Haemophilia (2010), 16, 256262

DOI: 10.1111/j.1365-2516.2009.02122.x

ORIGINAL ARTICLE Clinical haemophilia

New early prophylaxis regimen that avoids immunological

danger signals can reduce FVIII inhibitor development
K. KURNIK,* C. B IDLINGMAIER,* W. ENGL, H. CHEHADEH, B. REIPERT and G. AUERSWALD
*Klinikum der Universitat Munchen, Dr. von Haunersches Childrens Hospital, Munich, Germany; Baxter Innovations
GmbH, Vienna, Austria; and Prof. Hess Childrens Hospital, Klinikum Bremen-Mitte, Bremen, Germany

Summary. The most problematic complication of

haemophilia A treatment is the development of
inhibitors to FVIII. The highest risk of developing
inhibitors is during the first 20 exposure days (EDs).
If the patient can be brought through this high risk
period without inhibitor development, the subsequent risk is low. Therefore, as a pilot project,
we developed a prophylaxis regimen for the first
2050 EDs specifically designed to induce tolerance
to the administered FVIII and to minimize inhibitor
development by avoiding immunological danger
signals. Twenty-six consecutive previously untreated
patients (PUPs) with severe haemophilia A were
treated with the new prophylaxis regimen and the
incidence of inhibitor development in this group was
compared with that in a historical control group of
30 consecutive PUPs treated with a standard joint
protection prophylaxis regimen (4050 IU kg)1,
three times a week). There were no significant

Introduction
Today, the most problematic and costly complication of the treatment of haemophilia A that remains
to be overcome is the development of inhibitory
antibodies (FVIII inhibitors) to FVIII replacement
therapy, particularly in previously untreated patients
(PUPs). It is now becoming clear that inhibitor
development is a complex, multi-factorial immune
response involving both patient-specific and treatment-related factors [13]. It has been shown that
Correspondence: Karin Kurnik, PD, MD, Klinikum der Universitat
Munchen, Dr. von Haunersches Childrens Hospital, D-80337
Munich, Germany.
Tel.: +49 89 5160 2811/2853; fax: +49 89 5160 4453;
e-mail: karin.kurnik@med.uni-muenchen.de
Accepted after revision 7 September 2009

256

differences between the study and control groups

in patient-related inhibitor risk factors such as
ethnicity (all Caucasian), severity of haemophilia
(all <1% FVIII), severity of FVIII gene mutation
(P < 0.0006) nor in some treatment-related factors
such as product type, age at first exposure, vaccination regimen or the need for surgery. 14 of 30
subjects given standard prophylaxis but only one of
the 26 subjects given the new regimen developed an
inhibitor (P = 0.0003, odds ratio 0.048, 95% CI:
0.0010.372). Our results indicate that minimizing
danger signals during the first 20 EDs with FVIII
may reduce the risk of inhibitor formation. These
results should be confirmed in a larger prospective
clinical study.
Keywords: danger theory, early prophylaxis, FVIII
inhibitors, haemophilia, immunological danger signals, PUPs

patients with severe defects in the FVIII gene, such as

large deletions, inversions (most commonly intron 22
inversion) and stop mutations are significantly more
likely to develop inhibitors than are those with more
minor defects such as missense mutations, small
deletions or insertions and splice site mutations [1].
Severe mutations in the FVIII gene are predicted to
cause a complete deficit of any endogenous FVIII
production. In these circumstances, FVIII cannot be
presented to the immune system during negative
selection of high-affinity autoreactive T cells in the
thymus [4,5] and central immune tolerance against
FVIII cannot establish itself. FVIII in FVIII products
that are given for replacement therapy to patients
who carry such mutations would be seen as a foreign
protein by their immune system. Why some of these
patients develop FVIII inhibitors while others do not
is far from clear. For many years immunologists

2009 Blackwell Publishing Ltd

NEW REGIMEN CAN REDUCE INHIBITOR DEVELOPMENT

believed that the immune systems primary goal was

to discriminate between self and non-self [6,7].
Matzinger introduced the concept that the primary
driving force of the immune system is the need to
detect and protect against danger [8]. If a foreign or a
self-antigen is not dangerous, immune tolerance is
the expected outcome [8]. In recent years, it has been
suggested that the ability of the immune system to
sense danger is part of a more general surveillance,
defence and repair system that enables multicellular
organisms to control whether their cells are alive or
dead and to recognize when micro-organisms intrude
[912]. Danger is transmitted by various signals that
are associated either with pathogens or with tissue
and cell damage [912]. Pathogens express pathogenassociated molecular patterns (PAMPS) that are
recognized by pattern recognition receptors such as
toll-like receptors (TLR), Nod1-like receptors
(NLRs) or Rig-I like receptors (RLRs) that are
expressed on a range of cells of the innate and
the adaptive immune system. Once these receptors
are triggered, several signaling pathways are activated that can induce inflammatory responses and
the activation of specific anti-pathogen immune
responses. Evidence is accumulating that trauma,
ischemia and tissue damage can cause inflammatory
responses that are very similar to responses induced
by pathogens [912]. Damaged cells release so called
damage-associated molecular patterns (DAMPs) that
recruit and activate receptor-expressing cells of the
innate immune system, including dendritic cells,
granulocytes, monocytes or eosinophils, and thus
directly or indirectly promote adaptive immune
responses [912].
Based on the increasing evidence that both pathogen-associated as well as cell-damage associated
molecules present danger signals that can stimulate
inflammatory responses of the innate immune system
and thereby up-regulate antibody responses, we
asked whether the prevention of such danger signals
during treatment with FVIII products could decrease
the risk for the development of FVIII inhibitors in
PUPs with severe haemophilia A. We minimized the
exposure to immunological danger signals by avoiding first treatment with FVIII in a bleeding situation
or during infection, by avoiding surgery during the
first 20 exposure days (EDs) and by avoiding
vaccinations on the same day as FVIII treatments.
Furthermore, any bleeds that did occur were treated
early by giving higher doses immediately, thereby
avoiding long and intensive treatment and shortening
the time of tissue damage.
Our results indicate that minimizing danger signals
during the first 20 EDs with FVIII might indeed

2009 Blackwell Publishing Ltd

257

reduce the risk of inhibitor formation. However,

these results should be interpreted as hypothesis
generating and need to be confirmed in a larger
prospective clinical study.

Patients and methods

Twenty six PUPs in two centers in Germany with
severe haemophilia A (all <1% FVIII baseline
activity) with a variety of FVIII gene mutations,
the majority high risk, were treated with a prophylaxis regimen designed to induce immune tolerance
by avoiding immunological danger signals. The
incidence of inhibitor development in this group
was compared with that in a historical control
group of 30 children treated with a standard joint
protection prophylaxis regimen. To avoid selection
bias both study and control group consists of
consecutive PUPs with severe haemophilia A (<1%
FVIII) as they appeared in the respective haemophilia center during a given time period. Based on
the immunological danger theory and their potential
impact on FVIII inhibitor development the new
prophylaxis regimen was prospectively planned and
implemented as standard of care by January 2001 in
center A (Bremen) and by January 2005 in center B
(Munich).
Study aim
The overall risk of developing inhibitors to FVIII
during the first 150 EDs is 2030% for PUPs [13]. Of
those developing inhibitors, 50% will do so within
the first 20 days and 95% during the first 50 days
[13]. If the patient can be brought through this high
risk period without inhibitor development, the
subsequent risk is low [14].
We therefore decided to test the efficacy in
overcoming the high risk of the first 50 EDs of a
prophylaxis regimen specifically designed to induce
tolerance to the administered FVIII and to minimize
inhibitor development.
Treatment
According to the German haemophilia treatment
guidelines prophylaxis in children with haemophilia
is standard of care [15]. Patients in the study group
were treated with low dose prophylaxis, starting
with 250 IU once a week (corresponds to approximately 25 IU kg)1 week)1) as soon as a bleeding
tendency manifested, either through soft tissue and
muscle bleeds or a significant tendency for haematomas (Fig. 1). It was also introduced for safety

Haemophilia (2010), 16, 256262

258

K. KURNIK et al.

1 x weekly
25 lU kg1

When bleeding pattern requests:

2 x 25 lU week1 3 x 25 lU week1

Fig. 1. Treatment scheme for PUPs receiving the new prophylaxis regimen.

reasons as bleed prophylaxis after child-felt trauma

(i.e. typical head trauma without bleeding signs).
Prophylaxis was initiated without insertion of a
Port-A-Cath after a minimal number of on-demand
FVIII exposures. In patients with early joint bleeds
prophylaxis was introduced at the higher frequency
of 25 IU kg)1 twice a week, and in those with early
severe joint or life threatening bleeds at 25
50 IU kg)1 three times a week. When required by
the severity of the bleeding tendency the frequency
was increased from one per week to two per week
or three per week. For tolerization (as also known
from ITI programs in inhibitor patients) it seems to
be important to give prophylactic FVIII doses
always on the same weekday and to avoid interrupting the prophylaxis regimen even when additional on-demand FVIII doses to manage bleeds are
given.
During this tolerization period, immunological
danger signals were minimized by avoiding giving
first FVIII in a severe bleeding situation or during
an infection, avoiding surgery during the first 20
EDs, avoiding giving vaccinations on the same day
as FVIII and giving all vaccinations subcutaneously
rather than intramuscularly. Any bleeds that did
occur were treated early by giving a higher than
the prophylactic dose immediately, thereby avoiding long or intensive treatment. Patients in the
study group were tested for inhibitors every 34
EDs.
Patients in the control group were treated with a
standard joint-protection prophylaxis regimen of
4050 IU kg)1 FVIII three times a week, starting at
or after the first joint or other severe bleed. Please
note that some of the patients in the control group

Haemophilia (2010), 16, 256262

(n = 8) developed their inhibitors already during ondemand therapy before they entered a standard
prophylaxis program. The vaccination guidelines
have been the same for both the study and the
control group.
Statistical analysis
Differences in inhibitor development between the
study group and the historical control group were
analysed by Fishers exact test and odds ratios (OR).
The effect of potential determinants on inhibitor
risk such as FVIII gene mutation and type of product
(recombinant vs. plasma-derived FVIII) was evaluated for the two groups in a logistic regression
model.
Differences between the two study groups of
treatment-related parameters such as median EDs
before prophylaxis and age at start of prophylaxis
were assessed by Wilcoxon test.

Results
Fifty six of the 58 subjects studied had more than
100 EDs to FVIII therapy. Data from these were
analysed for inhibitor development and both patientrelated and treatment-related factors which might
have affected inhibitor development.
There were no significant differences between the
study and control groups in any patient-related
factors (Table 1), nor in the majority of treatmentrelated factors (Table 2). In a logistic regression
model for inhibitor development with factors for
study group (standard vs. new regimen prophylaxis),
genetic risk for inhibitor development (low vs. high),

2009 Blackwell Publishing Ltd

NEW REGIMEN CAN REDUCE INHIBITOR DEVELOPMENT

259

Table 1. Patient-related risk factors for inhibitor development in the study group compared with the control group.

Demographics Bremen
Born between
Ethnicity
Demographics Munich
Born between
Ethnicity
Genetic factors
Severity of haemophilia A
FVIII mutation type*:
High risk (%)
Low risk (%)
Unknown (%)

Control group
(standard prophylaxis regimen)
(n = 30)

Study group
(new prophylaxis regimen)
(n = 26)

March 1995December 2000

All Caucasian (n = 15)

January 2001July 2007

All Caucasian (n = 13)

Not significant

January 2002September 2004

All Caucasian (n = 15)

January 2005October 2007

All Caucasian (n = 13)

Not significant

All <1% FVIII activity

All <1% FVIII activity

Not significant

24 (80)
5 (17)
1 (3)

18 (69)
8 (31)

Not significant
Not significant

Statistical
significance

*Categorization of genetic risk according to Oldenburg J, Pavlova A. Genetic risk factors to inhibitors against FVIII and IX. Haemophilia
2006; 12 (Suppl. 6): 18.

and type of factor concentrate (recombinant vs.

plasma-derived), only the type of prophylaxis regimen had a significant effect (P = 0.005). Logistic
regression analysis was not performed for the risk of
high responder inhibitors due to lack of events in
patients given the new regimen.
There were however highly significant differences
between groups for the prophylaxis-related factors:
age at start of prophylaxis and the number of EDs
before the introduction of prophylaxis (Table 3).
Whereas the new prophylaxis regimen was started
after a median of 1 FVIII EDs at a median age of
10.7 months compared to the historical control
group were high dose prophylaxis was started later

after a median of 30 FVIII on-demand EDs at a

median age of 19 months (P < 0.006).
Age at start of prophylaxis was available for 23 of
the 30 subjects in the standard prophylaxis group and
all 26 subjects given the new regimen. The median age
at start of prophylaxis was 19 months (range 0.887)
for those given standard prophylaxis and 10.7 months
(range 0.524.5) for those given the new regimen. This
difference is highly significant (P < 0.0006).
Standard prophylaxis had been introduced after a
median of 30 EDs (range 1infinity) whereas the new
regimen was introduced after a median of 1 ED
(range 014). This difference too is highly significant
(P < 0.0001).

Table 2. Treatment-related risk factors for

inhibitor development in the study group
compared with the control group.

Control group
(standard prophylaxis
regimen)
(n = 30)
Product type
rFVIII (%)
pdFVIII (%)
Age at first exposure
Median (months)
Range (months)
Reason for first
exposure
Bleed (%)
Safety (%)
Total EDs*
Surgery
Within the first 20 EDs
Vaccination
Given i.m.
Given on a FVIII day

16 (53)
14 (47)
9.8
0.122

Study group
(new prophylaxis
regimen)
(n = 26)
15 (58)
11 (42)

Statistical
significance
Not significant

8.1
021

Not significant

21 (70)
9 (30)
>100

12 (46)
14 (54)
>100

Not significant:
P = 0.103
Not significant

Not significant

1
All no

All no
All no

Not significant

*As on August 2009.

2009 Blackwell Publishing Ltd

Haemophilia (2010), 16, 256262

260

K. KURNIK et al.
Study group
Control group
(standard prophylaxis (new prophylaxis
regimen)
regimen)
(n = 26)
(n = 30)

Age at start of prophylaxis

Median (months)
Range (months)
EDs before prophylaxis
Median
Range

(n = 23)
19
0.887
30
1infinity

(n = 26)
10.7
Highly significant:
0.524.5
P < 0.0006
1
014

Fourteen of the 30 subjects given standard prophylaxis and one of the 26 subjects given the new
prophylaxis regimen developed an inhibitor. The
difference between the groups was highly significant
(P = 0.0003, OR 0.048, 95% CI: 0.0010.372)
(Table 4).
Eight subjects given standard prophylaxis but none
of those given the new regimen were high responders.
The difference between groups was again significant
(P = 0.005, OR for high response 0.00, 95% CI:
0.000.57) (Table 4). Inhibitors in the control group
developed after a median of 11 EDs (range: 3170
EDs) which is well in agreement with a recent
international study [16].
The cumulative inhibitor incidence in the study
group on the new prophylaxis regimen was reduced
by 95% (OR 0.048) as compared to the control
group on a standard protocol (P = 0.0003, 95% CI:
0.0010.372) (Fig. 2).
As a post-hoc analysis, these results should be
interpreted as hypothesis generating. Confirmation in
a prospectively planned, historically controlled study
would be warranted.

Discussion
It may be considered that the overall risk of
developing an inhibitor reflects the level of danger
signals perceived by the patients immune system. It

Inhibitors (%)

High responders (%)

Low responders (%)

Statistical
significance

Highly significant:
P < 0.0001

is not, therefore, surprising that on-demand treatment which is, by definition, given in the presence of
bleeding should cause inhibitor development more
frequently than prophylaxis.
The value of prophylactic factor replacement
therapy in the prevention of severe joint bleeds and
arthropathy is now well established [17], and is
increasingly being adopted as the standard approach
to treatment of haemophilia A. However, even in
those countries, such as Sweden, where prophylaxis
is virtually universal there has been no reduction in
the overall incidence of inhibitors in PUPs [18]. The
prophylaxis regimens employed have been designed
for joint protection, with relatively high doses of
concentrate such as 50 IU kg)1 three times per week.
Because they are usually introduced at or just after
the first significant joint bleed, the FVIII is being
introduced at a time when there are strong immunological danger signals present, to an immune system
which has already been primed by previous ondemand therapy. Therefore, prophylaxis might start
too late to prevent inhibitor formation.
An effective prophylactic regimen for the treatment
of PUPs without the development of inhibitors must
take into account and avoid known danger signals,
such as bleeding associated with tissue damage,
immunological challenges such as vaccination, or
infection. This would permit the immune system to
develop tolerance to the foreign protein in a non-

Control group
(standard prophylaxis
regimen)
(n = 30)

Study group
(new prophylaxis
regimen)
(n = 26)

Statistical
significance

14 (47)

1 (3.8)

8 (27)
6 (20)

0
1 (3.8)

Highly significant:
P = 0.0003
OR 0.048 (95% CI:
0.0010.372)
Highly significant:
P = 0.005
OR of high response
0.00 (95% CI:
0.000.57)

Haemophilia (2010), 16, 256262

Table 3. Prophylaxis-related factors for

inhibitor development in the study group
compared with the control group.

Table 4. Inhibitor development in the

study group compared with the control
group.

2009 Blackwell Publishing Ltd

NEW REGIMEN CAN REDUCE INHIBITOR DEVELOPMENT

Fig. 2. Cumulative inhibitor incidence with increasing number of

EDs: control vs. study group.

threat situation. The results of this study demonstrate

that this approach with an early start of low dose
prophylaxis once weekly might have the capacity to
dramatically reduce the incidence of inhibitors, even in
high-risk patients, from the normally expected level,
which in PUPs has been around 30% [1,13].
It remains difficult to judge which parameters of
the new prophylaxis regimen were of major influence
on inhibitor development: the low number of
on-demand exposures before prophylaxis, the low
dose/frequency of the prophylaxis regimen, the
young age at start of prophylaxis or a combination
of some or of all of them. The avoidance of first FVIII
exposure during a severe bleeding episode might be a
direct protector from inhibitor development whereas
the age, however, might play only an indirect role as
the earlier prophylaxis is started the more likely the
PUP can reach >50 tolerizing EDs without the need
for intensive treatment due to a severe joint bleed.
However future studies will have to evaluate the
significance of single treatment-related factors and
further refine the optimal regimen for inducing
immunotolerance.
We are aware of the fact that our results can only
be considered as hypothesis generating and need to
be confirmed in a larger prospective clinical study.
Our results also suggest that early introduction of
FVIII is a more satisfactory way of avoiding inhibitors than attempting to delay the use of FVIII, for
example by treating bleeds with rFVIIa [19]. Starting
with prophylaxis early in life, in our study at a
median age of 10.7 months, was not associated with
an increased inhibitor risk, a finding that is well in
line with other recent studies [20,21].
A low dose, escalating regimen may also provide a
better long-term outcome for patients, with less
frequent joint bleeds and better joint scores, due to
the earlier start on prophylaxis. The beneficial effect
on joint outcomes is hard to explain, since a weekly
prophylaxis regimen cannot maintain FVIII levels

2009 Blackwell Publishing Ltd

261

above 1%. Nevertheless, benefit from a regimen

similar to ours has been demonstrated in a 10-year
study into the Canadian tailored primary prophylaxis regimen [22]. This regimen differs from our
proposed regimen in using higher doses, introducing
prophylaxis only after a joint bleed has occurred and
stepping up only after inadequacy of dosage is
demonstrated by several joint bleeds or development
of a target joint. This beneficial effect should be the
subject of further study.
As well as its key role in preventing inhibitor
development, the new prophylaxis regimen offers a
number of other advantages. With once a week
administration, it is not necessary to insert a Port-ACath, thereby avoiding surgery. If the initial dosage
proves inadequate, it may still be possible to avoid the
need for a Port-A-Cath by increasing the individual
dose rather than the frequency of dosing. Avoiding the
need for a Port-A-Cath is probably a major advantage
for the induction of immune tolerance to FVIII
because any surgical procedure is likely to be associated with some form of tissue damage together with
the generation of danger signals.
Once weekly administration is also simpler for
parents, requiring only one visit to the haemophilia
center each week, so that concordance is easier to
achieve with a consequent improvement in control.
There is also a pharmacoeconomic benefit in that
lower doses and less frequent treatments can allow
considerable cost savings compared with standard
prophylactic regimens.

Conclusions
Summarizing our results, we conclude that early start
of prophylaxis associated with minimizing immunological danger signals during the first 20 EDs with
FVIII should be considered for future therapy of
patients with severe haemophilia A to reduce the risk
of inhibitor formation. Once the patients have
developed tolerance to FVIII, usually after about
2050 EDs on the low dose regimen, and venous
access permitted, prophylaxis might be changed to
the normal three times weekly regimen for optimal
joint protection (Fig. 1).

Acknowledgements
The authors thank Baxter for support in development of this manuscript. The Munich centre thanks
Martin Olivieri and Susan Jenkins for valuable
support on patient care and data collection. It also
acknowledges greatly the work of the coagulation
laboratory of Prof. Dr W. Schramm. The Bremen

Haemophilia (2010), 16, 256262

262

K. KURNIK et al.

centre thanks Dr Julia Johne and Dr David Overberg

for intensive support on data collection.

Disclosures
G. Auerswald, K. Kurnik and C. Bidlingmaier have
been reimbursed for attending and/or speaking at
and/or organizing several symposia on the behalf of
several pharmaceutical industries.
K. Kurnik received funding for research by Baxter,
CSL Behring, Bayer, Wyeth/Pfizer; C. Bidlingmaier
by CSL Behring, Bayer, and Wyeth/Pfizer, and G.
Auerswald by Baxter, CSL-Behring and NovoNordisk.
B. Reipert, W. Engl and H. Chehadeh are Baxter
employees.

References
1 Oldenburg J, Pavlova A. Genetic risk factors for inhibitors to factors VIII and IX. Haemophilia 2006; 12(Suppl. 6): 1522.
2 Astermark J, Lacroix-Desmazes S, Reding MT. Inhibitor development. Haemophilia 2008; 14(Suppl. 3): 3642.
3 Lee CA, Lillicrap D, Astermark J. Inhibitor development in
hemophiliacs: the roles of genetic versus environmental factors.
Semin Thromb Hemost 2006; 32(Suppl. 2): 104.
4 Palmer E. Negative selection clearing out the bad apples from the
T-cell repertoire. Nat Rev Immunol 2003; 3: 38391.
5 Siggs OM, Makaroff LE, Liston A. The why and how of thymocyte
negative selection. Curr Opin Immunol 2006; 18: 17583.
6 Bretscher P, Cohn M. A theory of self-nonself discrimination.
Science 1970; 169: 10429.
7 Burnet FM. A modification of Jernes theory of antibody production using the concept of clonal selection. CA Cancer J Clin 1976;
26: 11921.
8 Matzinger P. Tolerance, danger, and the extended family. Annu
Rev Immunol 1994; 12: 9911045.
9 Bianchi ME. DAMPs, PAMPs and alarmins: all we need to know
about danger. J Leukoc.Biol 2007; 81: 15.

Haemophilia (2010), 16, 256262

10 Lotze MT, Zeh HJ, Rubartelli A et al. The grateful dead: damageassociated molecular pattern molecules and reduction/oxidation
regulate immunity. Immunol Rev 2007; 220: 6081.
11 Kaczorowski DJ, Mollen KP, Edmonds R, Billiar TR. Early events
in the recognition of danger signals after tissue injury. J Leukoc.Biol 2008; 83: 54652.
12 Kono H, Rock KL. How dying cells alert the immune system to
danger. Nat Rev Immunol 2008; 8: 27989.
13 EMEA. Report on expert meeting on FVIII products and inhibitor development, February 28March 2, 2006. EMEA/CHMP/
BPWP/123835/2006, http://www.emea.europa.eu/pdfs/human/bpwg/
12383506en.pdf, accessed April 27, 2009.
14 Hay CR. The epidemiology of factor VIII inhibitors. Haemophilia
2006; 12(Suppl. 6): 238.
15 Executive Committee of the German Medical Association on
the recommendation of the Scientific Advisory Board. Crosssectional Guidelines for Therapy with Blood Components and
Plasma Derivatives; 4th revised edition, 2008; http://www.
bundesaerztekammer.de/page.asp?his=0.6.3288.6716, accessed June
4, 2009.
16 Gouw SC, Van Den Berg HM, Le Cessie S, Van Der Bom JG.
Treatment characteristics and the risk of inhibitor development: a
multicenter cohort study among previously untreated patients with
severe hemophilia A. J Thrombosis Haemostasis 2007; 5: 138390.
17 Coppola A, Di CM, De SC. Primary prophylaxis in children with
haemophilia. Blood Transfus 2008; 6(Suppl. 2): 411.
18 Knobe KE, Sjorin E, Tengborn LI, Petrini P, Ljung RC. Inhibitors in
the Swedish population with severe haemophilia A and B: a 20-year
survey. Acta Paediatr 2002; 91: 9104.
19 Rivard GE, Lillicrap D, Poon MC et al. Can activated recombinant
factor VII be used to postpone the exposure of infants to factor VIII
until after 2 years of age? Haemophilia 2005; 11: 3359.
20 Gouw SC, van der Bom JG, van den Berg HM (for the CANAL
Study group). Treatment-related risk factors of inhibitor development in previously untreated patients with severe hemophilia A: the
CANAL cohort study. Blood 2007; 109: 464854.
21 Chalmers EA, Brown SA, Keeling D et al. (on behalf of the paediatric working party of UKHCDO) Early factor VIII exposure and
subsequent inhibitor development in children with severe haemophilia A. Haemophilia 2007; 13(Suppl. 2): 14955.
22 Blanchette VS, Rivard GE, Pai MK, Israels SJ, McLimont M,
Feldman BM 10 Year Musculoskeletal Outcomes with Tailored
Primary Prophylaxis: the Canadian Hemophilia Prophylaxis Study
[abstract]. Blood 2007; 110: Abstract 84.

2009 Blackwell Publishing Ltd