose chemotherapy is well tolerated
survival of less than 1 year. More intensive chemo- with no increase in the incidence of opportunistic
therapy regimens do not appear to improve prognosis, infections, at least in the short term. Recently, within
and a significantly worse outcome has been associated the GICAT, a study has been performed on a group of
with an ECOG performance score ≥2, advanced clin- patients with NHL or refractory or recurred Hodgkin’s
ical stage, and the presence of myc gene rearrange- disease: the results support the feasibility of an
ments, the latter being found in 40% of patients in adequate peripheral blood stem cell collection, with
one series. Since conventional chemotherapy is largely no transplant-related mortality and a very good out-
ineffective, novel agents are being investigated and come, similar to that observed in HIV-negative
drugs such as bortezomib and lenalidomide may patients.
prove to be useful. Allogeneic transplantation remains an experimen-
tal procedure in patients with HIV infection, with no
prospective trials. However, it does appear to be fea-
Primary central nervous system lymphoma sible in HIV-positive patients with hematological
Primary central nervous system lymphoma (PCNSL) malignancies. A retrospective study from the Center
occurs in the setting of severe immunosuppression, for International Blood Marrow Transplant Research
thus in the HAART era the incidence has reduced in the USA has reported the outcome for 23 HIV-
markedly. Historically, standard treatment has been positive patients who received allogeneic transplants
whole brain irradiation, but this is merely palliative. between 1987 and 2003, 10 of whom had a diagnosis of
Survival has improved with the use of HAART and the NHL. In 19 patients (83%) the donor was an HLA-
resultant immune recovery. This has encouraged the identical sibling, and 20 patients (87%) received mye-
use of CNS penetrating, systemic chemotherapy regi- loablative conditioning. The incidence of acute and
mens as in the HIV-negative setting. A small study in chronic graft-versus-host disease was 30% and 28%,
15 HIV-positive patients treated with high-dose respectively, and the 2-year OS was 30%. A signifi-
methotrexate, 3 g/m2, reported a CR rate of 47% and cantly better outcome was apparent for those patients
a median survival of 19 months. Other strategies being treated after the introduction of HAART.
explored are combinations with rituximab and
temozolamide. Hodgkin lymphoma
Hodgkin lymphoma is a non-AIDS defining malig-
nancy with an increased incidence in HIV-positive
Salvage treatment patients. Unexpectedly, some studies have reported a
Lymphoma progression is the leading cause of death further increase in incidence in the HAART era with
in 35–55% of patients with HIV–NHL receiving che- possibly a higher risk shortly after commencing
motherapy, of whom around half need second line HAART. With the introduction of HAART, there 279
chemotherapy following progression or relapse of the has been a shift in histologies, with the mixed
Chapter 16: Lymphoma in the immunosuppressed
cellularity subtype of HL occurring most commonly in infection have been reported. All patients have
the context of severe immunosuppression and the achieved CR, and nine are still alive and disease-free
nodular sclerosis subtype occurring most commonly after a median follow-up of 49 months. The results of a
in those with a higher CD4 count. The prognosis for large prospective phase II study with ABVD have been
patients with HIV-related HL has certainly improved published. Within a cooperative network in Spain, 62
in the HAART era and it is becoming the norm to treat patients with HIV–HL received the standard ABVD
patients with chemotherapy regimens used in the plus HAART. The scheduled six to eight ABVD cycles
HIV-negative setting; however, no randomized trials were completed in 82% of cases. Six patients died
but only prospective phase II studies have been per- during induction, 54 (87%) achieved a CR, and two
formed in this patient group. were resistant. The 5-year OS and event-free survival
A GICAT prospective trial carried out between probabilities were 76% and 71%, respectively. The
March 1989 and March 1992 enrolled 17 patients to immunological response to HAART had a positive
test a combined chemotherapy regimen based on epi- impact on OS (P = 0.002) and EFS (P = 0.001). The
rubicin, bleomycin, and vinblastine (EBV protocol). GICAT have also reported a prospective phase II study
Overall, the CR rate was 53%, with an 11-month aiming to evaluate feasibility and activity of a novel
median survival and a 2-year DFS rate of 55%. In an regimen including epirubicin, bleomycin, vinorelbine,
attempt to improve the above findings, another trial cyclophosphamide, and prednisone (VEBEP regi-
was performed between 1993 and 1997 on the feasi- men). Seventy percent of patients had advanced stages
bility of the EBVP (epirubicin, bleomycin, vinblastine, of disease, and 45% had an IPS >2. The CR was 67%,
and prednisone) combination. Thirty-five patients and 2-year OS, DFS, time to failure (TTF), and EFS
were admitted. The results showed a CR rate of 74%, were 69%, 86%, 59%, and 52%, respectively. Recently, a
but a recurrence was observed after 2 years in 38% of retrospective study with long follow-up has demon-
the cases. Survival and 3-year DFS rates were 32% and strated that, in patients treated with ABVD chemo-
53%, respectively. ACTG reported the findings of a therapy, the outcome is similar to that of HIV-negative
phase II study on 21 patients receiving six cycles of HL patients. In this study, which included 93 HIV-
ABVD regimen. Antiretroviral therapy was not positive patients with HL, the complete remission/
administered concomitantly but at the end of chemo- complete remission – unconfirmed (CR/CRu) rate
therapy. A CR rate of 43% was observed, with an was 77%, with a 5-year DFS and OS of 87% and 81%,
18-month median survival. respectively. Similarly to HIV-negative patients,
The GICAT has reported the final results of a [18F]-fluoro-2-deoxy-D-glucose positron emission
prospective phase II study to determine the feasibility tomography [FDG–PET] after 2–3 cycles of ABVD
of short-term (12 weeks) chemotherapy with classical chemotherapy appears to be predictive of outcome.
Stanford V regimen and adjuvant radiotherapy. In the In a retrospective study of 23 HIV-positive patients
time period from May 1997 to October 2001, 59 with advanced HL, treatment failure was not seen in
patients were treated and 69% completed the treat- those with a negative interim FDG–PET
ment plan. No dose reduction or delayed administra- scan. Table 16.1 summarizes all these studies.
tion was required. The main dose-limiting side effects In general, continuation of HAART during che-
were myelosuppression and peripheral neurotoxicity. motherapy for HL is recommended. However, care is
Overall, 89% of the patients had an objective response, required in the choice of antiviral combination as
with CR and PR rates of 81% and 8%, respectively. The interactions with chemotherapy have been reported
estimated 3-year OS, DFS, and time-to-progression to increase toxicity. To this end, ritonavir-boosted
rates are 51%, 68%, and 60%, respectively. regimens are avoided because of reports of severe
Multivariate analysis showed that the IPS was the neurotoxicity.
only significant survival-related variable (p = 0.001) It can be drawn from the above that prognosis for
with higher survival in patients with an IPS <2 com- patients with lymphomas and concomitant HIV infec-
pared to those with an IPS >2. tion could be improved by using better combined
The results of the combined BEACOPP (bleomy- chemotherapy protocols, incorporating anticancer
cin, etoposide, doxorubicin, cyclophosphamide, vin- treatments and antiretroviral drugs. The administra-
cristine, procarbazine, prednisone) regimen on a small tion of the newly available effective antiretroviral
280 group of patients with Hodgkin disease and HIV drugs (protease inhibitors) during chemotherapy
Chapter 16: Lymphoma in the immunosuppressed
together with nucleoside analogs can improve control Bibas M, Grisetti S, Alba L, Picchi G, Del Nonno F, Antinori
of the underlying HIV infection. Actually, a reduction A. Patient with HIV-associated plasmablastic lymphoma
in the viral load to indeterminable levels and an responding to bortezomib alone and in combination with
dexamethasone, gemcitabine, oxaliplatin, cytarabine, and
increased CD4+ count may lower the risk for oppor-
pegfilgrastim chemotherapy and lenalidomide alone.
tunistic infections during chemotherapy. The inclu- J Clin Oncol, 2010; 28(34):e704–708.
sion of hematopoietic growth factors in the treatment
Biggar RJ, Jaffe ES, Goedert JJ, Chaturvedi A, Pfeiffer R,
of this patient group makes it possible to increase
Engels EA. Hodgkin lymphoma and immunodeficiency
chemotherapy doses and prolong the administration in persons with HIV/AIDS. Blood, 2006;108
of antiretroviral drugs with an intent to improve (12):3786–3791.
survival. Bose P, Thompson C, Gandhi D, Ghabach B, Ozer H.
AIDS-related plasmablastic lymphoma with dramatic,
Further reading early response to bortezomib. Eur J Haematol,
2009;82(6):490–492.
An J, Sun Y, Fisher M, et al. Antitumor effects of bortezomib
(PS-341) on primary effusion lymphomas. Leukemia, Bohlius J, Schmidlin K, Boue F, et al. HIV-1-related
2004;18(10):1699–1704. Hodgkin lymphoma in the era of combination
antiretroviral therapy: incidence and evolution of
Armstrong R, Bradrick J, Liu YC. Spontaneous regression of CD4(+) T-cell lymphocytes. Blood, 2011;117
an HIV-associated plasmablastic lymphoma in the oral (23):6100–6108.
cavity: a case report. J Oral Maxillofac Surg,
2007;65:1361–1364. Boue F, Gabarre J, Gisselbrecht C, et al. Phase II trial of
CHOP plus rituximab in patients with HIV-associated
Balsalobre P, Díez-Martín JL, Re A, et al. Autologous stem- non-Hodgkin’s lymphoma. J Clin Oncol,
cell transplantation in patients with HIV-related 2006;24:4123–4128.
lymphoma. J Clin Oncol, 2009;27(13):2192–2198.
Boulanger E, Daniel MT, Agbalika F, et al. Combined
Barnes JA, Lacasce AS, Feng Y, et al. Evaluation of the chemotherapy including high-dose methotrexate in
addition of rituximab to CODOX-M/IVAC for Burkitt's KSHV/HHV8-associated primary effusion lymphoma.
lymphoma: a retrospective analysis. Ann Oncol, Am J Hematol, 2003;73(3):143–148.
2011;22(8):1859–1864.
Boulanger E, Gérard L, Gabarre J, et al. Prognostic factors
Barta SK, Lee JY, Kaplan LD, Noy A, Sparano JA. Pooled
and outcome of human herpesvirus 8-associated primary
analysis of AIDS malignancy consortium trials evaluating
effusion lymphoma in patients with AIDS. J Clin Oncol,
rituximab plus CHOP or infusional EPOCH
2005;23(19):4372–4380.
chemotherapy in HIV-associated non-Hodgkin
lymphoma. Cancer, 2011. Bower M, McCall-Peat N, Ryan N, et al. Protease inhibitors
potentiate chemotherapy-induced neutropenia. Blood,
Baumgartner JE, Rachlin JR, Beckstead JH, et al. Primary
2004;104(9): 2943–2946.
central nervous system lymphomas: natural history and
response to radiation therapy in 55 patients with acquired Caillard S, Porcher R, Provot F, et al. Post-transplantation
immunodeficiency syndrome. J Neurosurg, lymphoproliferative disorder after kidney 281
1990;73(2):206–211. transplantation: report of a nationwide French registry
Chapter 16: Lymphoma in the immunosuppressed
and the development of a new prognostic score. J Clin Engels EA, Pfeiffer RM, Fraumeni JF Jr., et al. Spectrum of
Oncol, 2013;31(10):1302–1309. cancer risk among US solid organ transplant recipients.
Castillo JJ, Echenique IA. Rituximab in combination with JAMA, 2011;306:1891–1901.
chemotherapy versus chemotherapy alone in HIV-associated Errante D, Tirelli U, Gastaldi R, et al. Combined antineoplastic
non-Hodgkin lymphoma: A pooled analysis of 15 and antiretroviral therapy for patients with Hodgkin’s
prospective studies. Am J Hematol, 2012;87(3):330–333. disease and human immunodeficiency virus infection. A
Castillo J, Pantanowitz L, Dezube BJ. HIV-associated prospective study of 17 patients. Cancer, 1994;73:437–444.
plasmablastic lymphoma: lessons learned from 112 Errante D, Gabarre J, Ridolfo AL, et al. Hodgkin’s disease in 35
published cases. Am J Hematol, 2008;83(10):804–809. patients with HIV infection: an experience with epirubicin,
Chen YB, Rahemtullah A, Hochberg E. Primary effusion bleomycin, vinblastine and prednisone chemotherapy in
lymphoma. Oncologist, 2007;12(5):569–576. combination with antiretroviral therapy and primary use of
G-CSF. Ann Oncol, 1999;10:189–195.
Cheung MC, Hicks LK, Leitch HA. Excessive neurotoxicity
with ABVD when combined with protease inhibitor- Evens AM, David KA, Helenowski I, et al. Multicenter
based antiretroviral therapy in the treatment of AIDS- analysis of 80 solid organ transplantation recipients with
related Hodgkin lymphoma. Clin Lymph, myel leuk, post-transplant lymphoproliferative disease: outcomes
2010;10(2):E22–25. and prognostic factors in the modern era. J Clin Oncol,
2010;28(6):1038–1046.
Choquet S, Levy V, Leblond V. Identification of prognostic
factors in post-transplantation lymphoproliferative Gabarre J, Marcelin AG, Azar N. High-dose therapy plus
disorders. Bull Cancer, 2004;91(4):373–378. autologous hematopoietic stem cell transplantation for
human immunodeficiency virus (HIV)-related
Choquet S, Leblond V, Herbrecht R, et al. Efficacy and safety of lymphoma: results and impact on HIV disease.
rituximab in B-cell post-transplantation Haematologica, 2004;89(9):1100–1108.
lymphoproliferative disorders: results of a prospective
Gupta V, Tomblyn M, Pedersen TL, et al. Allogeneic
multicenter phase 2 study. Blood, 2006;107(8):3053–3057.
hematopoietic cell transplantation in human
Choquet S, Trappe R, Leblond V, et al. CHOP-21 for the immunodeficiency virus-positive patients with
treatment of post-transplant lymphoproliferative hematologic disorders: a report from the center for
disorders (PTLD) following solid organ transplantation. international blood and marrow transplant research. Biol
Haematologica, 2007;92(2):273–274. Blood and Marrow transpl, 2009;15(7):864–871.
Coiffier B, Lepage E, Briere J, et al. CHOP chemotherapy Habermann TM, Weller EA, Morrison VA, et al.
plus rituximab compared with CHOP alone in elderly Rituximab-CHOP versus CHOP alone or with
patients with diffuse large-B-cell lymphoma. N Engl J maintenance rituximab in older patients with diffuse large
Med, 2002;346:235–242. B-cell lymphoma. J Clin Oncol, 2006;24(19):3121–3127.
Dharnidharka VR, Lamb KE, Gregg JA, et al. Associations Halfdanarson TR, Markovic SN, Kalokhe U, et al. A non-
between EBV serostatus and organ transplant type in chemotherapy treatment of a primary effusion
PTLD risk: an analysis of the SRTR National Registry Data lymphoma: durable remission after intracavitary
in the United States. Am J Transplant, 2012;12:976–983. cidofovir in HIV negative PEL refractory to
Diamond C, Taylor TH, Im T, Anton-Culver H. Presentation chemotherapy. Ann Oncol, 2006;17(12):1849–1850.
and outcomes of systemic non-Hodgkin’s lymphoma: a Hartmann P, Rehwald U, Salzberger B, et al. BEACOPP
comparison between patients with acquired therapeutic regimen for patients with Hodgkin’s disease
immunodeficiency syndrome (AIDS) treated with highly and HIV infection. Ann Oncol, 2003;14:1562–1569.
active antiretroviral therapy and patients without AIDS. Hentrich M, Maretta L, Chow KU, et al. Highly active
Leuk Lymphoma, 2006;47(9):1822–1829. antiretroviral therapy (HAART) improves survival in
Díez-Martín JL, Balsalobre P, Re A, et al. Comparable HIV-associated Hodgkin's disease: results of a
survival between HIV+ and HIV– non-Hodgkin and multicenter study. Ann Oncol, 2006;17(6):914–919.
Hodgkin lymphoma patients undergoing autologous Hocqueloux L, Agbalika F, Oksenhendler E, Molina JM.
peripheral blood stem cell transplantation. Blood, Long-term remission of an AIDS-related primary
2009;113(23):6011–6014. effusion lymphoma with antiviral therapy. AIDS,
Dunleavy K, Wilson WH. How I treat HIV-associated 2001;15(2):280–282.
lymphoma. Blood, 2012;119(14): 3245–3255. Hoffmann C, Tabrizian S, Wolf E, et al. Survival of AIDS
Dunleavy K, Little RF, Pittaluga S, et al. The role of tumor patients with primary central nervous system lymphoma
histogenesis, FDG-PET, and short-course EPOCH is dramatically improved by HAART-induced immune
with dose-dense rituximab (SC-EPOCH-RR) in recovery. AIDS, 2001;15(16):2119–2127.
282 HIV-associated diffuse large B-cell lymphoma. Blood, Jacomet C, Girard PM, Lebrette MG, Farese VL, Monfort L,
2010;115(15):3017–3024. Rozenbaum W. Intravenous methotrexate for primary
Chapter 16: Lymphoma in the immunosuppressed
central nervous system non-Hodgkin's lymphoma in Montoto S, Wilson J, Shaw K, et al. Excellent immunological
AIDS. AIDS, 1997;11(14):1725–1730. recovery following CODOX-M/IVAC, an effective
Jiang Y, Villeneuve PJ, Wielgosz A, et al. The incidence of cancer intensive chemotherapy for HIV-associated Burkitt's
in a population-based cohort of Canadian heart transplant lymphoma. AIDS, 2010;24(6): 851–856.
recipients. Am J Transplant, 2010;10:637–645. Montoto S, Shaw K, Okosun J, et al. HIV status does not
influence outcome in patients with classical Hodgkin
Kaplan LD, Straus DJ, Testa MA, et al. Low-dose compared
lymphoma treated with chemotherapy using
with standard-dose m-BACOD chemotherapy for non-
doxorubicin, bleomycin, vinblastine, and dacarbazine in
Hodgkin’s lymphoma associated with human
the highly active antiretroviral therapy era. J Clin Oncol,
immunodeficiency virus infection. National Institute of
2012;30(33):4111–4116.
Allergy and Infectious Diseases AIDS Clinical Trials
Group. N Engl J Med, 1997;336:1641–1648. Mounier N, Spina M, Gabarre J, et al. AIDS-related non-
Hodgkin lymphoma: final analysis of 485 patients treated
Kaplan LD, Lee JY, Ambinder RF, et al. Rituximab does not
with risk-adapted intensive chemotherapy. Blood,
improve clinical outcome in a randomized phase 3 trial of
2006;107:3832–3840.
CHOP with or without rituximab in patients with HIV-
associated non-Hodgkin lymphoma: AIDS-Malignancies Nasta SD, Carrum GM, Shahab I, et al. Regression of a
Consortium Trial 010. Blood, 2005;106:1538–1543. plasmablastic lymphoma in a patient with HIV on highly
active antiretroviral therapy. Leuk Lymphoma,
Klass CM, Krug LT, Pozharskaya VP, et al. The targeting of
2002;43:423–426.
primary effusion lymphoma cells for apoptosis by inducing
lytic replication of human herpesvirus 8 while blocking Navarro JT, Lloveras N, Ribera JM, Oriol A, Mate JL. Feliu E
virus production. Blood, 2005;105(10):4028–4034. (2005). The prognosis of HIV-infected patients with
diffuse large B-cell lymphoma treated with
Krishnan AY, Molina A, Zaia J, et al. Durable remissions with chemotheraphy and highly active antiretroviral therapy is
autologous stem cell transplantation for high-risk HIV- similar to that of HIV-negative patients receiving
associated lymphomas. Blood, 2005;105(2):874–878. chemotherapy. Haematologica, 2005;90(5): 704–706.
LaCasce AS. Post-transplant lymphoproliferative disorders. Okosun J, Warbey V, Shaw K, et al. Interim fluoro-2-deoxy-
Oncologist, 2006;11:674–680. D-glucose-PET predicts response and progression-free
Lanoy E, Rosenberg PS, Fily F, et al. HIV-associated Hodgkin survival in patients with Hodgkin lymphoma and HIV
lymphoma during the first months on combination infection. AIDS, 2012;26(7):861–865.
antiretroviral therapy. Blood, 2011;118(1):44–49. Opelz G, Daniel V, Naujokat C, Dohler B. Epidemiology of
Lester R, Li C, Phillips P, et al. Improved outcome of human pretransplant EBV and CMV serostatus in relation to
immunodeficiency virus-associated plasmablastic posttransplant non-Hodgkin lymphoma.
lymphoma of the oral cavity in the era of highly active Transplantation, 2009;88:962–967.
antiretroviral therapy: a report of two cases. Leuk Oriol A, Ribera JM, Esteve J, et al. Lack of influence of
Lymphoma, 2004;45:1881–1885. human immunodeficiency virus infection status in the
Levine AM, Li P, Cheung T, et al. Chemotherapy consisting response to therapy and survival of adult patients with
of doxorubicin, bleomycin, vinblastine and dacarbazine mature B-cell lymphoma or leukemia. Results of the
with granulocyte-colony-stimulating factor CSF in HIV PETHEMA-LAL3/97 study. Haematologica,
infected patients with newly diagnosed Hodgkin’s 2003;88(4):445–453.
disease: a prospective, multi-institutional AIDS Clinical Oriol A, Ribera JM, Bergua J, et al. High-dose chemotherapy
Trials Group study (ACTG 149). J Acquir Immune Defic and immunotherapy in adult Burkitt lymphoma:
Syndr, 2000;24:444–450. comparison of results in human immunodeficiency
Lim ST, Karim R, Nathwani BN, et al. AIDS-related Burkitt’s virus-infected and noninfected patients. Cancer,
lymphoma versus diffuse large-cell lymphoma in the 2008;13(1):117–125.
pre-highly active antiretroviral therapy (HAART) and Panos G, Karveli EA, Nikolatou O, et al. Prolonged survival
HAART eras: significant differences in survival with of an HIV-infected patient with plasmablastic lymphoma
standard chemotherapy. J Clin Oncol, 2005;23:4430–4438. of the oral cavity. Am J Hematol, 2007;82:761–765.
Little RF, Pittaluga S, Grant N, et al. (2003). Highly effective Parker A, Bowles K, Bradley JA, et al. Management of post-
treatment of acquired immunodeficiency syndrome- transplant lymphoproliferative disorder in adult solid
related lymphoma with dose-adjusted EPOCH: impact of organ transplant recipients–BCSH and BTS guidelines.
antiretroviral therapy suspension and tumor biology. Br J Haematol, 2010;149(5):693–705.
Blood, 2003;101(12):4653–4659. Pfreundschuh M, Trumper L, Osterborg A, et al. CHOP-like
Luppi M, Trovato R, Barozzi P, et al. Treatment of herpesvirus chemotherapy plus rituximab versus CHOP-like
associated primary effusion lymphoma with intracavity chemotherapy alone in young patients with good- 283
cidofovir. Leukemia, 2005;19(3):473–476. prognosis diffuse large-B-cell lymphoma: a randomised
Chapter 16: Lymphoma in the immunosuppressed
controlled trial by the MabThera International Trial Sparano JA. Clinical aspects and management of AIDS-
(MInT) Group. Lancet Oncol, 2006;7(5):379–391. related lymphoma. Eur J Cancer, 2001;37:1296–1305.
Polesel J, Clifford GM, Rickenbach M, et al. Non-Hodgkin Sparano JA, Lee JY, Kaplan LD, et al. Rituximab plus
lymphoma incidence in the Swiss HIV Cohort Study concurrent infusional EPOCH chemotherapy is highly
before and after highly active antiretroviral therapy. effective in HIV-associated B-cell non-Hodgkin’s
AIDS, 2008;22(2):301–306. lymphoma. Blood, 2010;115(15):3008–3016.
Quinlan SC, Pfeiffer RM, Morton LM, Engels EA. Risk Spina M, Gaidano G, Carbone A, Capello D, Tirelli U.
factors for early-onset and late-onset post-transplant Highly active antiretroviral therapy in human
lymphoproliferative disorder in kidney recipients in the herpesvirus-8-related body-cavity-based lymphoma.
United States. Am J Hematol, 2011;86:206–209. AIDS, 1998;12 (8):955–956.
Raez LE, Patel P, Feun L, Restrepo A, Raub WA, Jr., Cassileth Spina M, Vaccher E, Juzbasic S, et al. Human
PA. Natural history and prognostic factors for survival in immunodeficiency virus-related non-Hodgkin
patients with acquired immune deficiency syndrome lymphoma: activity of infusional cyclophosphamide,
(AIDS)-related primary central nervous system doxorubicin and etoposide as second-line chemotherapy
lymphoma (PCNSL). Crit Rev oncogen, in 40 patients. Cancer, 2001;1(92):200–206.
1998;9(3–4):199–208. Spina M, Gabarre J, Rossi G, et al. Stanford V and
Ratner L, Lee J, Tang S, et al. Chemotherapy for human concomitant HAART in 59 patients with Hodgkin’s
immunodeficiency virus-associated non-Hodgkin's disease and HIV infection. Blood, 2002;100:1984–1988.
lymphoma in combination with highly active Spina M, Carbone A, Vaccher E, et al. Outcome in patients
antiretroviral therapy. J Clin Oncol, with non-Hodgkin lymphoma and with or without
2001;19(8):2171–2178. human immunodeficiency virus infection. Clin Infect Dis,
Re A, Cattaneo C, Michieli M, et al. High-dose therapy and 2004;38(1):142–144.
autologous peripheral-blood stem-cell transplantation as Spina M, Jaeger U, Sparano JA, et al. Rituximab plus
salvage treatment for HIV-associated lymphoma in infusional cyclophosphamide, doxorubicin, and
patients receiving highly active antiretroviral therapy. etoposide (R-CDE) in HIV-associated non-Hodgkin’s
J Clin Oncol, 2003;21(23):4423–4427. lymphoma: pooled results from three phase II trials.
Re A, Michieli M, Casari S, et al. High-dose therapy and Blood, 2005;105(5):1891–1897.
autologous peripheral blood stem cell transplantation as Spina M, Simonelli C, Talamini R, Tirelli U. Patients with
salvage treatment for AIDS-related lymphoma: long-term HIV with Burkitt’s lymphoma have a worse outcome
results of the Italian Cooperative Group on AIDS and than those with diffuse large-cell lymphoma also in the
Tumors (GICAT) study with analysis of prognostic highly active antiretroviral therapy era. J Clin Oncol,
factors. Blood, 2009;114(7):1306–1313. 2005;23(31):8132–8133.
Ribera JM, Oriol A, Morgades M, et al. Safety and efficacy Spina M, Rossi G, Antinori A, et al. VEBEP regimen and
of cyclophosphamide, adriamycin, vincristine, highly active antiretroviral therapy (HAART) in patients
prednisone and rituximab in patients with human (pts) with HD and HIV infection (HD-HIV). Ann Oncol,
immunodeficiency virus-associated diffuse large B-cell 2008;19:iv152, abs 227.
lymphoma: results of a phase II trial. Br J Haematol,
2008;140(4):411–419. Taylor AL, Bowles KM, Callaghan CJ, et al. Anthracycline-
based chemotherapy as first-line treatment in adults with
Riedel DJ, Gonzalez-Cuyar F, Zhao XF, et al. Plasmablastic malignant post-transplant lymphoproliferative disorder
lymphoma of the oral cavity: a rapidly progressive after solid organ transplantation. Transplantation,
lymphoma associated with HIV infection. Lancet Infect 2006;82(3):375–381.
Dis, 2008;8:261–267.
Tirelli U, Spina M, Vaccher E, et al. Clinical evaluation of 451
Schichman SA, McClure R, Schaefer RF, Mehta P. HIV and patients with HIV related non-Hodgkin’s lymphoma:
plasmablastic lymphoma manifesting in sinus, testicles, experience of the Italian Cooperative Group on AIDS and
and bones: a further expansion of the disease spectrum. Tumors (GICAT). Leukemia Lymphoma, 1995;20:91–96.
Am J Hematol, 2004;77:291–295.
Trappe R, Oertel S, Leblond V, et al. Sequential treatment
Siddiqi T, Joyce RM. A case of HIV-negative primary with rituximab followed by CHOP chemotherapy in
effusion lymphoma treated with bortezomib, pegylated adult B-cell post-transplant lymphoproliferative disorder
liposomal doxorubicin, and rituximab. Clin Lymphoma (PTLD): the prospective international multicentre phase
Myeloma, 2008;8(5):300–304. 2 PTLD-1 trial. Lancet Oncol, 2012;13(2):196–206.
Simonelli C, Spina M, Cinelli R, et al. Clinical features and Tsai De, Hardy CL, Tomaszewski JE, et al. Reduction in
outcome of primary effusion lymphoma in HIV-infected immunosuppression as initial therapy for posttransplant
284 patients: a single-institution study. J Clin Oncol, lymphoproliferative disorder: analysis of prognostic
2003;21(21):3948–3954.
Chapter 16: Lymphoma in the immunosuppressed
variables and long-term follow-up of 42 adult patients. immunodeficiency virus-associated Burkitt lymphoma.
Transplantation, 2001;71(8):1076–1088. Cancer, 2003;98:1196–1205.
Vaccher E, Spina M, Talamini R, et al. Improvement of Weiss R, Mitrou P, Arasteh K, et al. Acquired
systemic human immunodeficiency virus-related immunodeficiency syndrome-related lymphoma:
non-Hodgkin’s lymphoma outcome in the era of highly simultaneous treatment with combined
active antiretroviral therapy. Clin Infect Dis, cyclophosphamide, doxorubicin, vincristine, and
2003;37:1556–1564. prednisone chemotherapy and highly active antiretroviral
Vaccher E, Tirelli U, Spina M, et al. Age and serum lactate therapy is safe and improves survival–results of the
dehydrogenase level are independent prognostic factors German Multicenter Trial. Cancer,
in HIV-related non-Hodgkin’s lymphoma: a single 2006;106(7):1560–1568.
institution study of 96 patients. J Clin Oncol, Won JH, Han SH, Bae SB, et al. Successful eradication
1996;14:2217–2233. of relapsed primary effusion lymphoma with
Waddington TW, Aboulafia DM. Failure to eradicate AIDS- high-dose chemotherapy and autologous stem cell
associated primary effusion lymphoma with high-dose transplantation in a patient seronegative for human
chemotherapy and autologous stem cell reinfusion: case immunodeficiency virus. Int J Hematol, 2006;83
report and literature review. AIDS Patient Care STDS, (4):328–330.
2004;18(2):67–73. Xicoy B, Ribera JM, Miralles P, et al. Results of treatment with
Wang ES, Straus DJ, Teruya-Feldstein J, et al. Intensive doxorubicin, bleomycin, vinblastine and dacarbazine and
chemotherapy with cyclophosphamide, doxorubicin, highly active antiretroviral therapy in advanced stage
high-dose methotrexate/ifosfamide, etoposide, and high- human immunodeficiency virus-related Hodgkin’s
dose cytarabine (CODOX-M/IVAC) for human lymphoma. Haematologica, 2007;92:191–198.
285
Chapter
17
Atypical lymphoproliferative, histiocytic,
and dendritic cell disorders
Matthew S. Davids and Jennifer R. Brown
286 Lymphoma, Second Edition, ed. Robert Marcus, John W. Sweetenham, and Michael E. Williams. Published by
Cambridge University Press. © Cambridge University Press 2014.
Chapter 17: Atypical disorders
Table 17.1 Classification of rare lymphoproliferative presentation. In cases where symptoms are more
disorders. severe, a trial of glucocorticoids may be effective, and
anecdotal success has been reported with hydroxy-
Benign lymphoproliferative disorders
chloroquine. In general, patients should be followed
* Kikuchi’s disease
*
for several years after the resolution of symptoms,
Progressive transformation of germinal centers
* Vascular transformation of sinuses given the risk of late recurrence and the increased
* Inflammatory pseudotumor of lymph nodes likelihood of the development of SLE. Patients with
Histiocytic disorders recurrent disease often are symptomatic for longer
*
than their initial episode, though there is no evidence
Rosai–Dorfman disease
* Histiocytic sarcoma that early intervention can prevent recurrence or
* Langerhans cell histiocytosis decrease its severity.
* Interdigitating dendritic cell sarcoma
* Follicular dendritic cell sarcoma
* Erdheim–Chester disease Progressive transformation
* Hemophagocytic lymphohistiocytosis of germinal centers
Atypical lymphoproliferative disorders Progressive transformation of germinal centers
* Polyclonal (PTGC) occurs primarily in young, asymptomatic
* Unicentric Castleman’s disease adults, who develop otherwise unexplained lymph-
* Multicentric Castleman’s disease
* Malignant
adenopathy. There is approximately a 3:1 male pre-
* Lymphomatoid papulosis dominance. Cervical lymphadenopathy is the most
* Lymphomatoid granulomatosis common presentation, although extranodal presenta-
tions of PTGC in oral and gastrointestinal mucosa,
most frequently in the colon, have also been reported.
In contrast to nodal PTGC, extranodal cases have a
may be present in more advanced cases. Most com-
female predominance and tend to occur at an older
monly, these include a transient skin rash similar to a
age. Both nodal and extranodal PTGC can occur at
drug eruption or viral exanthem, but a variety of other
multiple sites. Enlarged lymph nodes may wax and
skin lesions have been described.
wane over a period of months or years, and sometimes
Excisional biopsy of an involved lymph node is the
regress spontaneously. No specific treatment is
preferred diagnostic procedure for Kikuchi’s disease, as
generally required. Of note, PTGC can be associated
an evaluation of the tissue architecture is crucial to dis-
with both classical Hodgkin lymphoma and nodular
tinguish it from malignant lymphoma. Inflammatory
lymphocyte-predominant Hodgkin lymphoma
markers, such as the erythrocyte sedimentation rate,
(NLPHD). Small prospective studies have shown that
are usually elevated, and mild leukopenia and anemia
about 2.5% of patients with PTGC will later develop
are also commonly seen. Although serologic testing with
NLPHD, but causal factors underlying this association
antinuclear antibodies (ANA) or rheumatoid factor (RF)
have not been identified. In rare cases, florid presenta-
is usually negative, an ANA should be checked at base-
tions of PTGC have been reported, and in such sit-
line because patients with Kikuchi’s disease are at
uations they can be mistaken for lymphoma. It is
increased risk for the development of systemic lupus
important to distinguish these cases pathologically
erythematosus (SLE), with at least 35 such cases reported
from lymphoma, as they will often be self-limited
in the literature. No clear predictive markers for the
and may not require treatment. These more aggressive
development of SLE in Kikuchi’s patients have been
cases tend to occur in young men, and present most
identified.
often as a rapidly enlarging, solitary lymph node.
Prognosis/treatment
Kikuchi’s disease is generally self-limited, with most Vascular transformation of sinuses
patients having resolution of their symptoms by 3 Although vascular transformation of the sinuses
months. In a series of 102 patients followed in Korea (VTS) can present as lymphadenopathy, more com-
from 2001 to 2006, eight patients relapsed early and monly it is an incidental finding noted on pathologic
13 relapsed late. When patients did relapse, their examination of lymph nodes. The condition usually 287
symptoms tended to last longer than at initial arises in abdominal lymph nodes, but cervical lymph
Chapter 17: Atypical disorders
node presentations have also been reported. Since VTS often the key to confirming these diagnoses. For exam-
may be caused by lymph node obstruction, the diag- ple, expression of CD1a is helpful in confirming the
nosis should prompt an evaluation for causes of nodal diagnosis of Langerhans cell histiocytosis. Clarifying
obstruction. A variety of factors can contribute to the diagnosis often may require multiple biopsies, but
lymph node obstruction, including vascular thrombo- is essential, as approaches to management vary widely.
sis, heart failure, previous surgery, or radiotherapy.
Occult malignancy leading to obstruction is a theoret- Rosai–Dorfman disease
ical concern, but little has been published describing
this phenomenon. A case of a plexiform variant of Presentation/diagnosis
VTS associated with myelodysplastic syndrome was Rosai–Dorfman disease (RDD), also known as sinus
reported, but there are not enough data to conclude histiocytosis with massive lymphadenopathy, is a rare
whether a causal relationship exists between these two disorder characterized by histiocytic proliferation in
entities. lymph node sinuses. The disease is more common in
males and in those of African descent. RDD has been
Inflammatory pseudotumor described following allogeneic stem cell transplanta-
tion for hematologic malignancies and concurrently
of lymph nodes or after treatment for both Hodgkin and non-Hodgkin
Inflammatory pseudotumor of lymph nodes (IPLN), lymphoma (NHL).
also known as plasma cell granuloma, is a benign Patients generally present in early adulthood with
condition that usually occurs in young adults who massive painless lymphadenopathy, predominantly in
present with unexplained lymphadenopathy. Patients the cervical area. Although the majority of cases occur
with a single site of disease may be asymptomatic, but within lymph nodes, extranodal presentations are
those with multiple sites often also have systemic common, occurring in about 40% of cases, usually in
symptoms and elevated inflammatory markers in the the cervical region. Isolated RDD in the CNS has been
blood. Occasionally, IPLN will be discovered as part of reported, generally in an extra-axial or dura-based
a fever of unknown etiology work-up. As the disease location. These patients can present with headache,
develops, marked enlargement of both peripheral and seizure, or other neurologic manifestations. CNS pre-
central lymph nodes can occur, and it is not uncom- sentations occur more often in older patients and can
mon for nodes to enlarge to >3 cm. These large nodes be mistaken radiographically for meningioma.
are usually self-contained, but focal infiltration of the Extranodal presentations have also been described in
liver and spleen by nodes growing in, as well as paren- a variety of other locations; for example, in the thyroid,
chymal involvement, have also been described. kidney, and bone, the latter of which can manifest as
IPLN is thought to be a consequence of a resolving lytic lesions. A cutaneous form of the disease, with
infection. Although the inciting pathogens are not well isolated papular or firm nodular lesions, has also
defined, an association with resolved toxoplasmosis been described. Whether cutaneous RDD is truly a
infection has been proposed based on a similar patho- distinct entity is not certain. Systemic symptoms,
logic appearance. Treatment is usually not required, such as fever or night sweats, may be present, as may
although surgery has been used either with curative inflammatory markers, such as the erythrocyte sedi-
intent in single site disease, or for palliation of symp- mentation rate and polyclonal elevation in immuno-
toms in multifocal disease. globulin levels.
patients suggested that there may be a role for this cavity, skin, soft tissue, liver, and spleen. The most
combination. There has been growing interest in common presentation is that of a painless mass that
using targeted therapy with imatinib to inhibit differ- slowly enlarges, but can grow to large sizes (median size
entiation of dendritic cells from peripheral blood pro- 5 cm). Systemic symptoms are uncommon, but there
genitor cells. A recent report also described recurrent are reports of a possible association with pemphigus.
BRAF mutations in LCH, suggesting that a subset of The disease course is generally indolent, and most
patients may respond to RAF pathway inhibitors. patients can be successfully treated with complete sur-
gical excision. Local recurrence occurs in more than
50% of cases, and may be delayed for several years.
Interdigitating dendritic cell sarcoma Larger tumors (>6 cm), high-grade histology, or an
Only single case reports or very small series have been intra-abdominal location are all poor prognostic fac-
published on the extremely rare interdigitating den- tors. Adjuvant radiotherapy and/or chemotherapy
dritic cell sarcoma (IDC). In the largest series of four have been utilized in selected cases, but are generally
cases, the disease occurred predominantly in older reserved for recurrent disease. More widespread dis-
adults, and may be associated with NHL. The median ease may eventually develop in up to 25% of patients,
age at diagnosis of reported cases is 51, with a range of and at least 10% of patients ultimately die of the
2–86 years. Most commonly, patients are asympto- disease.
matic, and present with a single enlarged lymph
node. Extranodal presentations may occur in the
lung, bone marrow, bladder, testis, oral cavity, or Erdheim–Chester disease
breast. More extensive disease is frequently accompa- Erdheim–Chester disease (ECD) is a rare disorder
nied by systemic symptoms, such as weight loss, night similar in its distribution of disease sites to LCH but
sweats, fever, or fatigue. Hepatosplenomegaly or more pathologically distinct, in that ECD is typically S-100
diffuse lymphadenopathy are unusual. negative, and lacks Birbeck granules. In contrast to
The disease tends to behave aggressively, with pro- LCH, ECD tends to present at an older age, with a
gressive growth of lesions that can then invade into median age at diagnosis of 53 years. ECD also involves
organs, eventually affecting the liver, lungs, or kidneys. symmetric, sclerotic lesions in the long bones, as
About half of patients eventually die of the disease, and opposed to the lytic, axial lesions in LCH. The most
effective treatment options have been limited. For local- common symptom at presentation in ECD is bone
ized disease, radiation has been reported to be useful. pain, with systemic symptoms also frequently present
With extensive disease, attempts have been made to at diagnosis. Cough and dyspnea are frequent, as 20–
utilize NHL-like therapy, with limited success. The 30% of patients have lung involvement, often seen as
Hodgkin chemotherapy regimen, doxorubicin, vinblas- interstitial infiltrates or pleural disease on X-ray. X-
tine, bleomycin, dacarbazine (ABVD), was also utilized rays may also reveal sclerosis of the diaphyses and
successfully in a patient with extensive IDC. metaphyses, which typically spares the epiphyses.
Radiologic circumferential sheathing, or coating, of
the aorta, is considered to be pathognomonic of
Follicular dendritic cell sarcoma ECD. Extraosseous disease can include pituitary
Follicular dendritic cell (FDC) sarcoma is a rare neo- involvement, leading to DI, cerebral infiltration, pul-
plasm of unknown etiology. The hyaline-vascular type monary fibrosis, pericardial infiltration, renal infiltra-
of Castleman’s disease may be associated with FDC, tion, retroperitoneal involvement, and cutaneous
usually developing either prior to or simultaneous xanthomas. In general, the greater the burden of extra-
with FDC. The disease presents at a median age at osseous disease the more aggressive the clinical course.
diagnosis of 44 years. FDC is confined to the lymph Most patients die within 3 years of diagnosis, often
nodes in up to two-thirds of cases, most often in the due to progressive renal dysfunction, pulmonary fib-
cervical region. Intra-abdominal lymph nodes may be rosis, and/or congestive heart failure. Various combi-
larger, as they often present later in the course of dis- nations of corticosteroids, chemotherapy, and
ease. When FDC sarcoma spreads, the lungs and liver radiation have been attempted, but none have shown
are the most common sites involved. Extranodal pre- a consistent effect in reducing the mortality of the
sentations have also been described in up to one-third disease. Improvement of clinical manifestations, 291
of patients, including in the gastrointestinal tract, oral including regression of bony lesions, has been
Chapter 17: Atypical disorders
reported with interferon alfa-2a, which is now used splenomegaly, and hypertriglyceridemia or hypofibri-
commonly as a first line treatment for ECD. As under- nogenemia. Minor criteria include serum ferritin con-
standing of the molecular underpinnings of this dis- centration >500 ug/L, and specialized testing, such as
order improves, there is hope that more targeted the soluble CD25 (sIL-2 receptor) >2400 U/mL, and low
agents will be even more effective. A recent report or absent NK-cell activity. If the clinical suspicion is
found PDGFRB expression in 32/37 patients (86.5%) high, but hemophagocytosis cannot be demonstrated
and utilized imatinib to achieve stable disease in 2/6 pathologically, a repeat biopsy should be performed, as
(33%) patients with advanced, treatment-refractory this finding is frequently missed on marrow
disease. Another group recently reported using imati- examination. Notably, serum ferritin concentrations
nib as frontline therapy in one patient with ECD, and may be highly elevated (in the tens of thousands
was able to achieve both symptomatic and radio- range) and values in this range should prompt consid-
graphic improvement. eration of HLH in the differential diagnosis.
aggressive, rapidly progressive form of the disease, free survival of nearly 80%, though reactivation of KS
which can lead to death within weeks of diagnosis. was observed in several patients. There has been inter-
This more aggressive variant appears to be more com- est in blocking the cytokine-mediated symptoms in
mon in HIV-positive patients. In between these two MCD that are primarily related to elevated IL-6 levels.
extremes, the disease exhibits a large variety of behav- Neutralizing IL-6 monoclonal antibodies have been
ior. For example, many patients have disease that studied, and appear to be efficacious in providing
behaves in an aggressive manner for a limited period symptom relief, but the results are usually transient,
of time, only to have spontaneous regression for a with a return of symptoms upon discontinuation of
period of time, and then subsequent relapse. therapy. Antiviral agents, such as ganciclovir, are able
When viewed in the aggregate, however, MCD to reduce HHV8 viral load, but have shown only
carries a poor prognosis. HIV-positive patients with modest clinical benefit in the small number of patients
MCD fare the worst, with a median survival time studied to date.
of only 8–14 months, and an overall mortality rate of
70–85%. HIV-negative MCD patients do slightly bet-
ter, but still only have median survival times of 26–30
Malignant disorders
months. The most common causes of death are the
development of overwhelming infection or of second
Lymphomatoid papulosis
malignancies, such as KS and/or NHL. In particular, Lymphomatoid papulosis (LP) is a histologically
growing evidence suggests that MCD patients are at malignant cutaneous T-cell lymphoproliferative dis-
risk for developing HHV8-associated, plasmablastic order that clinically usually behaves indolently. The
microlymphomas within affected nodes, which are disease typically presents with multiple, small
difficult to treat and confer a poor prognosis. (<3 cm) skin papules that can wax and wane in size
A wide variety of treatment options have been tried over a period of several months to years. The disease
in MCD, but no clear standard of care has been tends to be a chronic, self-healing condition, but does
defined. Unlike in UCD, surgery generally is not fea- carry a substantial risk of transformation into frank
sible or effective in MCD, because of the multifocal lymphoma, in particular cutaneous T-cell lym-
nature of the disease. In HIV-negative MCD patients, phoma, Hodgkin lymphoma, or anaplastic large cell
steroids may help alleviate systemic symptoms, and in lymphoma. Estimates of the overall risk for progres-
some cases induce a transient remission. The response sion to lymphoma vary widely in the literature, from
rate is of the order of 65%, with about 15% achieving a 20% to 80%. It does appear that the risk of progres-
complete response. HIV-positive patients generally sion is cumulative, as summarized in a review of
benefit more from chemotherapy-based approaches. four major studies which each showed substantial
Both etoposide and vinblastine have been studied as cumulative lymphoma risk over time. Some have
single agents, with most patients at least achieving a proposed that, because of this substantial risk of
partial response and relief of symptoms; however, the transformation, early treatment with agents such as
disease tends to relapse upon discontinuation of ther- methotrexate or interferon-alpha may slow or
apy, leading some to advocate for intermittent main- even prevent this progression. However, it should
tenance therapy. Patients with more aggressive MCD, be noted that, to date, no currently available treat-
regardless of HIV status, should be considered for ment has been convincingly shown to lower the
combination chemotherapy approaches. NHL-type risk of transformation. Furthermore, the survival of
regimens such as CHOP have demonstrated about a patients who develop lymphoma after LP has not
50% durable complete response rate. been systematically studied. Until more data are
Recently, there has been promising activity of the available, treatment of LP should therefore generally
anti-CD20 monoclonal antibody rituximab in MCD. be reserved for symptomatic patients.
In HIV-positive patients recently treated with single
agent chemotherapy, four weekly infusions of rituxi- Lymphomatoid granulomatosis
mab were able to maintain a sustained remission in
over 70% of patients at 1 year. Another study utilized Presentation/diagnosis
weekly rituximab for 4 weeks, and achieved a partial Lymphomatoid granulomatosis (LYG) was first
294 response in two-thirds of patients, with 2-year relapse- described in the early 1970s, primarily in patients with
Chapter 17: Atypical disorders
multiple nodular pulmonary lesions that pathologically of 14 months. Patients most commonly succumb to
resembled a pulmonary angitis similar to Wegener’s pulmonary complications, but development of CNS
granulomatosis. Most patients with LYG are middle- disease, infection, or lymphoma can also be potentially
aged, with a median age at diagnosis of 48 years. A 2:1 fatal. Treatment strategies are guided by the histopa-
male predominance has been described, and the disease thologic grade, severity of symptoms, and extent of
is more common in western countries. Patients most extrapulmonary disease. Asymptomatic patients with
commonly present with respiratory symptoms, includ- low-grade (grade 1 or 2) disease in the lungs only can
ing cough and shortness of breath, and fever is also be followed closely with observation, including serial
often present at initial diagnosis. Other systemic symp- imaging. Many of these patients will spontaneously
toms, such as malaise and weight loss, are present in remit. Those who develop progressive disease, those
about one-third of patients. On chest imaging, a bilat- with grade 3 disease, or those with CNS involvement
eral, nodular pattern of infiltrates is often observed, require more aggressive systemic therapy. A small
classically with small lesions (1 cm or less) in the mid prospective study of 15 patients suggested that combi-
to lower lungs fields. Extrapulmonary presentations are nation cyclophosphamide and steroids could lead to
less common, with the skin (25–50%), kidney (32%), prolonged complete remissions in the majority of
and neurologic system (20–30%) involved most com- patients. Low-risk patients have also been treated
monly. Skin lesions can vary from an erythematous with some success with the antiproliferative cytokine
rash to subcutaneous nodules or ulceration. Lymph interferon alfa-2b. A recent NIH study of 31 grade I/II
node involvement is less common at diagnosis, but LYG patients treated with interferon alfa-2b showed a
may develop as the disease progresses. 61% complete response (CR) rate, with a 90% CR rate
The lung is almost always involved at some point in patients with CNS disease. They also reported a
in the disease even if the primary site of presentation progression-free survival of 56%, with a median
is extrapulmonary. The neoplastic cell component is follow-up time of 5 years. High-risk (grade 3) patients
mainly angiocentric and angioinvasive but there is may benefit from combination chemotherapy regi-
often a florid background population of reactive mens such as CHOP in combination with rituximab,
small lymphocytes, immunoblastic cells, plasma which was highly efficacious in a recent case report.
cells, and histiocytes. Lesions are graded according Autologous stem cell transplantation may also play a
to the mix of neoplastic cellular appearances. Grade 1 role in the treatment of LYG, although more data are
lesions contain a lymphoid infiltrate with absent large needed.
cells and little pleomorphism, grade 2 lesions have
scattered large cells, while grade 3 lesions have more
abundant large cells that are easily apparent and may
Further reading
include cells with morphology reminiscent of Reed– Andriko JA, Morrison A, Colegial CH, Davis BJ, Jones RV.
Rosai–Dorfman disease isolated to the central nervous
Sternberg cells. system: a report of 11 cases. Mod Pathol,
Immunophenotypically the neoplastic cells of this 2001;14(3):172–178.
disorder are CD20-positive B-cells. CD79a may be
Appel S, Boehmler AM, Grunebach F, et al. Imatinib
absent in some cases. Staining for CD30 is variable. mesylate affects the development and function of
There is always a population of Epstein–Barr virus dendritic cells generated from CD34+ peripheral blood
(EBV)-positive cells that usually express LMP-1, and progenitor cells. Blood, 2004;103(2):538–544.
the proportion of these cells correlates with the histo- Badalian-Very G, Vergilio JA, Degar BA, et al. Recurrent
logical grading, being scanty in grade 1 lesions and BRAF mutations in Langerhans cell histiocytosis. Blood,
numerous in grade 3 lesions. The background popu- 16(11):1919–1923.
lation is mainly T-cells, most of which express CD4. Beck JT, Hsu SM, Wijdenes J, et al. Brief report: alleviation of
systemic manifestations of Castleman’s disease by
monoclonal anti-interleukin-6 antibody. N Engl J Med,
Prognosis/treatment 1994;330(9):602–605.
The prognosis for LYG patients is quite variable; while Bower M, Powles T, Williams S, et al. Brief communication:
spontaneous resolution occurs in about 20% of rituximab in HIV-associated multicentric Castleman
patients, the majority of patients develop progressive disease. Ann Intern Med, 2007;147(12):836–839.
disease. In a large study of 152 patients, the overall Cabanillas F, Armitage J, Pugh WC, Weisenburger D, Duvic 295
mortality rate due to LYG was 67%, with a median OS M. Lymphomatoid papulosis: a T-cell dyscrasia with a
Chapter 17: Atypical disorders
propensity to transform into malignant lymphoma. Ann Gadner H, Grois N, Arico M, et al. A randomized trial of
Intern Med, 1995;122(3):210–217. treatment for multisystem Langerhans’ cell histiocytosis.
Casper C, Nichols WG, Huang ML, Corey L, Wald A. J Pediatr, 2001;138(5):728–734.
Remission of HHV-8 and HIV-associated multicentric Gadner H, Grois N, Potschger U, et al. Improved outcome
Castleman disease with ganciclovir treatment. Blood, in multisystem Langerhans cell histiocytosis is
2004;103(5):1632–1634. associated with therapy intensification. Blood,
Castleman B, Iverson L, Menendez VP. Localized 2008;111(5):2556–2562.
mediastinal lymph node hyperplasia resembling Gerard L, Berezne A, Galicier L, et al. Prospective study of
thymoma. Cancer, 1956;9(4):822–830. rituximab in chemotherapy-dependent human
Chadburn A, Cesarman E, Nador RG, Liu YF, Knowles DM. immunodeficiency virus associated multicentric
Kaposi’s sarcoma-associated herpesvirus sequences in Castleman’s disease: ANRS 117 Castleman Trial. J Clin
benign lymphoid proliferations not associated with Oncol, 2007;25(22):3350–3356.
human immunodeficiency virus. Cancer, Gergis U, Dax H, Ritchie E, et al. Autologous hematopoietic
1997;80(4):788–797. stem-cell transplantation in combination with thalidomide
Chang Y, Cesarman E, Pessin MS, et al. Identification as treatment for histiocytic sarcoma: a case report and
of herpesvirus-like DNA sequences in AIDS- review of the literature. J Clin Oncol, 2011 Jan 10.
associated Kaposi’s sarcoma. Science, 1994;266 Gessain A, Briere J, Angelin-Duclos C, et al. Human herpes
(5192):1865–1869. virus 8 (Kaposi’s sarcoma herpes virus) and malignant
Davis RE, Warnke RA, Dorfman RF. Inflammatory lymphoproliferations in France: a molecular study of 250
pseudotumor of lymph nodes. Additional observations cases including two AIDS-associated body cavity based
and evidence for an inflammatory etiology. Am J Surg lymphomas. Leukemia, 1997;11(2):266–272.
Pathol, 1991;15(8):744–756. Gruber R, Sepp NT, Fritsch PO, Schmuth M. Prognosis of
Derenzini E, Fina MP, Stefoni V, et al. MACOP-B regimen lymphomatoid papulosis. Oncologist,
in the treatment of adult Langerhans cell histiocytosis: 2006;11(8):955–957; author reply 7.
experience on seven patients. Ann Oncol, Haroche J, Amoura Z, Trad SG, et al. Variability in the
2010;21(6):1173–1178. efficacy of interferon-alpha in Erdheim–Chester disease
Dunleavy K, Chattopadhyay P, Kawada J, et al. Immune by patient and site of involvement: results in eight
Characteristics Associated with Lymphomatoid patients. Arthritis Rheum, 2006;54(10):3330–3336.
Granulomatosis and Outcome Following Treatment with Haroche J, Amoura Z, Charlotte F, et al. Imatinib
Interferon-Alpha. ASH Annual Meeting; 2010. mesylate for platelet-derived growth factor receptor-
Fauci AS, Haynes BF, Costa J, Katz P, Wolff SM. beta-positive Erdheim–Chester histiocytosis. Blood,
Lymphomatoid granulomatosis. Prospective clinical and 2008;111(11):5413–5415.
therapeutic experience over 10 years. N Engl J Med, Hehn ST, Grogan TM, Miller TP. Utility of fine-needle
1982;306(2):68–74. aspiration as a diagnostic technique in lymphoma.
Foucar E, Rosai J, Dorfman RF. Sinus histiocytosis with J Clin Oncol, 2004;22(15):3046–3052.
massive lymphadenopathy. An analysis of 14 deaths Henter JI, Horne A, Arico M, et al. HLH-2004: diagnostic
occurring in a patient registry. Cancer, and therapeutic guidelines for hemophagocytic
1984;54(9):1834–1840. lymphohistiocytosis. Pediatr Blood Cancer,
Foucar E, Rosai J, Dorfman R. Sinus histiocytosis with 2007;48(2):124–131.
massive lymphadenopathy (Rosai–Dorfman disease): Herrada J, Cabanillas F, Rice L, Manning J, Pugh W. The
review of the entity. Semin Diagn Pathol, clinical behavior of localized and multicentric Castleman
1990;7(1):19–73. disease. Ann Intern Med, 1998;128(8):657–662.
Frizzera G. Castleman’s disease and related disorders. Semin Horne A, Janka G, Maarten Egeler R, et al. Haematopoietic
Diagn Pathol, 1988;5(4):346–364. stem cell transplantation in haemophagocytic
Frizzera G, Peterson BA, Bayrd ED, Goldman A. A systemic lymphohistiocytosis. Br J Haematol,
lymphoproliferative disorder with morphologic features 2005;129(5):622–630.
of Castleman’s disease: clinical findings and Horneff G, Jurgens H, Hort W, Karitzky D, Gobel U. Sinus
clinicopathologic correlations in 15 patients. J Clin Oncol, histiocytosis with massive lymphadenopathy
1985;3(9):1202–1216. (Rosai–Dorfman disease): response to methotrexate and
Gaba AR, Stein RS, Sweet DL, Variakojis D. Multicentric mercaptopurine. Med Pediatr Oncol, 1996;27(3):187–192.
giant lymph node hyperplasia. Am J Clin Pathol, Hornick JL, Jaffe ES, Fletcher CD. Extranodal histiocytic
1978;69(1):86–90. sarcoma: clinicopathologic analysis of 14 cases of a rare
296
Chapter 17: Atypical disorders
epithelioid malignancy. Am J Surg Pathol, Minkov M, Prosch H, Steiner M, et al. Langerhans cell
2004;28(9):1133–1144. histiocytosis in neonates. Pediatr Blood Cancer,
Jang YJ, Park KH, Seok HJ. Management of Kikuchi’s 2005;45(6):802–807.
disease using glucocorticoid. J Laryngol Otol, Minkov M, Steiner M, Potschger U, et al. Reactivations in
2000;114(9):709–711. multisystem Langerhans cell histiocytosis: data of the
Janku F, Amin HM, Yang D, et al. Response of histiocytoses international LCH registry. J Pediatr,
to imatinib mesylate: fire to ashes. J Clin Oncol, 2008;153(5):700–705.
28(31):e633–636. Miras-Parra FJ, Parra-Ruiz J, Gomez-Morales M,
Jindal B, Vashishta RK, Bhasin DK. Vascular transformation Gomez-Jimenez FJ, de la Higuera-Torres-Puchol J.
of sinuses in lymph nodes associated with Inflammatory pseudotumor of lymph nodes with focal
myelodysplastic syndrome–a case report. Indian J Pathol infiltration in liver and spleen. Dig Dis Sci, 2003; 48(10):
Microbiol, 2001;44(4):453–455. 2003–2004.
Johnston A, Coyle L, Nevell D. Prolonged remission of Nakamura S, Koshikawa T, Kitoh K, et al.
refractory lymphomatoid granulomatosis after Interdigitating cell sarcoma: a morphologic and
autologous hemopoietic stem cell transplantation with immunologic study of lymph node lesions in four cases.
post-transplantation maintenance interferon. Leuk Pathol Int, 1994;44(5):374–386.
Lymphoma, 2006;47(2):323–328. Oksenhendler E, Duarte M, Soulier J, et al.
Jung KH, Sung HJ, Lee JH, et al. A case of pulmonary Multicentric Castleman’s disease in HIV infection: a
lymphomatoid granulomatosis successfully treated by clinical and pathological study of 20 patients. AIDS,
combination chemotherapy with rituximab. 1996;10(1):61–67.
Chemotherapy, 2009;55(5):386–390. Olnes MJ, Nicol T, Duncan M, Bohlman M, Erlich R.
Katzenstein AL, Carrington CB, Liebow AA. Lymphomatoid Interdigitating dendritic cell sarcoma: a rare malignancy
granulomatosis: a clinicopathologic study of 152 cases. responsive to ABVD chemotherapy. Leuk Lymphoma,
Cancer, 1979;43(1):360–373. 2002;43(4):817–821.
Keller AR, Hochholzer L, Castleman B. Hyaline-vascular Osborne BM, Butler JJ, Gresik MV. Progressive
and plasma-cell types of giant lymph node hyperplasia of transformation of germinal centers: comparison of 23
the mediastinum and other locations. Cancer, pediatric patients to the adult population. Mod Pathol,
1972;29(3):670–683. 1992;5(2):135–140.
Kojima M, Nakamura N, Sakamoto K, et al. Progressive Peterson BA, Frizzera G. Multicentric Castleman’s disease.
transformation of the germinal center of extranodal Semin Oncol, 1993;20(6):636–647.
organs: a clinicopathological, immunohistochemical, Pulsoni A, Anghel G, Falcucci P, et al. Treatment of sinus
and genotypic study of 14 cases. Pathol Res Pract, histiocytosis with massive lymphadenopathy (Rosai–
2010;206(4):235–240. Dorfman disease): report of a case and literature review.
Kudo K, Ohga S, Morimoto A, et al. Improved outcome of Am J Hematol, 2002;69(1):67–71.
refractory Langerhans cell histiocytosis in children with Rezai K, Kuchipudi S, Chundi V, et al. Kikuchi–Fujimoto
hematopoietic stem cell transplantation in Japan. Bone disease: hydroxychloroquine as a treatment. Clin Infect
Marrow Transplant, 2010;45(5):901–906. Dis, 2004;39(12):e124–126.
Leger-Ravet MB, Peuchmaur M, Devergne O, et al. Rosai J, Dorfman RF. Sinus histiocytosis with
Interleukin-6 gene expression in Castleman’s disease. massive lymphadenopathy. A newly recognized
Blood, 1991;78(11):2923–2930. benign clinicopathological entity. Arch Pathol,
Marsh RA, Vaughn G, Kim MO, et al. Reduced-intensity 1969;87(1):63–70.
conditioning significantly improves survival of patients Samet A, Gilbey P, Talmon Y, Cohen H. Vascular
with hemophagocytic lymphohistiocytosis transformation of lymph node sinuses. J Laryngol Otol,
undergoing allogeneic hematopoietic cell 2001;115(9):760–762.
transplantation. Blood, 2010;116(26):5824–5831.
Scott D, Cabral L, Harrington WJ, Jr. Treatment of
McCarty MJ, Vukelja SJ, Banks PM, Weiss RB. HIV-associated multicentric Castleman’s
Angiofollicular lymph node hyperplasia (Castleman’s disease with oral etoposide. Am J Hematol,
disease). Cancer Treat Rev, 1995;21(4):291–310. 2001;66(2):148–150.
Minkov M, Grois N, Heitger A, et al. Response to initial Song JY, Lee J, Park DW, et al. Clinical outcome
treatment of multisystem Langerhans cell histiocytosis: and predictive factors of recurrence among patients
an important prognostic indicator. Med Pediatr Oncol, with Kikuchi’s disease. Int J Infect Dis,
2002;39(6):581–585. 2009;13(3):322–326. 297
Chapter 17: Atypical disorders
298
Index
autologous stem cell transplantation blood hyperviscosity syndrome, CALGB regimen, 179, 182
(ASCT) (cont.) 139–141 Campath, 129, 240, 264
in T-cell non-Hodgkin lymphoma, bone marrow transplant, 241 Campylobacter jejuni
245–247 Borrelia burgdorferi association with IPSID, 116
relapse following, 77–78 association with cutaneous MALT association with Miller Fisher
Avastin, 240 lymphoma, 24, 114–115 syndrome, 142
azathioprine, 274, 275 bortezomib, 98, 144, 149, 168, 201, 240, association with small intestine
249, 265, 278, 279 MALT lymphoma, 114
bacterial infection mantle cell lymphoma studies, 166 cancer registration
role in lymphoma aetiology, 8 use in Waldenström’s impact of classification issues, 2
B-cell acute lymphoblastic leukemia, 172 macroglobulinemia, 147 candida infection
See also lymphoblastic lymphoma. brentuximab vedotin (SGN-35), 240 imaging, 43
B-cell lymphoma, unclassifiable, with Bruton’s tyrosine kinase (Btk) carboplatin, 76, 77, 203
features intermediate between inhibitors, 206 cardiac toxicity
DLBCL and Burkitt lymphoma bryostatin, 206 treatment related, 82
cytogenetics, 175 B-symptoms carmustine, 76, 203, 261
molecular pathology, 175 definition, 27 lung toxicity, 41–42
pathology, 173 prognostic significance in Hodgkin CD30 + T-cell lymphoproliferative
B-cell lymphoma, unclassifiable, with lymphoma, 27 disorders, 266–268
features intermediate between Burkitt lymphoma CD30 expression
DLBCL and cHL, 66–67 as post-transplant in HRS cells, 30
B-cell prolymphocytic leukemia, 122 lymphoproliferative disorder, prognostic significance in Hodgkin
B-cell receptor (BCR) pathway 276 lymphoma, 30
inhibitors association with EBV, 8 cell cycle inhibitors, 168
mantle cell lymphoma studies, 168 B-cell lymphoma, unclassifiable, centrocytic lymphoma, 155
bcl-2 inhibitors, 168 with features intermediate chemotherapy
BCNU, 181, 214, 261 between DLBCL and Burkitt therapeutic strategies, 11
BDR regimen, 149 lymphoma, 173, 175 Chlamydophila psittaci
BEACOPP regimen, 71–75, 84, 280 classification, 173 association with adnexal ocular
BEAM regimen, 76, 181, 214 clinical presentation, 172 MALT lymphoma, 24, 114–115
Bence Jones proteinuria, 143 clinical variants, 172 chlorambucil, 41, 93, 94, 111, 127, 128,
bendamustine, 97, 98, 144, 162 CNS prophylaxis and treatment, 221 144, 264
mantle cell lymphoma studies, cytogenetics, 174–175 use in Waldenström’s
165–166 endemic Burkitt lymphoma, 172 macroglobulinemia, 145–146
use in Waldenström’s HIV-associated Burkitt lymphoma, 2-chlorodeoxyadenosine, 264
macroglobulinemia, 148 182 CHOEP regimen, 199, 234
benign lymphoproliferative disorders, immunodeficiency associated chonic lymphocytic leukemia
286–288 Burkitt lymphoma, 172 cell immunophenotypes, 122
classification, 286 immunophenotypes, 173 CHOP regimen, 14, 55, 94, 161, 196,
inflammatory pseudotumor of in HIV patients, 277–278 198, 234, 236, 239, 241, 264,
lymph nodes, 288 management of relapse, 181 277, 279, 289, 294
Kikuchi’s disease, 286–287 molecular pathology, 174–175 addition of rituximab, 17
progressive transformation of natural history, 172 variations, 17
germinal centers, 287 pathology, 172–173 CHOP-bleomycin regimen, 94
vascular transformation of sinuses, prognostic factors, 24–25 CHOP-DHAP regimen, 163
287–288 age, 24–25 CHOP-rituximab regimen, 199
bevacizumab, 240, 251 at diagnosis, 24–25 ChP regimen, 94
bexarotene, 261, 264 clinical factors at diagnosis, chromosome translocations
Bexxar, 205 24–25 MALT lymphomas, 24
BFM 90 regimen, 179, 186 refractory disease, 25 chronic lymphocytic leukemia 121
BFM 95 regimen, 186 relapsed disease, 25 advances in understanding and
BFM regimen, 183 treatment-related, 25 treatment, 121
BH3 mimetics, 168 sporadic Burkitt lymphoma, 172 allogenic stem cell transplantation,
bleomycin, 55, 69, 70, 74, 80, 92, 94, St Jude classification, 38 131
197, 234, 276, 280 staging, 177 B-cell prolymphocytic leukemia, 122
lung toxicity, 41–42 supportive care, 187–188 chemotherapeutic approach, 126
blood-brain barrier (BBB) treatment, 178–181 choice of first line therapy, 127–129
effects in CNS lymphoma treatment principles, 177–178 choice of therapy in relapse,
chemotherapy, 209–210 treatment recommendations, 182 130–131
300 reversible disruption, 214 tumor lysis syndrome, 187 differential diagnosis, 160
Index
factors defining choice of therapy, historical background, 45 intravascular large B-cell lymphoma,
126–127 informed consent, 46 222
future treatment challenges, 134–135 intent-to-treat analysis, 50, 57 leptomeningeal lymphoma,
lymph-node structure, 121–122 interim analysis, 50, 53 214–216
minimal residual disease, 124–125 Kaplan-Meier estimator, 50 low-grade marginal zone B-cell
monoclonal antibody therapy, lymphoma specific survival lymphoma (MALT type), 222
129–130 endpoint, 47 methotrexate therapy, 210
novel therapeutic agents, 131–134 maximum tolerated dose (MTD), plasmacytoma, 222–223
recommended therapeutic 50–52 primary CNS Hodgkin lymphoma,
approaches, 134 Modified Cheson criteria, 47 223
treatment by risk stratification, outcome data, 47 primary DLBCL of the CNS, 208
123–124 overall survival (OS) endpoint, 47 prophylaxis against CNS relapse,
use of alkylating agents, 127 patient eligibility criteria, 48–49 220–221
CHVP/interferon regimen, 95 phase I trials, 50–52 radioimmunotherapy, 212
cisplatin, 76, 77, 163, 203, 234, 244 phase II trials, 52–53, 54 radiotherapy, 212–213
cladribine, 144, 242, 290 phase III trials, 54–57 rare forms, 222–223
use in Waldenström’s power of a study, 50 relapsed disease, 217
macroglobulinemia, 146 precision, 48–50 reversible BBB disruption, 214
clarithromycin, 111 progression-free survival (PFS) rituximab therapy, 212
classical Hodgkin lymphoma (cHL) endpoint, 47 small lymphocytic lymphoma, 208
associations with chronic qualitative outcome measures, 47 spinal cord lymphoma, 217
inflammatory rheumatic quantitative outcome measures, 47 T/NK cell lymphoma, 208
disease, 9 randomization methods, 57 T-cell phenotype of primary CNS
characteristic cell types, 61 randomized phase II trials, 54 lymphoma, 222
Hodgkin Reed-Sternberg (HRS) response criteria, 47 temozolomide therapy, 212
cells, 61, 68 study population, 48–49 topotecan therapy, 212
pathology, 63–64 time to event data, 47 treatment of secondary CNS
Reed–Sternberg cells, 63–64, 66 time to progression (TTP) lymphomas, 221–222
classification of haematological endpoint, 47 types of, 208
malignancies type I and type II errors, 49–50 vanishing tumor effect of
challenges, 1 use of placebo, 54–55 corticosteroids, 212
WHO classification, 1 CNS lymphomas whole-brain radiotherapy (WBRT),
clinical features aggressive lymphomas, 222–223 212–213
use in classification of lymphomas, 1 allogenic stem cell transplantation, CODOX-M regimen, 180, 182
clinical trials in lymphoma 222 CODOX-M/IVAC regimen, 277
accelerated titration method, 52 autologous stem cell transplantation coeliac disease
confidence intervals for results, 48 (ASCT), 221–222 association with lymphomas, 9
continual reassessment method chemotherapy, 209–212 cold agglutinin hemolytic anemia, 142
(CRM), 51–52 combined chemo-radiotherapy, common variable immunodeficiency,
contribution to therapy 213–214 273
development, 58 conventional therapeutic strategies COMP regimen, 183
Cox regression model, 50 for primary CNS lymphoma, Competence Network Malignant
definition of p-value, 50 208–209 Lymphoma, 48
definition of randomization, 50 corticosteroid therapy, 212 complete response (CR)
definition of sample size, 50 cytarabine therapy, 210 IWC definition, 39
definition of significance level, 50 dural based MALT lymphoma, 208 complete response unconfirmed (CRu)
definition of significance test, 50 EBV positive CNS lymphomas, 208 IWC definition, 39
definitions of endpoints, 47 effects of the blood-brain barrier complications of therapy
design complexity and challenges, 45 (BBB), 209–210 acute respiratory distress syndrome
determining endpoints, 46–47 experimental therapeutic strategies, (ARDS), 41–42
development of new therapeutic 214 angioinvasive aspergillosis, 42–43
approaches, 45 high dose chemotherapy with ASCT, bleomycin-related lung injury, 41–42
dose-limiting toxicity (DLT), 50–51 214 carmustine-related lung injury, 41–42
error control, 49–50 histopathology, 208 cyclophosphamide-related lung
evaluating treatment effect, HIV-related lymphomas, 217–220 injury, 41–42
47–48 immunocompromised patients, 208, diffuse alveolar damage, 41–42
event-free survival (EFS) endpoint, 217–220 frank radiation pneumonitis, 42
47, 50 immunocytoma, 222 imaging, 41–43
future developments, 58 indolent lymphomas, 222 immunocompromised patients,
hazard function, 50 intraocular lymphoma, 216–217 42–43 301
Index
complications of therapy (cont.) dacarbazine, 70, 79, 80, 84, 280 factors predicting response to H.
non-specific interstitial pneumonia DA-EPOCH regimen, 277 pylori eradication, 112–113
(NSIP), 42 DA-EPOCH-R regimen, 181, 182, 199 germinal center-like (GC) DLBCL, 16
organizing pneumonia, 42 data reporting high dose therapy and ASCT,
paramediastinal fibrosis, 42 challenges related to cancer 199–200
pulmonary haemorrhage, 42 registration, 2 high dose therapy and ASCT in
radiotherapy, 42 decadron, 234 relapsed disease, 203–204
computed tomography (CT), 32 dendritic cell disorders histone deacetylase (HDAC)
PET-CT, 37–38, 48 classification, 286 inhibitors, 205–206
staging of lymphomas, 32–34 diagnostic challenges, 286 immunoblastic variant, 191–192
treatment response assessment, follicular dendritic cell sarcoma, 291 immunophenotypes, 192
38–40 interdigitating dendritic cell in HIV patients, 276–277
computerized axial tomography sarcoma, 291 intravascular large B-cell lymphoma,
(CT scan), 48 denileukin diftitox, 236, 240, 249, 264 193
confidence interval 2-deoxycoformycin, 264 IPI and R-IPI classification systems,
definition, 50 Dexa-BEAM regimen, 76 12–14
for study results, 48 dexamethasone, 55, 76, 77, 129, 149, management of antibiotic-resistant
continual reassessment method 163, 181, 203, 234, 244, 276, 292 cases, 114
(CRM), 51–52 DHAP regimen, 76, 77, 203, 234 management of relapsed DLBCL,
COPBLAM regimen, 234 diagnosis of haematological neoplasms 203–206
COPDAC regimen, 80 multiple parameters used, 1 mediastinal (thymic) large B-cell
COPP regimen, 80 diagnosis of lymphoma lymphoma, 193–194
COPP/ABVD regimen, 72 changing diagnostic practice, 7 molecular follow-up, 113–114
core biopsy, 32 methods used, 1 molecular pathology, 194–195
corticosteroids, 94, 241, 261 diffuse alveolar damage morphological variants, 191–192
use in CNS lymphomas, 212 complication of therapy, 41–42 mTOR inhibitors, 206
vanishing tumor effect, 212 imaging, 41–42 new therapeutic targets, 205–206
Cotswolds modification. See Ann diffuse large B-cell lymphoma, pathology, 191–194
Arbor classification (DLBCL) post-treatment histologic evaluation,
Cox regression model, 50, 57 activated B cell (ABC)-like DLBCL, 16 112
Crohn’s disease addition of new agents to primary DLBCL of the CNS, 208
association with lymphomas, 9 chemotherapy, 201 prognostic factors, 12–18, 196
cryoglobulinemia, 141 addition of rituximab to at diagnosis, 12–16
Cumulative Illness Rating Scale chemotherapy, 200–201 at relapse, 17–18
(CIRS), 126 allogenic stem cell transplantation, 204 CHOP regimen, 17
cutaneous large B-cell lymphoma, 271 Ann Arbor classification, 12–14 clinical factors at diagnosis,
cutaneous MALT lymphoma anti-Helicobacter therapy, 111 12–14
association with Borrelia burgdorferi, associated with inflammation, 193 clinical factors at relapse, 17–18
24, 114–115 association with autoimmune DLBCL subtypes, 15
cutaneous γδ T-cell lymphoma, 226, diseases, 9 dose intensity and schedule, 17
241 B-cell lymphoma, unclassifiable, gene expression profiling, 16
pathology, 241 with features intermediate gene expression profiling at
treatment, 241 between DLBCL and Burkitt relapse, 18
CVP regimen, 94, 234 lymphoma, 173, 175 genetic abnormalities, 16
CVP-bleo regimen, 92 B-cell lymphoma, unclassifiable, histological factors, 15–16
cyclophosphamide, 17, 25, 55, 69, 72, with features intermediate immunophenotypes, 15–16
77, 78, 80, 92, 93, 94, 95, 127, between DLBCL and cHL, PET-18F-FDG, 17
144, 162, 163, 180, 181, 183, 66–67 relapsed DLBCL, 204–205
196, 197, 199, 234, 245, 247, centroblastic lymphoma, 191 time to complete remission (CR),
276, 280, 295 classification of subtypes, 15 16–17
lung toxicity, 41–42 clinical follow-up, 113–114 treatment-related factors, 16–17
cyclosporine, 236, 274, 275, 292 clinical presentation, 191 treatment-related factors at
cytarabine, 41, 55, 76, 77, 163, 178, 180, cytogenetics, 194–195 relapse, 17–18
181, 197, 203, 214, 221, 234, 290 diffuse large B-cell lymphoma NOS, protein kinase C β (PKCβ)
use in CNS lymphomas, 210 191–192 inhibitors, 206
cytomegalovirus (CMV), 274 dose dense therapy, 198–199 radiolabeled monoclonal antibodies
imaging of infection, 43 dose intensified therapy, 198–199 in relapsed disease, 205
cytophagic histiocytic panniculitis EBV-positive DLBCL of the elderly, second line therapy prior to ASCT,
(CHP), 240 193 203
302 cytosine arabinoside, 76 extranodal disease, 191 staging investigations, 195
Index
allogenic stem cell transplantation, sequential dose intensification and mantle cell lymphoma, 167
165 autologous transplantation, use in DLBCL, 206
autologous stem cell transplantation 163–164 mucosa-associated lymphoid tissue. See
(ASCT), 162 staging, 159 MALT lymphoma
B-cell receptor (BCR) pathway temsirolimus studies, 167 multicentric Castleman’s disease,
inhibitors, 168 therapeutic challenges, 155 293–294
bcl-2 inhibitors/BH3 mimetics, 168 Mantle Cell Lymphoma International multi-parameter, four-color flow
bendamustine studies, 165–166 Prognostic Index (MIPI), 21, 161 cytometry (MRD flow), 125
bortezomib studies, 166 mantle zone lymphoma, 155 mycophenolate mofetil, 275
cell cycle inhibitors, 168 marginal zone lymphomas (MZL) mycosis fungoides, 254
chromosomal aberrations, 157–158 association with autoimmune clinical features, 255–255
classification, 155, 156 diseases, 9 clinicopathologic variants, 265–266
clinical presentation, 155–156 normal counterparts, 24 folliculotropic variant, 265
clinical prognostic markers, 160–161 prognostic factors, 24 granulomatous variants, 266
CNS relapse, 155 subtypes, 24 molecular pathology, 259–260
conventional chemotherapy, 161 maximum tolerated dose (MTD), pagetoid reticulosis, 265
cytogenetics, 157–158 50–52 pathology, 256–259
differential diagnosis, 160 M-BACOD regimen, 55, 234, 276 staging, 255–255
diffuse type, 21 MCP regimen, 94 syringotropic variant, 265
disease course and timescale, 155–156 MDX-060 (Medarex), 78, 240 treatment, 261–265
dose-intensified regimens, 163 MDX-1401, 240 myelodysplasia, 145, 163
everolimus studies, 167 MEC regimen, 73
familial aspect, 155 mechlorethamine, 70, 74, 93, 261 National Cancer Institute (NCI), 45, 46
future directions, 168 mediastinal (thymic) large B-cell navitoclax (ABT-263), 132–133, 168
HDAC inhibitors, 168 lymphoma, 193–194 neuropathy
immunomodulatory drugs, 167 Mediterranean lymphoma, 115 IgM related, 141–142
immunophenotypes, 157 MegaCHOEP regimen, 245 nitrogen mustard, 261
incidence, 155 melphalan, 76, 78, 181, 203, 214 NK/T-cell lymphoma (NK/TCL), nasal
initial therapy, 161 6-mercaptopurine, 290 type, 226
lymph node architecture, 156–157 methotrexate, 26, 55, 77, 93, 163, 178, NK-cell neoplasms, 244
management of relapsed disease, 180, 181, 183, 186, 197, 213, nodal marginal zone lymphoma, 24
165–166 214, 219, 221, 234, 236, 241, clinical features, 118
minimal residual disease, 161 244, 264, 276, 292, 294 genetics, 118
molecular markers, 159–160 use in CNS lymphomas, 210 management, 118–119
molecular pathogenesis, 157–158 methyl-GAG, 234 pathology, 118
monoclonal antibody and combined methylprednisolone, 77 nodular lymphocyte predominant HL
immunochemotherapy, methylprednisone, 234 characteristic cell types, 61
161–162 metronidazole, 111, 116 lymphocytic and histiocytic (L&H)
morphologic subtypes, 159 Miller Fisher syndrome cells, 61–63
mTOR inhibitor studies, 167 association with Campylobacter pathology, 61–63
myeloablative therapy followed by jejuni, 142 popcorn cells, 61–63
ASCT, 163–164 mini-BEAM regimen, 203 treatment, 80
nodular type, 21 mitoxantrone, 94, 162, 234 nodular sclerosis HL
novel therapeutic approaches, mixed cellularity HL pathology, 65–66
167–168 pathology, 65 non-Hodgkin lymphoma (NHL)
pathology, 156–157 Modified Cheson criteria, 47 association with HIV/AIDS, 7–8
phenotypic markers, 159–160 monoclonal antibodies, 11, 18, 94 change in rates over time, 9
PI3K/AKT pathway therapeutic novel agents, 98–99, 133–134 changes in rates over time, 6–7
targets, 167–168 use in chronic lymphocytic HIV-related non-Hodgkin
prognostic factors, 159–161 leukemia, 129–130 lymphoma, 276
at diagnosis, 11–12 use in mantle cell lymphoma, International Prognostic Index (IPI),
clinical factors at diagnosis, 21 161–162 12, 13
genetic abnormalities, 21–23 use in Waldenström’s use as a reporting category, 2
histological factors, 21 macroglobulinemia, 130, See also specific subtypes.
prognostic indexes, 21 146–147 See also specific agents. non-Hodgkin’s Lymphoma Pathologic
radioimmunotherapy, 162 MOPP regimen, 84 Classification Project, 49
radioimmunotherapy in relapsed/ MOPP/ABV regimen, 71 non-specific interstitial pneumonia
refractory disease, 166 MOPP/ABVD regimen, 70 (NSIP)
rituximab maintenance therapy, motor neuron disease, 142 complication of therapy, 42
164–165 mTOR inhibitors, 77, 79 nuclear factor kappa B (NFκB), 24 307
Index
obatoclax (GX 15–070), 168 peripheral T-cell lymphoma, not pregabelin, 149
obesity otherwise specified (PTCL- primary CNS lymphoma, 208
potential risk factor for NOS), 23, 226, 232–234 HIV related, 279
lymphoma, 9 clinical presentation, 232 incidence, 25
OEPA regimen, 80 cytogenetics, 233 prognostic factors, 25–26
ofatumumab, 98, 133, 144 incidence, 232 at diagnosis, 25–26
use in Waldenström’s molecular pathology, 233 clinical factors at diagnosis, 25–26
macroglobulinemia, 147 pathology, 232–233 histological factors, 26
OKT3, 274 treatment, 233–234 immunodeficiency, 25
omeprazole, 111 PET scan organ transplantation, 25
Oncovin, 72, 74, 77 role in response evaluation, 47–48 treatment-related factors, 26
Ontak, 240, 249, 264 PET-18F-FDG, 11, 70 See also CNS lymphomas.
OPPA regimen, 80 in HIV positive HL patients, 280 primary cutaneous (anaplastic) CD30
organ transplantation predictive value in DLBCL, 17 positive large cell lymphoma,
lymphoma risk, 9 prognostic value in Hodgkin 268
potential influence on lymphoma lymphoma, 30 primary cutaneous aggressive
rates, 7 staging of lymphomas, 36–38 epidermotropic CD8+ T-cell
risk factor for primary CNS treatment response assessment, 40–41 lymphoma, 269
lymphoma, 25 PET-18FLT primary cutaneous anaplastic large-cell
See also post-transplant staging of lymphomas, 38 lymphoma (ALCL), 236
lymphoproliferative disorders. PET-CT, 48 primary cutaneous B-cell lymphoma,
organizing pneumonia staging of lymphomas, 37–38 254, 269
complication of therapy, 42 photochemotherapy, 261–263 classification system, 254
outcome measures, 47 phototherapy, 261–263 cutaneous large B-cell lymphoma, 271
overall survival (OS) PI3K/AKT pathway primary cutaneous large B-cell
definition, 47, 50 therapeutic targets, 167–168 lymphoma, leg type, 271
endpoint, 12 plasmablastic lymphoma, 278–279 primary cutaneous marginal zone
plasmacytoma B-cell lymphoma/
pagetoid reticulosis, 265 CNS involvement, 222–223 immunocytoma, 269–270
panobinostat, 77 plerixafor, 132 primary follicle centre-cell
pantoprazole, 111 Pneumocystis carinii pneumonia lymphoma, 270
paramediastinal fibrosis, 42 (PCP) primary cutaneous CD4-small/
partial response (PR) imaging, 43 medium-sized pleomorphic
IWC definition, 39 pneumonia T-cell lymphoma, 269
PCI-37265, 206 imaging, 42 primary cutaneous DLBCL, leg type, 15
PD 0332991, 168 POEMS (polyneuropathy, primary cutaneous follicle centre
pediatric conditions organomegaly, endocrinopathy, lymphoma, 89
EBV associated lymphoproliferative M protein, and skin changes) primary cutaneous large B-cell
disease, 244 syndrome, 142 lymphoma, leg type, 271
EBV-related hemophagocytic polymerase chain reaction (PCR), 7 primary cutaneous lymphoma
syndrome, 244 popcorn cells, 61–63 classification, 254
follicular lymphoma, 89 positron emission tomography. See incidence of skin lymphoma, 254
hemophagocytic PET primary cutaneous B-cell
lymphohistiocytosis, 292 post-transplant lymphoproliferative lymphoma, 254
Langerhans cell histiocytosis, disorders, 9, 273–276 See also primary cutaneous B-cell
289–291 B-cell lymphomas, 273 lymphoma; primary cutaneous
St Jude classification, 38 Burkitt lymphoma, 276 T-cell lymphoma.
treatment of Hodgkin lymphoma in diagnosis, 274 primary cutaneous marginal zone
children, 80–81 etiology, 274 B-cell lymphoma/
See also Burkitt lymphoma. Hodgkin lymphoma, 276 immunocytoma, 269–270
pegylated liposomal doxorubicin, 77 incidence, 273–274 primary cutaneous peripheral T-cell
pentostatin, 242 prognostic factors, 274 lymphoma, unspecified,
peripheral T-cell lymphoma T-cell lymphomas, 273, 275 268–269
prognostic factors, 23 treatment, 274–275 primary cutaneous T-cell lymphoma,
ALK protein expression, 23 power of a study, 50 254
at diagnosis, 23 pralatrexate, 250–251, 265 CD30 + T-cell lymphoproliferative
clinical factors at diagnosis, 23 prednisolone, 55, 92, 94, 95, 244 disorders, 266–268
histological factors, 23 prednisone, 17, 25, 55, 69, 70, 72, 74, 80, classification system, 254
treatment-related factors, 23 93, 181, 183, 196, 197, 199, 234, granulomatous slack skin disease,
308 prognostic value of IPI, 26–23 236, 245, 280, 290 257, 260, 266
Index
thyroid lymphoma vincristine, 17, 25, 55, 69, 70, 72, 74, 77, incidence, 138
association with Hashimoto 80, 92, 93, 94, 163, 180, 181, 183, laboratory investigations and
thyroiditis, 9 196, 199, 234, 245, 276, 280, 290 findings, 143–144
time to event data, 47 vindesine, 74, 197, 234 lenalidomide therapy, 147–148
time to progression (TTP) vinorelbine, 76, 77, 280 lymph node biopsy findings, 144
definition, 47 VIPD regimen, 244 maintenance rituximab
endpoint, 12 viral infections therapy, 150
time to treatment failure (TTF) association with marginal zone monoclonal antibody therapies,
endpoint, 12 lymphomas, 114 146–147
topical steroids, 241 role in lymphoma aetiology, 7–8 morbidity mediated by effects of
topotecan vorinostat, 168, 265 IgM, 139–143
use in CNS lymphomas, 212 nature of the clonal cell, 139
131
I tositumomab, 205 Wald statistics, 57 nucleoside analog therapies, 146
total skin electron beam (TSEB) Waldenström’s macroglobulinemia ofatumumab therapy, 147
therapy, 263 auto-antibody activity, 141–142 prognosis, 144
toxin therapies, 264–265 bendamustine therapy, 148 response criteria, 150–151
treatment response assessment biochemical investigations, 143 rituximab therapy, 146–147
International Workship Criteria blood hyperviscosity syndrome, serum viscosity investigation, 143
(IWC), 38–41 139–141 thalidomide therapy, 147–148
PET-18F-FDG, 40–41 bone marrow findings, 143–144 tissue deposition, 142
use of CT, 38–40 bone marrow involvement, 139 treatment indications, 147–144
use of imaging, 38–41 bortezomib therapy, 147 treatment options, 144–150
tumor lysis syndrome, 178, 187 chlorambucil therapy, 145–146 whole brain radiotherapy (WBRT),
tumor necrosis factor alpha (TNF-α) classification, 138 212–213
inhibitors, 226 clinical features, 139 Wiskott–Aldrich syndrome, 8, 273
tumor suppressor genes cold agglutinin hemolytic anemia, 142 World Health Organization (WHO)
inactivation, 227 combination therapies, 148–150 classification of hematological
cryoglobulinemia, 141 malignancies, 1
ultrasound (US) examination, 32 cytogenetics, 138–139 classification, fourth edition, 49
unicentric Castleman’s disease, 293 effects of tissue infiltration by International Agency for Research
urticaria, 143 neoplastic cells, 142–143 on Cancer (IARC), 2
epidemiology, 138 PTLD classification, 9
vanishing tumor effect of etiology, 138 Wotherspoon score, 112
corticosteroids, 212 everolimus therapy, 148
vascular transformation of sinuses, familial aspect, 138 X-linked lymphoproliferative
287–288 hematological abnormalities, 143 syndrome, 273
VEBEP regimen, 280 high-dose therapy and stem cell
Velcade, 240 transplantation, 150 zanolimumab, 265
VIM-3 regimen, 234 IgM related neuropathy, 141–142 Zevalin, 205, 212
vinblastine, 70, 74, 79, 84, 280, imaging studies, 144 zidovudine, 220
290, 294 immunomodulatory agents, 147–148 Zubrod, Gordon, 45
311