From www.bloodjournal.org by on October 7, 2010. For personal use only.

1980 56: 329-343

Autoimmune thrombocytopenic purpura

S Karpatkin

Information about reproducing this article in parts or in its entirety may be found online at:
http://bloodjournal.hematologylibrary.org/misc/rights.dtl#repub_requests

Information about ordering reprints may be found online at:

http://bloodjournal.hematologylibrary.org/misc/rights.dtl#reprints

Information about subscriptions and ASH membership may be found online at:
http://bloodjournal.hematologylibrary.org/subscriptions/index.dtl

Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published semimonthly

by the American Society of Hematology, 1900 M St, NW, Suite 200, Washington
DC 20036.
Copyright 2007 by The American Society of Hematology; all rights reserved.
 From www.bloodjournal.org by on October 7, 2010. For personal use only.

The Journal of
B L 0 OD The American Society of Hematology

VOL. 56, NO. 3 SEPTEMBER 1980

REVIEW

Autoimmune Thrombocytopenic Purpura

By Simon Karpatkin

SUMMARY platelet count to normal. This can generally be accom-

Adult autoimmune throbocytopenic purpura (ATP) plished with a platelet count of >40,000/cu mm with
is a platelet disorder that develops in certain individu- patients having this disorder. Approximately 50% of
als with a genetic as well as sex (female) predisposition patients respond to steroids by a significant elevation
following an environment event (?viral). This results of platelet count and improvement of purpura.
in the production of an IgG antiplatelet antibody However, cessation of therapy results in eventual
capable of reacting with the host’s platelets, as well as relapse if the disease is of the chronic variety. Splenec-
crossing the placenta. This leads to the rapid clearance tomy is successful in approximately 659’o-7S% of
and destruction of opsonized platelets by the reticu- patients, resulting in a restoration of the platelet count
loendothelial system, particularly the spleen, by to normal or safe levels by removing a major source of
greater than tenfold the normal rate. Bound platelet platelet destruction as well as antibody production;
lgG correlates with disease severity, whereas serum platelet survival improves. At least 50% of patients “in
antiplatelet IgG does not. It has not been rigorously remission” following steroids or splenectomy generally
established whether bound platelet IgG is directed have a compensated thrombocytolytic state in which
against a platelet antigen or represents an immune increased platelet production keeps up with increased
complex bound to the platelet Fc receptor. Neverthe- platelet destruction. Antiplatelet IgG can often be
less, several lines of evidence suggest that antiplatelet found in the serum of these patients.
IgG binds directly to a platelet antigen(s). Patients refractory to steroids and/or splenectomy
Megakaryocyte number, volume, and mass are present with a serious therapeutic problem. Immuno-
increased commensurate with increased platelet turn- suppressive therapy is effective in approximately one-
over. Platelets of i ncreased size, megathrombocytes, third of refractory patients, but often relapses occur,
are noted on peripheral smear or via platelet volume requiring maintenance therapy with potentially
distribution analysis. Megathrombocyte number is mutagenic drugs. The use of ymca alkaloids has
proportionate to megakaryocyte number and to plate- recently been proposed. Its efficacy has yet to be
let turnover. Megathrombocyte diameter is inversely established and its use should be confined to research
proportional to platelet survival. centers.
Antiplatelet antibody is also associated with quali-
HISTORICAL BACKGROUND AND PATHOGENESIS
tative platelet functional defects, which are indistin-
guishable from those noted with thrombopathia (i.e., In the 1950’s, the term idiopathic thrombocytopenic
apparent platelet release defect). Antibody-induced purpura (ITP) referred to a clinical disorder of
functional defects are probably more common than unknown etiology associated with thrombocytopenia
quantitative thrombocytopenic defects and may repre-
sent a significant portion of those women with the From the New York University Medical School, New York, N.Y.
“easy bruising” syndrome and normal platelet count. Submitted November 28, 1979; accepted February 19. 1980.
Address reprint requests to Simon Karpatkin, M.D., New York
Adults who develop ATP generally develop the
University Medical School, Department of Medicine, 550 First
chronic variety, which remains premanently with the
Avenue, New York, N.Y. 10016.
patient. Treatment should be directed towards main- © 1980 by Grune & Stratton, Inc.
taming the patient free of purpura, not restoring the 0006-4971/80/5603--000J$02.o%

Blood, Vol. 56, No. 3 (September), 1980 329

 From www.bloodjournal.org by on October 7, 2010. For personal use only.

330 SIMON KARPATKIN

and purpura. The disorder could be acute and often capable of inducing thrombocytopenia when infused
transient, as in the case of childhood ITP, recurrent into volunteer recipients.
with episodes of “remission,” or chronic, as in adult
ITP. It was noted that mothers with ITP (with active SERUM ANTIBODY

disease or in remission) often gave birth to children Harrington then developed a simple platelet agglu-
who developed transient ITP, suggesting the transfer tinin test that consisted of the addition of heat-
of a humoral factor. It was also noted that patients inactivated “test” sera to an aliquot of normal plate-
with Coomb’s positive (autoimmune) hemolytic let-rich plasma followed by incubation overnight at
anemia often had episodes of ITP (Evan’s syndrome) 5#{176}C.3
The sera of patients with ITP were positive in
both transient and recurrent, suggesting an autoim- approximately 65% of cases.6 This antiplatelet factor
mune etiology.5 Furthermore, the favorable response was shown to be present in the gamma globulin frac-
to steroids (antilymphocytic) and/or splenectomy tion of sera. In 1965, Shulman et al.,7 employing the
(antireticuloendothelial) encouraged the thought that same in vivo test as Harrington,3 demonstrated that
this disease might be mediated by the production of the antiplatelet factor was in the 75 gamma globulin
anti-platelet antibody. fraction of serum, was adsorbed by human platelets,
In 1951, Harrington clearly established the humoral was species specific, and affected autologous as well as
nature of such an antiplatelet factor and provided the homologous platelets.
strongest evidence that this disease was antibody- The platelet agglutinin test3’714 as well as other
mediated.3 He infused the plasma equivalent of a unit tests’#{176}”5 for antiplatelet antibody, particularly the
of blood from a patient with ITP intravenously into antiglobulin consumption test,7’8”6’8 although repro-
himself, and promptly developed thrombocytopenic ducible in som&4”618 laboratories, was not reproduc-
purpura (Fig. 1). A drop in platelet count occurred ible in others7’8’”3”9 Indeed, some workers claimed
immediately, reaching its nadir in 1-3 hr and return- that the antiplatelet factor of serum was no more than
ing to normal in 4-6 days. Of interest was the observa- thrombin,’3 a potent platelet agglutinin. It was argued
tion that only 63% of patients with the diagnosis of that the absence of platelets and their phospholipid
ITP contained a factor in their plasma that was platelet factor 3 (PF-3) in these patients prevented
adequate utilization of prothrombin, leading to a high
residual prothrombin in their sera. This could then be
converted to thrombin by the addition of platelets to
the test system.’3
In 1968, Karpatkin and Siskind2#{176} employed an
ammonium sulfate globulin fraction of sera to develop
a PF-3 immunoinjury technique for measuring anti-
platelet antibody that circumvented the problem of
residual prothrombin or thrombin. This technique was
I
capable of detecting an antiplatelet factor in approxi-
0
0 mately 60%65%421 of patients with ITP, confirming
0
Harrington’s original work with platelet agglutinins,
as well as the work of Steffen,’6 Dausset,’7 and Van
w
-I
w
deWiel et al.’ with the antiglobulin consumption test.
I- The PF-3 immunoinjury technique takes advantage of
4
-J
a- the requirement of platelet phospolipid to act as a
catalytic surface for the coagulation cascade and is
measured by the acceleration of the clotting of plate-
let-rich plasma following addition of a factor capable
of injuring platelets.420’2’ The PF-3 immunoinjury
I 23 I 2 3 4 5 6 ‘7 8 9 technique has been recently modified so that it can be
HOURS DAYS employed as a rapid, routine laboratory test employing
Fig. 1 . Thrombocytopenic effect produced by transfusing 500
frozen platelets and requiring 4 hr for its perfor-
cc of citrated blood or its plasma equivalent from 1 7 patients with mance.2’ Positive results have been found in 60% of
thrombocytopenic purpura. Transfusions were given at “0” time. patients with ITP. Results are also positive in systemic
Recipients were healthy laboratory workers or patients with
inoperable carcinoma. The mean effect is represented by the
lupus erythematosus (SLE), rheumatoid arthritis,
heavy line.(Reproduced by permission.3) lymphoma, autoimmune hemolytic anemia, chronic
 From www.bloodjournal.org by on October 7, 2010. For personal use only.

AUTOIMMUNE THROMBOCYTOPENIC PURPURA 331

hepatitis, and following numerous blood transfusions. (tests are positive in >92% of thrombocytopenic
This antiplatelet factor has the characteristics of an patients).
IgG immunoglobulin in that it is stable at 56#{176}C;20”6its The immunofluorescent technique is relatively
activity can be neutralized with a globulin fraction of insensitive. The complement lysis-inhibition technique
rabbit anti-human IgG antisera;4’2#{176} but not anti-IgM, is complicated, detects IgG at the 10-25-ng level, and
IgA, or IgD;2#{176} its activity can be adsorbed to plate- can be employed with as little as 106 platelets. The
lets;4’7’9”6’2#{176}
the adsorbed activity can be eluted and the Fab-anti-Fab technique is also relatively complicated,
eluted4”6 activity neutralized with rabbit anti-human requires 3 days to perform, is sensitive at the I 5-20-ng
IgG but not with normal rabbit globulin;4’20 the eluted level, and can be performed on approximately 2 x 10
material has a molecular weight of 1 50,00020 and 520 platelets. The radioactive IgG Coomb’s test can be
value of 77#{149}5,16 similar to that of human IgG; and it applied to routine use but has a complicated method of
can be transferred across the pIacenta.’’2’ The heavy standardization (IgG fixed to red blood cells), is sensi-
chain subclass of the IgG immunoglobulin appears to tive at the 5-10-ng level, and requires approximately
be ‘yG3, and the IgG contains both kappa and lambda iO freshly collected platelets. The solid-phase
light chains.22 Some degree of platelet specificity does radioimmune assay can be performed with commer-
exist, since the anti-platelet antibody of 3 of 5 patients cial reagents and is relatively simple, except for the
studied reacts more intensely with their own platelets iodination of protein A. It is I 0-50-fold more sensitive
than with platelets from unrelated donors.22 The than other techniques, can be performed with as little
serum antibody is useful for diagnostic purposes only, as 2 x I 0 platelets, and can be employed with frozen
since its titer does not correlate with severity of platelet extracts (samples can be stored at - 20#{176}C
in
diseases.4 Results confirming the reliability of the the presence of protein inhibitors for future assay).
PF-3 immunoinjury technique2326’7’ and indicating the The inverse relationship between platelet count and
presence of antiplatelet antibody in ITP patients have platelet IgG is best expressed by a log-log equation
been published by a number of workers employing a with a correlation coefficient, r = -0.71, p < 0.001,
variety of different techniques.233’ Because of the rather than an arithmetic plot, r = 0.56, p < 0.001.
immunologic nature of this disorder, in which the This indicates considerably more antibody bound to
host’s immune system destroys the host’s platelets, it platelets from patients with low platelet counts than
was recommended that the term idiopathic be changed would be expected from a simple inverse arithmetic
to autoimmune and the disorder designated ATP or relationship. This could represent more antiplatelet
autoimmune thrombocytopenic purpura.32 This desig- antibody bound to larger younger platelets, mega-
nation has been accepted by other investigators.26’33 thrombocytes,”#{176} which are increased in number in this
disorder. However, a recent study suggests other-
BOUND ANTIPLATELET ANTIBODY
wise.4’ It is conceivable that younger platelets may be
In I 975, Rosse and coworkers34 made a major more resistant to destruction and clearance by the
contribution to this field by developing a complement- reticuloendothelial (RE) system.
mediated red blood cell lysis-inhibition technique for Platelet-bound IgG has also been recently reported
the quantitative detection of bound antiplatelet anti- in 46% of patients following gram-positive or -negative
body, wherein 1 7 of I 7 patients tested were positive. septicemias,42 where positive test results are thought to
Bound platelet IgG was found to correlate with sever- represent nonspecific binding of bacterial antigen-
ity of disease. Treatment of patients with steroids or antibody complexes to platelets. Accordingly, positive
splenectomy lowered platelet IgG; and patients with results associated with septicemias should be inter-
considerably elevated platelet IgG were refractory to preted with caution, before making the diagnosis of
corticosteroids or splenectomy. This work was quickly ATP.
confirmed by several laboratories employing similar35
COMPLEMENT
as well as different techniques, which include immuno-
fluorescence,36 Fab-anti-Fab IgG assay,37 radioac- Serum complement levels are generally normal, and
tive 251 Coomb’s test.3’ and a solid-phase radioimmu- attempts to measure serum complement-dependent
noassay for IgG, employing rabbit anti-human IgG antibody have been negative.7 Nevertheless, C3 has
“sandwiched” to radioactive ‘ 25I staphylococcal recently been reported to be bound to platelets in the
protein A, which binds to the Fc domain of IgG.39 The presence of bound IgG (9 of 14 patients with elevated
various techniques described measure 1-1 1 ng of platelet IgG; 9 of 16 total patients tested). The C3
nonspecific IgG on 106 normal platelets and 5-13 bound was proportional to the IgG bound. In another
times this amount on platelets from ATP patients study3’ employing a radioactive 1251 Coomb’s test, 12
 From www.bloodjournal.org by on October 7, 2010. For personal use only.

332 SIMON KARPATKIN

of 21 patients had elevated C3 in the absence of coated latex particles by rheumatoid factor or Clq.
elevated IgG levels. The pathophysiologic significance The sera of 48 of 72 patients with ATP (58%)
of fixation of C3 to platelets in 58% of the 36 patients displayed a significant inhibitory effect toward both
(from both studies) remains to be determined. agglutinating agents. A negative correlation was
obtained between the platelet count and the titer of the
ROLE OF THE SPLEEN IN ANTIBODY PRODUCTION
inhibitory factors. When the sera were subjected to
In 1971, Karpatkin et al.4 demonstrated decreased Sephadex G-200 gel filtration, IgG and DNA could be
circulating antiplatelet antibody postsplenectomy in 7 detected in “heavy factions” corresponding to Ag-Ab
of 8 patients studied. In 1972 both Karpatkin et al.43 complexes in 6 different patients. However, DNA was
and McMillan et al.44 independently demonstrated not necessarily eluted with “heavy IgG” or fractions
synthesis of specific IgG antiplatelet antibody by the with inhibitory activity for agglutination of IgG-
spleen, which reflected approximately 0.6%-5% of coated latex particles. Dissociation and association of
total IgG produced. Both laboratories also noted a “immune complex” inhibitory activity could be
considerable increase in the synthesis of nonspecific obtained with sera from 2 patients. These authors also
IgG by splenic tissue from these patients that was found hepatitis HbsAg in 20 sera, Epstein-Barr virus
5-55 times that produced by control spleens;43’44 an (EBV) antigen in 5 sera, and adenovirus antigens in 6.
observation that remains unexplained. This very interesting report remains to be confirmed.
However, confirmation of the presence of immune
ON THE NATURE OF THE ANTIGEN
complexes would still not rule out the presence of
Present technology for the detection of antiplatelet specific IgG against platelet antigens. These com-
antibody in the serum or on the platelet does not plexes could represent an epi-phenomenon, wherein
distinguish between IgG and lgG complexed to an immune complexes are developed against platelet or
antigen, i.e., an Ag-Ab complex bound to the platelet other cellular antigens produced during cell destruc-
Fc receptor. It is conceivable that ATP is an immune tion by “specific” platelet antibodies. In order to
complex disease wherein the IgG antibody reacts with conclusively demonstrate the presence of specific anti-
a foreign (?viral) antigen or a (nonplatelet) circulat- body against platelet antigens, it will be necessary to
ing host antigen, to form an immune complex that demonstrate that Fab fragments of eluted platelet IgG
then binds to the platelet; or a foreign antigen binds to or serum antiplatelet IgG specifically binds to plate-
the platelet and then reacts with an antiforeign antibo- lets. Experiments designed to answer this question
dy. Six lines of evidence suggest that this may not be have as yet not been reported.
the case: (1) Antiplatelet factor classically crosses the
placenta in this disease, resulting in the passive trans- ROLE OF CELL-MEDIATED IMMUNITY
fer of neonatal thrombocytopenia. IgG is known to In 1970, Piessens et al.46 demonstrated increased
cross the placenta, whereas immune complexes are less lymphocyte transformation (incorporation of tritiated
likely to do so. (2) The in vivo infusion experiments thymidine into DNA) with autologous platelets from a
had demonstrated the thrombocytopenic factor to be patient with ATP as well as homologous platelets plus
in the 75 gamma globulin fraction of sera.6’7 (3) The the pateint’s lymphocytes, suggesting a role for cell-
serum factor demonstrates greater specificity for the mediated toxicity. Wybran and Fudenberg47 per-
patient’s platelets.22 (4) Lymphocytes from cultured formed similar studies on a group of patients with this
spleens are capable of synthesizing IgG, which specifi- disorder and noted positive lymphocyte stimulation
cally binds to washed platelets.43’44 (5) The eluate from with autologous platelets in 7 of 8 patients with severe
bound antiplatelet IgG migrates in the 75 (IgG) disease and 3 of 6 patients with mild disease; 10
region of a sucrose gradient39 (although it is conceiva- normal subjects and 6 patients with nonimmune
ble that the complex may have disassociated with the thrombocytopenic disorders gave negative results.
extraction procedure, or that the Ag may be too small Similar observations were made by Clancy4t in 6 of 7
to appreciably affect the IgG sedimentation value). patients with the disorder. He also noted inhibition of
(6) Antibody activity, as defined by IgG platelet leukocyte migration by autologous platelets in 9 of 10
adsorbtion, is eluted from Sephadex G200 and DEAE patients tested. This technique is based on the fact that
cellulose at the area where IgG is found.33 when antigen is incubated with sensitized lympho-
Nevertheless, circulating immune complexes have cytes, a number of soluble factors are released, includ-
been recently reported in this disorder by Lurhuma et ing a migration inhibition factor that inhibits the
al.45 These workers employed two methods for the migration of macrophages, polymorphs, and lympho-
detection of circulating immune complexes based on cytes. However, it now appears likely that the results
their inhibitory effect on the agglutination of IgG- reported in their studies actually represented nonspe-
 From www.bloodjournal.org by on October 7, 2010. For personal use only.

AUTOIMMUNE THROMBOCYTOPENIC PURPURA 333

cific lymphocyte stimulation by platelet-antibody (compared to normal subjects) parallels the increase
complexes, which have been shown to stimulate in platelet turnover (compared to normal subjects).
lymphocyte transformation.49 Platelets have recently Although the older literature refers to “abnormal
been shown to be coated by antiplatelet antibody (see megakaryocytes,” reflecting decreased platelet pro-
Bound Platelet Antibody). In 1979, Quagliata and duction secondary to splenic inhibition of the bone
Karpatkin5#{176} reported impaired lymphocyte transfor- marrow via a humoral antogonist,6’64 the kinetic
mation in whole blood with mitogenic agents (phyto- evidence of Harker56 as well as Garg et al.#{176}
suggests
hemagglutinin and concanavalin A) in patients with otherwise. Nevertheless, common antigens for the
ATP (but not with washed lymphocytes). The number megakaryocyte and platelet have been demonstrated
of T cells, B cells, and null cells were normal. An with heterologous antiplatelet sera,6567 and more
inverse relationship was found for the number of T recently, with radioactive homologous splenic anti-
cells and platelet count in 31 ATP patients, with body synthesized in vitro.68 It is conceivable that
r = -0.55, p < 0.001. However, “killer” cells could impaired megakaryocytopoiesis69 may obtain in some
not be demonstrated employing lymphocytes from 7 situations; however, this has not been demonstrated as
ATP patients incubated with 51Cr-labeled allogeneic yet.
or syngeneic platelets. Lymphocyte capping with
PLATELET SIZE
rabbit antihuman lymphocyte IgG was impaired in 7
of 7 patients with ATP. Large platelets or megathromobocytes are routinely
The apparent disappearance of impaired lympho- noted in this disorder on EDTA smear or by volume
cyte transformation with washed lymphocytes and the determination in a Coulter Counter with EDTA
abnormal lymphocyte capping suggests the presence employed as anticoagulant#{176} (Fig. 2). Megathrombo-
of a blocking factor(s) or cell or humoral antibody(s) cytes generally correlate with megakaryocyte number
that could be responsible for the observed qualitative in most acquired platelet disorders of increased
abnormalities of cell-mediated immunity. destruction or decreased production, r = 0.7, p <

0.00l.#{176} Mean platelet diameter is 1.6-fold greater

PLATELET KINETICS
than normal in ATP. The megathrombocytes in this
It is generally accepted that thrombocytopenia is disorder are probably “stress” platelets, which are
secondary to increased platelet destruction of anti- released during conditions of increased megakaryocyte
body-coated platelets by particularly turnover. The increase in megathrombocytes parallels
in the spleen. Platelet survival is markedly shortened the increase in megakaryocytes noted and can be in
to less than 10% of the normal 10-day survival.5559 the order of 3-4-fold.4#{176} There is generally a shift in the
Megakaryocytes are increased in number, volume, and platelet volume distribution curve to the right towards
immaturity.4056’57 Platelets are decreased in number larger platelets. Approximately 50% of patients in
but generally increased in volume (see Platelet Size).
More specifically, Branehog et al.6#{176}
have demon-
LU
strated that platelet survival is proportional to the
circulating platelet count (at a platelet survival of 0-3
days). Harker 56 and Branehog et al.57 have both z
studied a total of 49 patients with comparable platelet I-
LU
-J
counts, mean 30,000/cumm. The mean platelet LU
survival was 0.3456 and 0.67 days, respectively. The
-J
mean platelet turnover was increased to 4956 and 2.3
times normal,” respectively. Megakaryocyte mass has LU
>
been quantified by Harker et al. who studied 14
I-.
patients. Megakaryocyte number averaged 3 times the 4
-J
control value, and megakaryocyte volume averaged LU

I .6-fold. Thus, the increase in total megakaryocyte

4060 P100
mass averaged 4.8-fold, indicating that under suitable
stress, the bone marrow can increase platelet produc- WINDOW SIZE
tion by approximately fivefold. The ratio of platelet Fig. 2. Coulter platelet volume distribution curves in 20
turnover to megakaryocyte mass in ATP patients was normal subjects ± 2 SD (grey area). (A) A patient with a normal
platelet count (1 60,000/cu mm) in “remission” from ATP (with
the same as that found in normal subjects. This
elevated platelet lgG. 25 ng lg/10 platelets). (B) A patient with
implies that thrombopoiesis is effective. Thus, the ATP and a platelet count of 55,000/cu mm. Calibration: 1 win-
increase in megakaryocyte mass in ATP patients dow = 0.25/cu p.
 From www.bloodjournal.org by on October 7, 2010. For personal use only.

334 SIMON KARPATKIN

“apparent” clinical remission have increased mega- patients with the “easy bruising syndrome” and
thrombocytes7#{176} (as well as bound platelet IgG39), mdi- normal platelet count. Platelet function was abnormal
cating increased platelet turnover despite a normal in 31 patients (29 were female) and was indistinguish-
platelet count. This is known as a “compensated able from the “aspirin-like” platelet aggregation
thrombocytolytic state.” There is a significant inverse defect reported in patients with ATP,7’ SLE,72 and
relationship between platelet diameter and platelet thrombopathiaY79 Of particular interest was the
mean life-span, r = -0.8, p < 0.001 .“ There also is a presence of antiplatelet antibody in 38% of these
significant relationship between platelet diameter and patients. Increased number of megathrombocytes was
platelet production rate, r = 0.7, p 0.001 and < also noted in 71% of these patients, suggesting
between platelet size and percent young megakaryo- increased platelet turnover despite a normal platelet
cytes, P = 0.6, p < 0.005.” count.
Patients with particularly severe disease also have In 1967, five independent groups779 described an
evidence for intravascular thrombocytolysis, as noted apparently new qualitative platelet disorder in which
by the presence of small fragments that can be the prolonged bleeding time, impaired collagen-
detected with the Coulter Counter by a shift of the induced platelet aggregation, and impaired epineph-
platelet volume distribution curve to the left or by a rine-induced secondary wave of aggregation were
separate small particle peak. These have been shown attributed to defective release of platelet adenosine
to contain platelet fragments as well as red blood cell diphosphate (ADP). The disorder was termed throm-
fragments by electron microscopy. The red blood cell bopathia and is indistinguishable from the acquired
fragments are associated with weak complement sensi- aspirin defect in normal subjects with respect to plate-
tization of red blood cells in 7 of I 2 patients, suggest- let aggregation. Most of the patients described have
ing that autoantibody might also be directed (subclini- been females. Three of six patients studied had
cally) against red blood cells.54 i ncreased megathrombocytes. Four patients had
histories that included ankylosing spondylitis, inter-
QUALITATIVE PLATELET DEFECTS
mittent arthritis, arthralgia, and myositis. Upon
In 1972, Clancy, Jenkins, and Firkin7’ reported the further study, two subgroups were documented: one in
presence of qualitative (functional) platelet defects in which the storage pool of adenine nucleotide and
I I patients in “remission” (normal platelet counts). serotonin associated with platelet dense granules was
Although no correlation was found between the diminished, termed “storage pool” disease,79 and the
presence of antiplatelet antibody (PF-3 technique) other in which the storage pool was normal, but the
and abnormal platelet function, an antiplatelet func- release from this pool with initiation of irreversible
tion factor could be isolated from the globulin fraction platelet aggregation was impaired, termed “release
of the patient’s serum that could inhibit platelet func- defect.”74
tion when mixed with normal platelet-rich plasma. The patients with thrombopathia (both varieties)
The antiplatelet function factor could be specifically are indistinguishable from the 31 easy bruising
adsorbed to washed human platelets and specifically patients described by Lackner and Karpatkin73 with
neutralized with anti-human lgG, suggesting that it respect to sex, clinical history (a mild bleeding disor-
was an immunoglobuin. However, a similar antiplate- der), and platelet function. The presence of antiplate-
let function factor was also found in 3 patients with let antibody in 38% of these patients, as well as the
mild intermittent disease in whom platelet function finding of similar platelet function abnormalities in
was normal. such autoimmune disorders as compensated ATP in
In 1974, Regan, Lackner, and Karpatkin72 remission and SLE, strongly suggests an autoimmune
described abnormal platelet function in 1 2 of 2 1 con- etiology or association for many of the patients with
secutive patients with SLE that correlated with clini- thrombopathia and/or the easy bruising syndrome.
cal severity of disease. No correlation was found This is supported by the disappearance of abnormal
between the presence of serum PF-3 antiplatelet anti- platelet function in at least 3 of the 3 1 patients
body and abnormal platelet function in the “nonre- studied, as well as the description of an acquired
sponder” group. However, an antiplatelet function storage pool disorder associated with antiplatelet anti-
factor could be isolated from the serum globulin frac- body that responded to prednisone therapy in a patient
tion of 3 of the 7 nonresponder patients, which could with nephritis, polyarthralgia, chondritis, thrombo-
inhibit platelet function of normal platelet-rich plas- phlebitis, Raynaud’s phenomenon, and a thrombotic
ma, whereas no such factor could be isolated from the tendency.8#{176} It is intriguing that 2 of the nonresponder
serum of 4 of 4 “responder” patients with SLE who SLE patients, 3 of the easy bruising patients, 2 of the
had normal platelet function. thrombopathia patients, and the patient with the
In 1975, Lackner and Karpatkin73 studied 75 acquired storage pool disorder had transient episodes
 From www.bloodjournal.org by on October 7, 2010. For personal use only.

AUTOIMMUNE THROMBOCYTOPENIC PURPURA 335

of thrombocytopenia. It is therefore likely that anti- now available that have specificities closely “related”
platelet antibody can contribute to or be responsible to alleles of the D locus and are therefore called DR or
for both quantitative as well as qualitative platelet DRw for D-related workshop designation. The DRw2
disorders. This can explain those patients with autoim- alloantigen has been found in 75% of 20 consecutive
mune thrombocytopenic purpura with so-called “safe” AlP patients in the New York City area, compared to
platelet counts of >50,000/cumm who have purpura 23% (<0.001) in an ethnically matched control popu-
and bruise easily. It is further proposed that some lation.’5 This provides a genetic predisposition with a
patients with thrombopathia and/or the easy bruising relative risk of 10. (A relative risk of I would indicate
syndrome may have a forme fruste of ATP, wherein no association between the presence of the antigen and
the antiplatelet antibody damage is insufficient to the disease.) Of particular interest was the apparent
result in thrombocytopenia, but sufficient to elicit typing of a single allele, DRw2, in I 0 of the 20 patients
functional platelet defects.73’8’ Perhaps two varieties of tested. Also noted were a high association of the
antiplatelet antibody are present: one capable of apparent haplotypes A3-B7 and A26-Bw38 of the
destroying platelets (? high-affinity antibody), and the HLA-A and B loci, which appear to be in linkage
other capable of altering platelet function (? low- disequilibrium with DRw2 in the population studied.
affinity antibody). It is suggested that thrombopathia These data indicate a genetic predisposition to ATP
and/or the easy bruising syndrome may be the bottom that is inherited with a DRw2 gene of the major
of the ATP iceberg. The incidence of this qualitative histocompatibility system. It is of interest that SLE
disorder(s) is considerably more common than classic has recently been shown to be associated with DRw2
84 and DRw3, with relative risks of 3.9 and 6.5, respec-
tively 66.87

THE ATP ICEBERG

It should be emphasized that thrombocytopenic DIAGNOSIS

purpura is merely the top of the iceberg, and that The term autoimmune thrombocytopenic purpura
probably many thousands of individuals have platelet should be applied to those patients fulfilling the
counts below 1 50,000/cu mm without purpura (95% following criteria.
probability of being abnormal”). These individuals (I) Increased platelet destruction as manifested by
are unaware of their illness, since symptoms of thrombocytopenia or shortened platelet survival i.e.,
purpura or bleeding do not occur until the platelet compensated thrombocytolytic states.40’57’82 Both con-
count reaches dangerous levels, i.e., <30-50,000/cu ditions are associated with an increased percent or
mm. There are still other individuals with a compen- number of megathrombocytes (young platelets).’#{176}’57’82
sated thrombocytolytic state, i.e., normal platelet (2) Increased number of megakaryocytes in the
count but decreased platelet survival and increased bone marrow. In the older literature,60 emphasis has
megathrombocytes.82 These individuals probably have been placed on abnormal megakaryocyte morphology:
subclinical compensated autoimmune thrombocytoly- i.e., absence of platelet budding, lack of granularity,
sis, wherein increased bone marrow production of vacuolization, cytoplasmic and nuclear degenerative
megakaryocytes keeps up with increased peripheral changes. These observations are probably a reflection
platelet destruction. These individuals are in precar- of megakaryocyte turnover and maturation. Any
ious balance, and any stress on the productive capacity disorder with increased megakaryocyte turnover is
of the bone marrow, such as nutritional (folate or B,,), likely to contain an increased number of young mega-
drugs, toxic substances (alcohol), or viral or bacterial karyocytes, which in turn would probably reflect most
infections as well as stress on platelet survival such as of the above findings.
alcohol, infection, etc., can precipitate thrombocytope- (3) Presence of bound antiplatelet antibody in the
nia. These individuals may possibly include the absence of septicemia42 or hypergammaglobaline-
numerous women with normal platelet counts and easy mia’57.
bruising syndrome and/or thrombopathia, some of (4) Exclusion of other primary clinical disorders
whom have antiplatelet antibody and defective platelet that are capable of giving all of the above criteria or
function. some of the above criteria: systemic lupus erythemato-
sus, lymphoma, disseminated intravascular coagula-
HLA TYPING AND GENETIC
tion, hypersplenism, drug-induced thrombocytopenic
PREDISPOSITION WITH DRw2
purpura, sepsis, etc. The distinction between ATP and
Recent studies on the genetic predisposition to ATP drug-induced immunologic thrombocytopenia is often
have revealed a high association with an alloantigen of crucial. All drugs should be withheld and the platelet
the HLA-D locus, which is responsible for the mixed count closely observed. Failure to return to a normal
lymphocyte stimulation reaction. Specific antisera are platelet count in 7-10 days (in the absence of serious
 From www.bloodjournal.org by on October 7, 2010. For personal use only.

336 SIMON KARPATKIN

hepatic or renal disease) rules out the diagnosis of normal tourniquet test, and absent to normal clot
drug-induced immunologic thrombocytopenia. retraction, depending on the platelet count. There are
(5) Absence of splenomegaly in 97% of patients. Its diminished platelets and increased megathrombocytes
presence usually,88 but not always (as might be the on peripheral smear. Platelet and red blood cell frag-
case with children8991), excludes the diagnosis of mentation may be detectable via Coulter Counter if
ATP. the thrombocytopenia is severe enough. Anemia
The term ITP should be reserved for those 5%-I 0% secondary to blood loss may be present. The white
of patients without demonstrable bound antiplatelet blood cell count is generally normal or slightly
antibody but who fulfill the remaining criteria. increased. There is often a relative lymphocytosis with
atypical lymphocytes (particularly in children) and
CLINICAL SIGNS AND SYMPTOMS
eosinophilia. The bone marrow contains increased
The signs and symptoms of this disorder are directly numbers of young megakaryocytes and often an
related to the platelet count and are not specific for increase in eosinophiles.89’96 The myeloid: erythrocyte
ATP or ITP but will be present in any quantitative (M:E) ratio may be decreased if blood loss is a
platelet disorder. When the platelet level is insufficient prominent feature.
to support hemostasis, generally <30-50,000/cu mm,
TEMPORAL CLASSIFICATION AND
capillary bleeding or purpura ensue. The degree of
NATURAL HISTORY
thrombocytopenia required for hemorrhage can be
quite variable, 10-50,000/cu mm, and is probably Autoimmune thrombocytopenic purpura may be
dependent on such factors as platelet age (greater temporally classified into three categories: acute,
functional capacity of young platelets,9294 or mega- intermittent, and chronic.
thrombocytes95) antiplatelet antibody binding leading The acute variety most often occurs in children,
to functional defects,8’ and capillary vessel integrity. particularly following a seasonal (winter-spring) viral
The type of bleeding is predominantly dermal and illness96 in 50% of patients (or, rarely, following a
mucosal. Bleeding into the skin with pin-point hemor- vaccination against a viral illness.’33) These include
rhage (pctechiae), particularly in dependent areas typical childhood infections and upper respiratory
where there is increased capillary pressure (i.e., lower infections. The average interval between purpura and
extremities), is often the rule. Petechiae are viola- preceding infection is 2 wk.9#{176}Of interest is the recent
ceous, flat, round, intradermal, and submucous report of high levels of bound platelet IgG in this
hemorrhages that later turn blue and yellow. The disease, which the authors postulate represents Ag-Ab
petechial purpura may become confluent and present complexes.98 The disease has an average duration of
as ecchymoses. Common areas of hemorrhage are in 1-2 mo (usually less than 6 mo). The male:female
the nasal, buccal, gastrointestinal, and vaginal muco- ratio is I 3.89.90,96.97 Although the acute disease is
sa. Examination of the mouth often reveals violaceous usually considered a milder disorder with a better
hemorrhagic blebs, 0.25-0.5 cm in diameter, on the prognosis than the chronic disorder, 27 deaths have
buccal and glossal mucosa with bleeding gums. been reported out of a total of 709 cases not treated
Conjunctival and retinal hemorrhages may be noted. with steroids.89#{176}’9699 Approximately 7%-28% of chil-
In chronic ATP or ITP, there is often a history of easy dren go on to develop the chronic variety.8990’96
bruising in the absence of trauma, frequent epistaxes, The chronic form most often occurs in adults, is
and prolonged menses. If the thrombocytopenia is persistent, lasting years to indefinitely, and has a
chronic, gastrointestinal blood loss may ensue, and female:male ratio of 2_4:l.388971b01 The so-called
iron deficiency anemia with microcytic hypochromic remissions of the chronic form are remissions of
red blood cells may be evident on peripheral smear. purpura only. The platelet count usually remains at
The most serious complication and one which is one-third to one-half the normal value.97 Some
rarely seen (particularly when steroids are employed) patients have been followed with the chronic disease
is hemorrhage into the central nervous system. This for 30 yr.97
can be fatal89 and must be rapidly and vigorously The intermittent variety may occur in childhood or
prevented by suitable therapy when there is purpura adulthood and may by spaced by intervals free from
and signs or symptoms of CNS bleeding, such as disease, wherein both the platelet count and platelet
retinal hemorrhages, convulsions, meningismus, head- survival are normal.’#{176}2 The usual pattern is for the
ache, change in personality, or more specific neuro- platelet count to return to normal in less than 6 mo,
logic signs. and remain normal for at least 3 mo prior to the next
Laboratory findings reveal a prolongation of the episode.#{176} As many as 5 recurrences have been
bleeding time, a positive tourniquet test, an absent to reported over a 4-1 8-yr interval.’02
 From www.bloodjournal.org by on October 7, 2010. For personal use only.

AUTOIMMUNE THROMBOCYTOPENIC PURPURA 337

POSSIBLE INTERRELATIONSHIP BETWEEN phagocytes to digest intracellular particles and destroy

ATP AND SLE bacteria,’#{176} Steroids inhibit erythrophagocytosis by
ATP may be an early manifestation of SLE in 07 adherence of antibody-coated plate-
3%-16% of patients.88”#{176}3104 In a series of 34 patients, lets to granulocytes,’#{176}8 and phagocytosis of platelets by
the median time for the development of SLE post- granulocytes. ‘#{176} Steroids inhibit chemotaxis of mono-
splenectomy was 2.5 yr. with a range of 3 mo to I 0 cytes”#{176} as well as monocyte receptors for lgG and
yr.’#{176}5Data supporting the similarity of these two C3.” There is indirect evidence that steroids inhibit
autoimmune disorders (chronic ATP and SLE) can be binding of autoimmune antibody to red blood cells”2
obtained from the young female preponderance, the as well as platelets.3#{176} There is controversial evidence in
presence of antiplatelet antibody, and the high mci- man as to whether steroids in the dosages currently

dence of DRw28587 in both disorders. Indeed, the employed significantly inhibit antibody production to
claim has been made that splenectomy for ATP accel- a previously recognized antigen. ‘#{176}“ ‘ 2 4 However,
erates the expression of SLE or activates a “dormant” steroids have been shown to decrease the levels of free
form of the disease.’#{176}4This claim has been challenged, and cell-bound autoimmune anti-red blood cell anti-
however,88”05 and is not accepted by most workers in body in man following 4-90 days of treatment with 60
the field. mg/day ofprednisone.”2 And, with high-dose methyl-
prednisolone, 96 mg for 3-5 days, a significant 22%
TREATMENT AND DIAGNOSIS decrease in serum lgG has been reported in 86% of
The classical treatment for autoimmune thrombo- volunteers 2-4 wk later.”6 Steroids however, do
cytopenic purpura is palliative, not curative, and is increase the catabolic rate ofcirculating IgG.”4 16

directed toward blockage or removal of a major site of Prednisone at 40 mg/day can inhibit the response to
platelet destruction, the spleen. delayed hypersensitivity.”7 118 Steroids can enhance
capillary resistance.”9’22 The effect of steroids is
Treatment With Steroids usually noted within the first 24-48 hr7”23 if sufficient
Administration of steroids is usually successful in steroids are employed, i.e., prednisone, I mg/kg.
either raising the platelet count to safe levels or normal However, a delay of 3-4 days may sometimes be
levels and is often dose-dependent. It is equivalent to a required to effectively inhibit splenic sequestration.7
“medical splenectomy” in that it prevents sequestra- In some situations, a “response to steroids” may not be
tion of damaged or antibody-coated platelets by the noted until I or more weeks of therapy. An effective
spleen (Fig. 3). Steroid administration to normal mdi- response to steroids (i.e., platelet count rising to
viduals has no effect on platelet survival or platelet > 100,000/cu mm) occurs in 36%_44%I659l124I25126
count.7 Steroids are known to reduce the ability of of all patients treated. Raising the steroid dosage to
I .5-2 mg/kg’27”#{176}’ will have some effect on the platelet
RESPONSES TO DIFFERENT DOSES OF ONE ITP PLASMA
count (although not necessarily to an effective level) in
70%-80% of patients. The sole purpose of steroid
implementation is to maintain the platelet count at a
z
0
C-)
safe level, i.e., >50,000/cu mm. There is little value in
C-
Lu attempting to bring the platelet count to normal levels.
-I
LU
C- This is often associated with prohibitive levels of
.
-A
steroids, which lead to the well known deleterious side
-A
.
F-
effects of hypercorticism. There is no available
evidence that long-term steroid therapy has any effect
F-
z on the natural history of the disease (following cessa-
LU
C-)
tion of therapy).#{176}’96 On the contrary, there is evidence
that long-term large doses of prednisone can occasion-
ally cause thrombocytopenia.’28

Indicationsfor Splenectomy
Fig. 3. Comparison of response of normal and splenectomized
persons to ITP factor. The graph on the left shows the effects of A favorable response to moderate steroid dosage, I
ITP plasma in a normal individual. The graph on the right shows
the effects of the same ITP plasma in a splenectomized person.
mg/kg, is probably an indication that the spleen is the
Note that the ITP plasma dose that did not produce thrombocyto- major site for platelet destruction and that the patient
penia in the splenectomized person was greater than the dose will benefit from splenectomy. Harrington et al. have
that produced marked thrombocytopenia in the normal individual.
(Reproduced by permission of Transactions of the Association of
reported a success rate of I 00% in such steroid-
American Physicians.) responsive splenectomized patients.6 However, the
 From www.bloodjournal.org by on October 7, 2010. For personal use only.

338 SIMON KARPATKIN

duration of follow-up has not been reported. Further- that the quantity of antiplatelet factor determines the
more, Doan et al.88 have noted a 79% response to site of sequestration. Splenic sequestration was
splenectomy in patients who failed to respond to extremely sensitive to low concentrations of antiplate-
steroids. Splenectomy removes the potential site of let factor following in vivo infusion of “ITP” plasma
destruction of damaged platelets,3”29 as well as a into volunteer recipients,’29”32 as was platelet survival
significant source of antiplatelet antibody produc- in nonsplenectomized, isoimmunized dogs.’27 Higher
tion.4’43’’54 After splenectomy, platelet survival may concentrations of antiplatelet antibody infused in vivo
return towards normal within 65 hr to 5-8 days.’27 The into volunteer recipients resulted in hepatic sequestra-
antiplatelet antibody, however, is still present in the tion.’32
plasma3’4’9’2’ of patients with ATP, despite an apparent The above observations indicate that splenectomy is
clinical remission, as well as on the platelets of some in order for all chronic cases of ATP, i.e., >6 mo
individuals.38’39 The antiplatelet antibody can be duration, regardless of the site of sequestration or
demonstrated in vivo by passive transfer to normal response to steroids (if they cannot be maintained free
recipients3 and by the classical occurrence of neonatal of purpura on 5-15 mg prednisone/day). This is
thrombocytopenia in infants of mothers with splenec- further supported by the clinical observations that
tomy-induced remission from ATP. It is of interest unsuccessfully splenectomized patients sometimes
that splenectomy in a normal individual or animal has respond better to steroids’27 and/or immunosuppres-
no effect on platelet survival.7’””30 The response of the sive therapy,’27’35’3’ and that splenectomy makes
platelet count to splenectomy is usually immediate available to the active platelet pool the 40% of plate-
with a peak obtained between the first and second lets normally sequestered in the spleen.’39 These plate-
week, In one study, all 15 of 22 patients who lets are enriched with megathrombocytes,’4#{176} which
responded to splenectomy had a peak platelet count aggregate better in vitro than other platelets.95
>500,000/cu mm, while those who did not have a
remission had a peak below this level.96 Approximately Risk ofinfection and Death Postsplenectomy

109o-l2% of patients will relapse following initially The hazard of severe infection (particularly pneu-
successful splenectomy,88”24 the majority within the mococcal) and rapid death in splenectomized infants
first year, but some as long as 5 yr later.’8 and children has been recognized for the past 20
Shulman et al.’29 have shown that six times more yr.’4’ “ This has been associated with disseminated
antiplatelet factor (obtained from a patient with ITP intravascular coagulation and Waterhouse- Frider-
and infused into a volunteer recipient) is required to ichsen syndrome. It has been argued that only children
significantly lower the platelet count in a splenecto- with associated primary disease, such as thalassemia
mized recipient compared to a nonsplenectomized major, Wiskott-Aldrich syndrome, histiocytosis, portal
recipient (Fig. 3). This indicates that other reticuloen- hypertension, etc., are at greater risk to infection and
dothelial organs, such as the liver, although less sensi- death.’43 What has not been fully appreciated is the
tive as a “sieve,”
sensitized
sponse to
are
antibody-coated
splenectomy
still capable

varies
of destroying

from
The
3 ‘ “ 32

50%-85%
highly
re-
in
accumulative
adults.
adult
‘-
patients
evidence
Whittaker””
with severe
of similar
retrospectively
pneumococcal
hazards
studied
infection
in
77

reported 3988900024,1 33 However, Aster has (meningitis or bacteremia). Eight of these patients
suggested that hepatic and splenic sequestration can ( I 0%) had absent splenic function (seven had splenec-
vary in individual patients with the course and severity tomies and one had splenic atrophy). Six patients died.
of the disease’32.’34 Hepatic sequestration was found in Reasons for splenectomy varied from trauma (2
patients who were more severely affected, e.g., lower patients), incidental to surgery (2 patients), acquired
platelet counts and shorter platelet survivals.’34 There hemolytic anemia, spherocytic hemolytic anemia,
is some experimental evidence in animals that a autoimmune thrombocytopenic purpura, and thalas-
compensatory hepatic sequestration follows splenecto- semia major. The onset of pneumococcal sepsis post-
my,’3’ reaching its maximum 6-8 wk postsplenectomy. splenectomy varied from I .5 to I 4 yr and averaged 5
This was associated with increased erythrophagocyto- yr. Hemostatic defects were observed in 7 patients,
sis by Kupffer cells following treatment with damaged and in 4, the diagnosis of disseminated intravascular
red blood cells.’3’ Indeed, patients who relapse follow- coagulation could be confirmed at autopsy: hemor-
ing “successful” steroid response can be shown to have rhagic diathesis, fibrin thrombosis of renal glomerular
developed hepatic Nevertheless, Na- capillaries and adrenal sinusoids, focal ischemic and
jean et al.59 have reported a 30% response to splenec- hemorrhagic lesions of the adrenal cortex, kidneys,
tomy in patients with exclusively or predominantly liver, gastrointestinal mucosa, ovary, choroid plexus,
hepatic sequestration. Shulman,’29 Baldini,’27 and and myocardium, including 2 patients with the Water-
Aster and Jandl’32 have presented evidence suggesting house-Friderichsen syndrome. It is of interest that 14
 From www.bloodjournal.org by on October 7, 2010. For personal use only.

AUTOIMMUNE THROMBOCYTOPENIC PURPURA 339

of 1 9 cases of Waterhouse-Friderichsen syndrome refractory to steroids and/or steroids and splenectomy

reported in the literature and associated with pneumo- is less enthusiastic. We have experienced the rare
coccal sepsis had absent spleens or splenic atrophy.’ patient with prompt and dramatic recovery; as well as
In contrast, of 523 patients with nonpneumococcal the more common situation where there is a small rise
infection, only 1 patient had an absent spleen.’44 The in the platelet count of 10-40,000/cu mm, with usual
findings strongly suggest that the spleen is a main line relapse several days later. Side-effects of peripheral
of defense in the handling of pneumococcal sepsis as neuropathy and intense bone (particularly jaw) pain
well as disseminated intravascular coagulation. The are not uncommon. Ahn and coworkers have recently
former may operate via the ability of the spleen to introduced a new approach for the delivery of ymca
produce antipneumococcal opson i ns ‘ 50. S as well as act alkaloids to those patients.’55 This consists of the in
as a filter; the latter presumably acts via the clearance vitro incubation of the patient’s platelets (or donor
of activated coagulation factor products, since reticu- platelets) with ymca alkaloids, prior to the infusion of
loendothelial blockade or splenectomy enhances the I U of ymca alkaloid (vinblastine)-loaded platelets
hypercoagulable state.’52 into the patient. The patient’s circulating platelet
The splenectomized patient should be carefully antibody presumably reacts with these platelets and
scrutinized when developing any fever or infection thereby serves to deliver vinblastine-loaded platelets to
because of the rapidly fulminant nature of pneumo- the RE system. The RE cells that ingest the vinblas-
coccal sepsis and death in these patients. tine-loaded platelets are presumably “killed” by the
vinblastine. They have treated I I patients who were
refractory to intravenous vincristine, as well as sple-
Treatment ofPatients Refractory to Steroids
nectomy, steroids, and other immunosuppressive
and Splenectomy agents, with 1-4 such infusions over a 2-wk period
The role of extended immunosuppressive therapy (each containing 2-5 mg of vinblastine). Six patients
for the treatment of ATP (particularly azathioprine), have gone into remission; 3 of these relapsed and 3
although theoretically sound, has not proved over- patients have remained in remission for 5-10 mo. Of
whelmingly successful in the few patients stud- the remaining 5 patients, 3 were partial responders
ied.’26”27”35 38 Its use should be reserved for those and 2 failed to respond. Side effects included alopecia,
patients who are refractory to steroids and splenecto- mild confusion, and intense jaw pain. Similar studies
my. A controlled study in this regard is lacking. have recently been reported by another group’56 who
Twelve of 36 documented cases have had an excellent noted a complete remission in 2 of 10 patients (7-mo
response to immunosuppressive therapy. ‘ 2b. 35 Cytoxan duration). There were 6 transient responses (2-3 wk
has also been used for the treatment of refractory duration) and 2 complete failures. Of interest was
patients with possibly better success than azathio- their observation that bound antiplatelet antibody
prine.’36”53 The disadvantage of cytoxan as well as disappeared from the two patients who went into
azathioprine is the approximately 2-mo duration remission, with a T’/2 of 15 days, whereas it did not
required to obtain an effect. Cytoxan also has the disappear from the treatment failures. It should be
undesirable side-effect of hemorrhagic cystitis. The stressed that this interesting approach is still experi-
alkaloid, vincristine, has recently been employed for mental. The long-term side-effects of continued ymca
the treatment of such refractory patients by Ahn and alkaloid treatment, as well as RE damage has as of yet
coworkers,’54 with benefit in 6 of 7 patients refractory not been evaluated.
to steroids with intact spleens, 9 of I 3 patients refrac- It cannot be overemphasized that the goal of ther-
tory to both steroids and splenectomy, and I 0 of I 0 apy should not necessarily be directed toward restora-
patients with SLE and thrombocytopenia. Follow-up tion of a normal platelet count, but toward the restora-
studies have revealed that 4 of the original I 3 patients tion of a sufficient number of functioning platelets to
are in prolonged remission (>6 yr). Five patients maintain hemostasis for normal daily activity. This
require the continued use of ymca alkaloids, and four can generally be achieved with a platelet count of
patients are treatment failures. The drug is unique in 50-80,000/cumm. It is rare for patients to develop
that it is both immunosuppressive and thrombocytotic, purpura in this range if the defect is purely quantita-
and when effective, acts promptly (i.e., 7-10 days). tive. Indeed, the claim has been made that only young
Our own limited experience with vincristine in patients platelets are hemostatically effective.94

REFERENCES

I Robson HN, Davidson LSP: Purpura in pregnancy with 2. Epstein RD. Lozner EL, Coffey TS. Davidson CS: Congenital
special reference to idiopathic thrombocytopenic purpura. Lancet thrombocytopenic purpura. Purpura hemorrhagica in pregnancy
2:164, 1950 and in the newborn. Am J Med 9:44, 1950
 From www.bloodjournal.org by on October 7, 2010. For personal use only.

340 SIMON KARPATKIN

3. Harrington Wi, Minnich V. Arimura G: The autoimmune 26. Hirschman Ri, Gralnick HR: A simplified platelet factor 3
thrombocytopenias, in Tocantins LM (ed): Progress in Hematology, (PF-3) assay for the rapid detection of platelet isoantibodies and an
vol 1, New York, Grune & Stratton, 1956, p 166 anti-platelet factor in ATP and SLE. J Lab Clin Med 84:292, 1974
4. Karpatkin S. Strick N, Karpatkin MH, Siskind GW: Cumula- 27. Aster RH, Enright SE: A platelet and granulocyte
Live experience on the detection of anti-platelet antibody in 234 membrane defect in paroxysmal nocturnal hemoglobinuria; useful-
patients with idiopathic thrombocytopenic purpura, systemic lupus ness for the detection of platelet antibodies. J Clin Invest 48:1 199,
erythematosis and other clinical disorders. Am J Med 52:776, 1972 I969
5. Evans RS, Takahaski K, Duane AB, Payne R, Liu C: Primary 28. Hirschman RJ, Shulman NR: The use of platelet serotonin
thrombocytopenic purpura and acquired hemolytic anemia; release as a sensitive method for detecting anti-platelet antibodies
evidence for a common etiology. Arch Intern Med 87:48, 1951 and a plasma anti-platelet factor in patients with idiopathic throm-
6. Harrington Wi, Arimura G: Immune reactions of platelets, in bocytopenic purpura. Br i Haematol 24:793, 1973
Johnson SA, Monto RW, Rebuck JW, Horn RJ, Jr (eds): Henry 29. McMillan R, Smith RS, Longmire RL, Yelenosky R, Reid
Ford Hospital Symposium: BlOOd Platelets, Boston, Little Brown, RT, Craddock CG: Immunoglobulin associated with human plate-
1961, p659 lets. Blood 37:316, 1971
7. Shulman NR, Marder VJ, Weinrach RS: Similarities between 30. Tate DY, Sorenson RL, Gerrard JM, White JG, Krivit W:
known anti-platelet antibodies and the factor responsible for throm- An immunoenzyme histochemical technique for the detection of
bocytopenia in idiopathic purpura. Physiologic. serologic and platelet antibodies from the serum of patients with idiopathic
isotopic studies. Ann NY Acad Sci 124:499, 1965 (autoimmune) thrombocytopenic purpura. Br i Haematol 37:265,
8. Van de Wiel WM, Van de Wiel-Dorfmeyer H, Van Loghem I 977
JJ: Studies on platelet antibodies in man. Vox Sang 6:641, 1961 31. Handin RI, Stossel TP: Phagocytosis of antibody-coated
9. Tullis JL: Identification and significance of platelet antibod- platelets by human granulocytes. N EngI J Med 290:989, 1974
jes. N EngI J Med 255:541, 1956 32. Karpatkin 5, Garg 5K, Siskind GW: Autoimmune thrombo-
10. Stefanini M, Mdc RH: The significance of platelet antibod- cytopenic purpura and the compensated thrombocytolytic state. Am
ies with special reference to platelet agglutinins. Acta Hematol J Med 51:1, 1971
20:195, 1958 33. Rosse W: Annotation: Platelet antibody in autoimmune
I I . Dausett J, Cohn M, Colombani J: Immune platelet isoanti- thrombocytopenia. Br i Haematol 3 1 : I 29, 1975
bodies. Vox Sang 5:4, 1960 34. Dixon R, Rosse W, Ebbert L: Quantitative determination of
12. Corn M, Upshaw JD Jr: Evaluation of platelet antibodies in antibody in idiopathic thrombocytopenic purpura. N EngI J Med
idiopathic thrombocytopenic purpura. Arch Intern Med 109:85, 292:230, 1975
1962 35. Hedge UM, Gordon-Smith EC, Worlledge 5: Platelet anti-
13. Jackson DP, Schmid Hi, Zieve PD, Levin i, Conley CL: bodies in thrombocytopenic purpura. N EngI J Med 35:113, 1977
Nature of a platelet-agglutinating factor in serum of patients with 36. Karpatkin S. Friedman E, Siskind GW: Studies on antibody
idiopathic thrombocytopenic purpura. i Clin Invest 42:383, 1963 bound to platelets in autoimmune thrombocytopenic purpura
14. Jacklin HN, Furth FW, Lozner EL: Studies on the causes of (ATP). Clin Res 25:476A, 1977
false-positive results in the laboratory detection of platelet agglu- 37. Luiken GA, McMillan E, Lightsey AL, Gordon P. Zevely 5,
tinins in thrombocytopenic purpura. Am J Clin Pathol 36:95, 1961 Shulman I, Gribble Ti, Longmuire RL: Platelet associated IgG in
I 5. Fluckiger P. Hassig A, Koller F: The technique of the immune thrombocytopenic purpura. Blood 50:317,1977
platelet Coombs test. Acta Hematol 12:339, 1954 38. Cines DB, Schreiber AD: Immune thrombocytopenia. Use of
16. Steffen C: Results obtained with the antiglobulin consump- Coombs antiglobulin test to detect lgG and C3 on platelets N Engl i
Lion test and investigations of autoantibody eluted in immunohema- Med 300:106, 1979
tology. i Lab Clin Med 55:9, 1960 39. Hymes K, Shulman 5, Karpatkin 5: A solid-phase radioim-
17. Dausset J: Antigenes et anticorps antiplaquettes. Presented munoassay for bound anti-platelet antibody. Studies on 45 patients
at the VII International Congress on Hematology, Rome, 1958 with autoimmune platelet disorders. J Lab Clin Med 94:639, 1979
18. Moulinier i: Le test de consommation d’antiglobuline appli- 40. Garg SK, Amorosi EL, Karpatkin 5: Use of the megathrom-
que a Ia recherche des anticorps anti-thrombocytes. Sang 26:8 1 1, bocyte as an index of megakaryocyte number. N EngI J Med
I955 284:11, 1971
19. Dausett J: Les thrombo-anticorps. Acta Hematol 20:185, 41 . Kelton JG, Neame PB, Bishop J, AIi M, Gauldie i, Hirsch J:
I958 The direct assay for platelet-associated IgG (PAIgG): Lack of
20. Karpatkin 5, Siskind GW: In vitro detection of platelet association between antibody and platelet size. Blood 53:73, 1979
antibody in patients with idiopathic thrombocytopenic purpura and 42. Kelton JG, Neame PB: Elevated platelet associated lgG in
systemic Iupus erythematosus. BlOOd 33:795, 1969 the thrombocytopenia ofsepticemia. N Engl J Med 300:760, 1971
21. Karpatkin M, Siskind GW, Karpatkin S: The platelet factor 43. Karpatkin 5, Strick N, Siskind GW: Detection of splenic
3 immunoinjury technique reevaluated. Development of a rapid test anti-platelet antibody synthesis in idiopathic autoimmune thrombo-
for anti-platelet antibody. Detection in various clinical disorders, cytopenic purpura (ATP). Br J Haematol 23:167, 1972
including immunologic drug-induced and neonatal thrombocytope- 44. McMillan R, Longmire RL, Yelenosky R, Smith RS, Crad-
nias. J Lab Clin Med 89:400, 1977 dock CG: Immunoglobulin synthesis in vitro by splenic tissue in
22. Karpatkin 5, Siskind GW: Studies on the specificity of idiopathic thrombocytopenic purpura. N EngI J Med 286:681, 1972
anti-platelet autoantibodies. Proc Soc Exp Biol Med 147:71 5, 1974 45. Lurhuma AZ, Riccomi H, Masson PL: The occurrence of
23. Murphy 5, Ziegler Z, Bennett 5, Myers AR, Gottlieb A, circulating immune complexes and viral antigens in idiopathic
Gardner FH: Detection of anti-platelet factors with a platelet thrombocytopenic purpura. Clin Exp lmmunol 28:49, 1977
immunoinjury assay. Am Soc Hematol 14: 1 15, 1971 46. Piessens WF, Wybran J, Manaster J, Strijckmans PA:
24. Cowan DH: Personal communication Lymphocyte transformation induced by autologous platelets in a
25. Wilkins Ri, Hurvitz AL, Dodds-Laffin Wi: Immune- case of thrombocytopenic purpura. Blood 36:421 , 1970
mediated thrombocytopenia in the dog. J Am Vet Med Assoc 47. Wybran i, Fudenberg HH: Cellular immunity to platelets in
163:277, 1973 idiopathic thrombocytopenic purpura. Blood 40:856, 1972
 From www.bloodjournal.org by on October 7, 2010. For personal use only.

AUTOIMMUNE THROMBOCYTOPENIC PURPURA 341

48. Clancy R: Cellular immunity to autologous platelets and malities in idiopathic thrombocytopenic purpura. N Engl J Med
serum-blocking factors in idiopathic thrombocytopenic purpura. 286:622, 1972
Lancet 1:6, 1972 72. Regan MG, Lackner H, Karpatkin 5: Platelet function and
49. Handin RI, Piessens WF, Moloney WC: Stimulation of coagulation profile in lupus erythematosus. Ann Intern Med 81:462,
nonimmunized lymphocytes by platelet antibody complexes in idio- I 974
pathic thrombocytopenic purpura. N EngI J Med 289:714, 1973 73. Lackner H, Karpatkin 5: On the easy brusing’ syndrome
50. Quagliata F, Karpatkin 5: Impaired lymphocyte transforma- with normal platelet count. A study of 75 patients. Ann Intern Med
tion and capping in autoimmune thrombocytopenic purpura. Blood 83:190, 1975
46:1022, 1975 74. Weiss Hi: Platelet aggregation, adhesion and adenosine
51. Firkin BG, Wright R, Miller 5, Stokes E: Splenic macro- diphosphate release in thrombopathia (platelet factor 3 deficiency).
phages in thrombocytopenia. BlOOd 33:240, 1969 Ami Med43:570, 967
52. Summerall JM, Gibbs WN: Splenic histiocytosis associated 75. O’Brien JR: Platelets: A Portsmouth syndrome? Lancet
with thrombocytopenia. Acta Haematol 48:34, 1972. 2:258, 1967
53. Tavassoli M, McMillan R: Structure of the spleen in idio- 76. Hardisty RM, Hutton RA: Bleeding tendency associated
pathic thrombocytopenic purpura. Am i Clin Pathol 64:180, 1975 with ‘new’ abnormality of platelet behavior. Lancet I :983, I 967
54. Zucker-Franklin D, Karpatkin 5: Red cell and platelet frag- 77. Hirsh J, Castelan Di, Loder PB: Spontaneous bruising asso-
mentation in idiopathic autoimmune thrombocytopenic purpura. N ciated with a defect in the interaction of platelets with connective
EngI i Med 297:517, 1979 tissue. Lancet 2:18, 1967
55. Cohen P. Gardner RH, Barnett GO: Reclassification of the 78. Caen JP. Sultan T, Larricu M: A new familial platelet
thrombocytopenias by the Cr5’-labeling method for measuring disease. Lancet 1:203, 1968
platelet life span. N EngI J Med 264: 1 294, 1961 79. Holmsen I-I, Weiss Hi: Further evidence for a deficient
56. Harker LA: Thrombokinetics in idiopathic thrombocytopenic storage pool of adenine nucleotides in platelets from some patients
purpura. Br J Haematol 19:95, 1970 with thrombocytopathia-Storage pool disease.’ Blood 39: I 97.
57. Branehog I, Kutti i, Ridell B, Swolin B, Weinfeld A: The I972
relation of thrombokinetics to bone marrow megakaryocytes in 80. Zahavi J, Marder Vi: Acquired ‘storage pool disease’ of
idiopathic thrombocytopenic purpura. Blood 45:551, 1975 platelets associated with circulating anti-platelet antibodies. Am i
58. Ries CA, Price DC: Cr5’ platelet kinetics in thrombocytope- Med 56:883, 1974
nia. Correlation between splenic sequestration of platelets and 81. Karpatkin 5, Lackner HL: Association of antiplatelet anti-
response to splenectomy. Ann Intern Med 80:702, 1974 body with functional platelet disorders. Am J Med 59:599, 1975
59. Najean Y, Ardaillou N, Dresch C, Bernard i: The platelet 82. Karpatkin S. Garg 5K, Siskind GW: Autoimmune thrombo-
destruction site in thrombocytopenic purpuras. Br J Haematol cytopenic purpura and the compensated thrombocytolytic state. Am
13:409, 1967 J Med 51:1, 1971
60. Branehog I, Kutti i, Weinfeld A: Platelet survival and 83. Ten Cate iW, Dc Vries SI, Sixma ii, Van Berkel W:
platelet production in idiopathic thrombocytopenic purpura. Br J Defective collagen-induced platelet aggregation in a normal popula-
Haematol 27:127, 1974 tion. Thromb Diath Haemorrh 25:234, 1971
61. Dameshek W, Miller EB: The megakaryocytes in idiopathic 84. Hardisty RM, Hutton RA: Bleeding tendency associated
thrombocytopenic purpura, a form of hypersplenism. Blood I :27, with ‘new’ abnormality of platelet behavior. Lancet I :983, 1967
I 946 85. Karpatkin 5, Fotino M, Gibofsky A, Winchester Ri: Asso-
62. Troland CE, Lee FC: Thrombocytopen. A substance in the ciation of HLA-DRw2 with autoimmune thrombocytopenic purpu-
extract from the spleen of patients with idiopathic thrombocyto- ra. J Clin Invest 63:1085, 1979
penic purpura that reduces the number of blood platelets. JAMA 86. Reinertsen iL, Kippel iH, iohnson AH, Steinberg AD,
111:221, 1938 Decker JL, Mann DL: B lymphocyte alloantigens associated with
63. Cronkite EP: Further studies on platelet reducing substances systemic lupus erythematosus. N EngI i Med 299:5 1 5, 1978
in splenic extracts. Ann Intern Med 20:52, 1944 87. Gibofsky A, Winchester Ri, Hansen J, Patarroyo M, Dupont
64. Tarnuzi A, Smiley RK: Hematologic effects of splenic B, Paget S. Lahita R, Halper i, Fotino M, Yunis E, Kunkel HG:
implants. Blood 29:373, 1967 Contrasting patterns of histocompatibility determinants in patients
65. Vasquez ii, Lewis JH: Immunocytochemical studies on with rheumatoid arthritis and systemic lupus erythematosus.
platelets. The demonstration of a common antigen in human plate- Arthritis Rheum 21:5133, 1978
lets and megakaryocytes. Blood 16:968, 1960 88. Doan CA, Bouroncle BA, Wiseman BK: Idiopathic and
66. Humphrey JH: Origin of blood platelets. Nature 176:38, secondary thrombocytopenic purpura: Clinical study and evaluation
I 955 of 381 cases over a period of 28 years. Ann Intern Med 53:861,
67. Silber R, Bentez R, Eveland WC, Akeroyd JH, Dunne Ci: I 960
The application of fluorescent antibody methods to the study of 89. Komrower GM, Watson GH: Prognosis in idiopathic throm-
platelets. Blood 16:958, 1960 bocytopenic purpura ofchildhood. Arch Dis Child 29:502, 1954

68. McMillan R, Luiken GA, Levy R, Yelenosky R, Longmire 90. Choi SI, McClure PD: Idiopathic thrombocytopenic purpura
R: Platelet antibody against megakaryocytcs in idiopathic thrombo- in childhood. Can Med Assoc J 97:562, 1967
cytic purpura. JAMA 239:2460, 1978 9 1 . Haddy TB: Idiopathic thrombocytopenic purpura in children:
69. Rolovic Z, Baldini M, Dameshek W: Megakaryocytopoiesis An analysis of 25 cases. Lancet 85:465, 1965
in experimentally induced immune thrombocytopenia. Blood 92. Karpatkin 5: Heterogeneity of human platelets. II. Func-
34:173, 1970 tional evidence suggestive of young and old platelets. i Clin Invest
70. Garg SK, Lackner H, Karpatkin 5: The increased percentage 48:103, 1969
of megathrombocytes in various clinical disorders. Ann Intern Med 93. Hirsch i, Glynn MF, Mustard iF: The effect of platelet age
77:361, 1972 on platelet adherence to collagen. i Clin Invest 47:466, 1968
7 1 . Clancy R, Jenkins B, Firkin B: Qualitative platelet abnor- 94. Shulman NR, Watkins SP Jr. Itscoitz SB, Students AB:
 From www.bloodjournal.org by on October 7, 2010. For personal use only.

342 SIMON KARPATKIN

Evidence that the spleen retains the youngest and hemostatically 3 or 5 days of high doses of methyl prednisolone. J Clin Invest
most effective platelets. Trans Assoc Am Physician 81 :302, 1968 52:2629, 1973
95. Karpatkin 5: Heterogeneity of human platelets. VI. Correla- I 17. Long JB, Favour CB: The ability of ACTH and cortisone to
tion of platelet function with platelet volume. Blood 5 1 :307, 1978 alter delayed type bacterial hypersensitivity. Bull Johns Hopkins
96. Lusher iM, Zuelzer WW: Idiopathic thrombocytopenic Hosp87:l86, 1950
purpura in childhood. J Pediatr 68:971, 1966 I 18. Bovornkitti S. Kansadal P. Sathirapat P. Oonsombatti P:
97. Hirsch EO, Dameshek W: Idiopathic thrombocytopenia: Reversion and reconversion rate of tuberculin skin reactions in
Review of 89 cases with particular reference to the differentiation correlation with the use of prednisone. Dis Chest 38:5 1 , 1960
and treatment of acute (self-limited) and chronic types. Arch Intern 1 19. Robson HN, Duthie J Jr: Capillary resistance and adreno-
Med88:70l, 1951 cortical activity. Br Med J 2:971, 1950
98. Lightsey AL, Koenig HM: Platelet associated immunoglobu- I 20. Faboon WW, Greene RW, Lozner EL: The hemostatic
lin G in childhood idiopathic thrombocytopenic purpura. J Pediatr defect in thrombocytopenia as studied by the use of ACTH and
94:20k, 1979 cortisone. Am J Med 13:12, 1952
99. Watson-Williams Ei, MacPherson AIS, Davidson Sir S: The I 21 . Stefanini H, Martino NB: Use of prednisone in the manage-
treatment of idiopathic thrombocytopenic purpura. A review of 93 ment of some hemorrhagic states. N Engl J Med 254:3 1 3, 1956
cases. Lancet 2:221, 1958 I 22. Kitchens CS: Amelioration of endothelial abnormalities by
100. Meyers MC: Results of treatment in 71 patients with prednisone in experimental thrombocytopenia in the rabbit. J Clin
idiopathic thrombocytopenic purpura. Am J Med Sci 242:295, Invest 60:1 129, 1977
I 961 123. Jacobson BM, Sohier WD: The effects of ACTH and of
101. Dameshek W, Rubio F Jr. Mahoney iP, Reeves WH, cortisone on the platelets in idiopathic thrombocytopenic purpura.
Burgin LA: Treatment of idiopathic thrombocytopenic purpura. N N EngI J Med 246:247, 1952
Engli Med 269:647, 1963 124. Carpenter AF, Wintrobe MM, Fuller EA, Haut A,
102. Dameshek W, Ebbe S. Greenberg L, Baldini M: Recurrent Cartwright GE: Treatment of idiopathic thrombocytopenic purpura
acute idiopathic thrombocytopenic purpura. N Engl J Med 269:649, with prednisone. JAMA 166:1805, 1958
I963 I 25. Wilson SJ, Eisemann G: The effect of corticotropin
103. McClure PD: Idiopathic thrombocytopenic purpura in chil- (ACTH) and cortisone on idiopathic thrombocytopenic purpura.
dren. Diagnosis and management. Pediatrics 55:68, 1975 AmJ Med 13:21, 1952
104. Rabinowitz Y, Dameshek W: Systemic lupus erythemato- 126. Lo SS, Hitzig WH, Sigg P: Management of chronic idio-
sus after ‘idiopathic’ thrombocytopenic purpura: a review. A study pathic thrombocytopenic purpura in children, with particular refer-
of systemic lupus erythematosus occurring after 78 splenectomies ence to immunosuppressive therapy. Acta Hematol 4 1 : I , I 969
for ‘idiopathic’ thrombocytopenic purpura, with a review of the 127. Baldini M: Idiopathic thrombocytopenic purpura. N EngI
pertinent literature. Ann Intern Med 52:1, 1960 Med274:1245, 1966
105. Best WR, Darling DR: A critical look at the splenectomy- 128. Cohen P. Gardner FH: The thrombocytopenic effect of
SLE controversy. Med Clin North Am 46: 19, 1962 sustained high-dosage prednisone therapy in thrombocytopenic
106. Berenbaum MC: Immunosuppressive agents and allogeneic purpura. N EngI i Med 265:61 1, 1961
transplantation. i Clin Pathol (Suppl) 20:471, 1967 129. Shulman NR, Weinrach RS, Libre EP, Andrews HL: The
107. Greendyke RM, Bradley EM, Swisher SN: Studies of the role of the reticuloendothelial system in the pathogenesis of idio-
effects of administration of ACTH and adrenal corticosteroids on pathic thrombocytopenic purpura. Trans Assoc Am Physician
erythrophagocytosis. i Clin Invest 44:746, 1965 78:374, 1965
108. Verp M, Karpatkin S: Effect of plasma, steroids or steroid I 30. Hjort PF, Paputchis H: Platelet lifespan in normal, splenec-
products on the adhesion of human opsonized thrombocytes to tomized and hypersplenic rats. Blood I 5:45, 1960
human leukocytes. J LabClin Med 85:478, 1975 131. Jacob HS, MacDonald RA, Jandl JH: Regulation of spleen
109. McMillan R, Longmire RL, Tavassoli M, Armstrong S. growth and sequestering function. J Clin Invest 42:1476, 1963
Yelenosky R: In vitro platelet phagocytosis by splenic leukocytes in 132. Aster RH, iandl JH: Platelet sequestration in man. II.
idiopathic thrombocytopenic purpura. N EngI i Med 290:249, 1974 Immunological and clinical studies. J Clin Invest 43:856, 1964
I 10. Rhinehart ii, Balcerzak SP, Sagone AL and LoBuglio AF: 1 33. Dameshek W: Controversy in idiopathic thrombocytopenic
Effects of corticosteroids on human monocyte function. i Clin purpura. JAMA 173:123, 1960.
Invest 54:1337, 1974 I 34. Aster RH, Keene WR: Sites of platelet destruction in
I I 1. Schreiber AD, Parsons i, McDermott P. Cooper RA: Effect idiopathic thrombocytopenic purpura. Br i Hematol 16:61, 1969
of corticosteroids on the human monocyte lgG and complement 135. Hilgartner MW, Lanzkowsky PL, Smith CH: The use of
receptors. i Clin Invest 56:1 189, 1975 azathioprine in refractory idiopathic thrombocytopenic purpura in
I 12. Rosse WF: Quantitative immunology of immune hemolytic children. Acta Pediatr Scand 59:409, 1970
anemia. 11. The relationship ofcell-bound antibody to hemolysis and I 36. Laros RK, Penner iA: ‘Refractory’ thrombocytopenic
the effect of treatment. i Clin Invest 50:734, 1971 purpura treated successfully with cyclophosphamide. JAMA
113. Fischel EE, Vaughan JH, Photopoulos C: Inhibition of 215:445, 1971
rapid production of antibody by cortisone. Proc Soc Exp Biol Med 137. Sussman LN: Azathioprine in refractory idiopathic throm-
81:344, 1952 bocytopenic purpura. JAMA 202:259, 1967
114. Griggs RC, Condemi ii, Vaughan JH: Effect of therapeutic I 38. Bouroncle BA, Doan CA: Refractory idiopathic thrombocy-
dosages of prednisone on human immunoglobulin G metabolism. topenic purpura treated with azathioprine. N EngI J Med 275:630,
J LabClin Med49:267, 1972 1966
1 15. Wochner RD: Hypercatabolism of normal lgG; an unex- I 39. Aster RH: Splenic platelet pooling as cause of ‘hypersplen-
plained immunoglobulin abnormality in connective tissue diseases. J ic’ thrombocytopenia. Trans Assoc Am Physician 78:362, 1965
Clin Invest 49:454, 1970 140. Freedman ML, Karpatkin S: Heterogeneity of rabbit plate-
I 16. Butler WT, Rossen RD: Effects of corticosteroids on lets. V. Preferential splenic sequestration of megathrombocytes. Br
immunity in man. I. Decreased serum lgG concentration caused by J Haematol 31:255, 1975
 From www.bloodjournal.org by on October 7, 2010. For personal use only.

AUTOIMMUNE THROMBOCYTOPENIC PURPURA 343

141. King H, Shumacker HB Jr: Splenic studies. I. Susceptibility cal serum opsonizing activity in sickle-cell disease. N Engl J Med
to infection after splenectomy performed in infancy. Ann Surg 279:459, I968
136:239, 1952 151. Biggar WD, Bogart D, Holmes B, Good RA: Impaired
142. Erickson WD, Burgert EO Jr. Lynn HB: The hazard of phagocytosis of pneumococcus type 3 in splenectomized rats. J
infection following splenectomy in children. Am J Dis Child I 16:1, Reticuloendothel Soc 1 1 :77, 1972
I 968 152. Rabiner SF, Friedman LH. The role of intravascular
143. Eraklis AJ, Kevy SV, Diamond LK, Gross RE: Hazard of haemolysis and the reticulo-endothelial system in the production of
overwhelming infection after splenectomy in childhood. N Engl J a hypercoagulable state. Br i Haematol 14: 105, I 968
Med 276:1225, 1967
153. Finch SC, Castro 0, Cooper M, Covey W, Erichson R,
144. Whitaker AN: Infection and the spleen: association between
McPhedran P: Immunosuppressive therapy of chronic idiopathic
hyposplenism, pneumococcal sepsis and disseminated intravascular
thrombocytopenic purpura. Am i Med 56:4:1974
coagulation. Med J Aust 1:1213, 1969
154. Ahn YS, Harrington Wi, Seelman RC, Eytel CS: Vincris-
145. Lowdon AGR, Steward RHM, Walker W: Risk of serious
tine therapy of idiopathic and secondary thrombocytopenias. N
infection following splenectomy. Br Med J I :446, 1966
EngI i Med 291:376, 1974
146. Rosner F, Ritz ND: The defibrination syndrome. Arch
Intern Med I 17:17, 1966 155. Ahn YS, Byrnes ii, Harrington Wi, Cayer ML, Smith DS,
147. Bisno AL, Freeman JC: The syndrome ofasplenia, pneumo- Brunskill DE, Pall LM: The treatment of idiopathic thrombocyto-

coccal sepsis, and disseminated intravascular coagulation. Ann penic purpura with vinblastine-loaded platelets. N EngI i Med

Intern Med 72:389, 1970 298:1 101, 1978

148. Torres J, Bisno AL. 1-lyposplenism and pneumococcemia. I 56. Slichter Si, Schwartz K: Mechanism of action of vinblas-
Visualization of diplococcus pneumoniae in the peripheral blood tine-loaded platelets in the treatment ofchronic idiopathic thrombo-
smear. Am J Med 55:851, 1973 cytopenic purpura. Clin Res 27:307A, 1979
149. Cormia FE, Campos LT: Infections after splenectomy. Ann 157. McGrath KM. Stuart ii, Richards F II: Correlation
Intern Med 78:149, 1973 between serum lgG, platelet membrane IgG, and platelet function
150. Winkelstein JA, Drachman RH: Deficiency of pneumococ- in hypergammaglobulinaemic states. Br J Haematol 42:585, 1979