Journal of Thrombosis and Haemostasis, 9: 1163–1170 DOI: 10.1111/j.1538-7836.2011.04267.

ORIGINAL ARTICLE

Successful immune tolerance induction by FVIII in hemophilia A

patients with inhibitor may occur without deletion of
FVIII-specific T cells
B. PAUTARD,* R. DÕOIRON,  V. LI THIAO TE,* R. LAVENDÕHOMME,à J.-M. SAINT-REMY,à
K . P E E R L I N C K à and M . J A C Q U E M I N à
*Service dÕHématologie Pédiatrique CHU Nord Place Victor Pauchet, Amiens;  Centre pour Hémophiles, AP-HP Hôpital de Bicêtre, Université
Paris-XI, le Kremlin-Bicêtre, France; and àCentre for Molecular and Vascular Biology, University of Leuven, Leuven, Belgium

To cite this article: Pautard B, dÕOiron R, Li Thiao Te V, LavendÕhomme R, Saint-Remy J-M, Peerlinck K, Jacquemin M. Successful immune
tolerance induction by FVIII in hemophilia A patients with inhibitor may occur without deletion of FVIII-specific T cells. J Thromb Haemost 2011;
9: 1163–70.

Summary. Background: The development of an inhibitor is
the major complication facing patients with hemophilia A
treated by administration of factor (F) VIII concentrates.
Restoration of tolerance to FVIII can be achieved by
prolonged administration of FVIII (immune tolerance induc-
tion, ITI). Although ITI has been used for more than 30 years
in patients with hemophilia A and inhibitor, its mechanism of
action is still poorly understood. Objectives: As administration
of high doses of antigen can induce the apoptosis of the T cells
recognizing the antigen, a potential mechanism of action of ITI
may be the deletion of FVIII-specific T cells. Patients/
Methods: We studied the CD4+ T-cell response to FVIII in
five (one mild, one moderate and three severe) patients
successfully desensitized by administration of FVIII and in
control subjects. Results: Following repeated stimulation with
autologous dendritic cells loaded with FVIII, FVIII-specific T
oligoclonal cell lines were expanded from the blood of one of
the successfully desensitized patients. The FVIII-specific T cells
produced IL-5, IL-13 and IL-2. By contrast, FVIII-specific T-
cell lines could not be derived from three patients with mild
hemophilia A without inhibitor or from four normal control
subjects. Conclusions: These data represent the first analysis of
the cellular mechanisms regulating the induction of tolerance to
FVIII. They demonstrate that successful tolerance induction
may occur without deletion of FVIII-specific T cells.
Keywords: factor VIII, hemophilia A, inhibitor, T cell.
Correspondence: Marc Jacquemin, Centre for Molecular and Vascular
Biology, Herestraat 49, B-3000 Leuven, Belgium.
Tel.: +32 16 345776; fax: +32 16 345990.
E-mail: marc.jacquemin@med.kuleuven.be
Received 13 December 2010, accepted 14 February 2011
Introduction
The development of an immune response to FVIII is a major
complication facing hemophilia A patients treated by admin-
istration of FVIII concentrates. Prolonged administration of
FVIII frequently results in the disappearance of inhibitor
antibodies [1]. However, although immune tolerance induction
(ITI) has been used for more than 30 years, only limited
information is available regarding its mode of action.
By contrast, the cellular mechanisms leading to the devel-
opment of a primary immune response to FVIII have been
extensively studied in patients with hemophilia A and in animal
models [2]. Experimental models have defined the central and
peripheral mechanisms sustaining tolerance to self at the level
of the B and of the T lymphocytes. This knowledge has allowed
the development of novel methodologies to control the immune
response to FVIII [2–6]. However, it is still unknown whether
the regulatory mechanisms evaluated in these studies play a
role in ITI.
Some information is available regarding a potential regula-
tion of B cells during ITI. Gilles et al. [7] have reported that
successful ITI is associated with the development of anti-
idiotypic antibodies competing with inhibitor antibodies for
binding to FVIII. Studies are ongoing to determine whether
such anti-idiotypic antibodies could be administered to patients
with inhibitor in order to neutralize anti-FVIII antibodies or
kill the B cells producing them [8]. In a mouse model, high
doses of FVIII have been shown to inhibit the differentiation of
FVIII-specific B cells both in vitro and in vivo, possibly by
inducing B-cell apoptosis [9]. The potential role of additional
strategies targeting B cells, such as anti-CD20 antibodies or
immunosuppressive agents, [6,10,11] in ITI is still under
investigation.
Of note, no data are available regarding the impact of ITI on
FVIII-specific T cells, although such cells have been described
in hemophilia A patients with inhibitor [12–17] and are believed
to play a causal role in the development of the immune
response to FVIII [18,19]. In animal models, administration of

 2011 International Society on Thrombosis and Haemostasis

1164 B. Pautard et al

high doses of an antigen commonly results in elimination of the
antigen-specific T cells, a phenomenon called deletion [20,21].
To determine whether such a mechanism sustains restoration
of tolerance to FVIII following ITI, we studied FVIII-specific
T cells in successfully desensitized patients.
Materials and methods
PatientsÕ history and treatment
Five desensitized patients with hemophilia A who had devel-
oped FVIII inhibitor and had been successfully desensitized by
prolonged administration of FVIII were included in this study.
The criteria for successful desensitization were two consecutive
Bethesda tests < 0.6 BU, normalization of FVIII recovery
(‡ 66% of expected value) and FVIII half-life ‡ 7 h.
Patient Mi1 (7 years old) has mild hemophilia A (exon 14
substitution Arg698Trp, nomenclature used in the Hemophilia
A database [HADB; HAMSTeRS], FVIII:C = 16%–30%).
He developed an immune response to exogenous and self
FVIII, after 27 exposure days, with a peak inhibitor titer of 12
BU (Table 1) and a reduced basal FVIII of 3%. He was
desensitized by administration of FVIII (200 IU kg)1 per day).
After 2 months the inhibitor titer increased and Mi1 was
transiently treated with corticosteroids (starting dose,
2 mg kg)1 per day; Fig. 1A) until a complete elimination of
the inhibitor and normalization of FVIII recovery and half-life
were achieved. Following successful ITI, Mi1was treated by
regular infusion to FVIII to maintain tolerance to FVIII
(30 U kg)1, twice or three times weekly).
Patient Mo1 (14 years old) has moderate hemophilia A
(exon 8 substitution Ala396Val, nomenclature used in the
HADB database, FVIII:C = 1%–5%). He developed an
immune response to exogenous and self FVIII after more than
100 exposure days, with a peak inhibitor titer of 8 BU. He was
desensitized by administration of FVIII (200 IU kg)1 per day).
Given the persistence of a low titer inhibitor for more than
6 months after initiation of ITI, he received corticosteroids
(starting dose, 1.5 mg kg)1 per day) for 2 months until a
complete elimination of the inhibitor and normalization of
FVIII recovery and half-life (7–8 h) were achieved. Following
ITI, Mo1 was treated by prophylaxis (25–30 U kg)1, twice or
three times weekly). The baseline FVIII levels were then similar
to those before inhibitor development.
Patients Se1 has severe hemophilia A related to intron 22
inversion. Disclosure of anti-FVIII inhibitor (2.2 BU) occurred
at 33 months after 25 exposure days of plasma-derived FVIII.
ITI was started immediately with 40 U kg)1 of plasma-derived
FVIII three times a week. The maximum inhibitor titer
observed was 6.6 BU 6 months after the start of ITI. Inhibitor
titers were negative after 11 months of treatment but 17
additional months were needed to fully normalize the recovery
and half-life of infused FVIII at 65% and 7h30, respectively.
Prophylaxis was used during the next 11 years, then the patient
switched to on demand treatment without recurrence of
inhibitor. Samples used for this study were collected 19 years
after inhibitor detection.
Patient Se2 (substitution Arg1997Trp) has severe hemophilia
A with a mild bleeding phenotype. He developed an immune
response to FVIII at 33 years after 11 exposure days. He was
successfully desensitized by administration of FVIII
(100 IU kg)1 per 48 h). The sample used for this study was
collected immediately after inhibitor eradication.
Patient Se3 also has severe hemophilia A related to intron 1
inversion. Age and number of exposure days at inhibitor
discovery are unknown as no medical records were available
from his native country. When he arrived in France he was
13 years old and had an inhibitor at 2 BU. ITI was started
1 year later with recombinant FVIII dosed at 200 U kg)1 per
day. The maximum inhibitor titer on ITI was 56 BU and
complete success was achieved in 5 months (two inhibitor titers
< 0.6 BU, normal recovery and half-life of infused FVIII at
75% and 8 h, respectively). Samples used for this study were
collected immediately after inhibitor eradication.
One patient (Se4) with severe hemophilia A and a high titer
inhibitor persisting after unsuccessful ITI, three patients with

Table 1 Clinical characteristics of patients with hemophilia A included in the study

Maximal
Patient Hemophilia inhibitor
identification FVIII mutation type titer (BU) Type of ITI ITI outcome

Mi1* Arg698Trp Mild 12 200 U kg)1 per Successful

24 h glucocorticoids
Mo1* Ala396Val Moderate 8 200 U kg)1 per Successful
24 h glucocorticoids
Mi2* ND Mild 0 – –
Mi3* ND Mild 0 – –
Mi4* ND Mild 0 – –
Se2  R1997W Severe 10 100 U kg)1 per 48 h Successful
Se1à Intron 22 inversion Severe 6.6 40 U kg)1 · 3 per week Successful
Se3à Intron 1 inversion Severe 56 200 U kg)1 per 24 h Successful
Se4  Deletion Severe 2200 100 U kg)1 per 24 h Unsuccessful

ND, not determined. Patient treated in the Hemophilia Treatment Center in Amiens (*), Leuven ( ) or Le Kremlin-Bicêtre (à).

 2011 International Society on Thrombosis and Haemostasis

 FVIII-specific T cells in ITI 1165

A Blood collection
15 Mi1 100
Inhibitor titer (BU) ITI
Corticotherapy 75 Blood was collected after informed consent was obtained. The

FVIII (%)
10 Inhibitor titer study was approved by the University of Leuven Ethics
FVIII at 24 h 50 Committee.
FVIII at 72 h
5
25

0
T-cell expansion
0
0 3 6 9 12 24 36 For each blood collection, 20 microcultures were initiated with
Time (months) 100 000 CD4+ T cells and autologous dendritic cells loaded
B C
with FVIII as previously described [14]. T-cell clones were
100 Mi1 9 derived from oligoclonal cell lines as described [14].
Mi1
Stimulation index
FVIII (%)

6
Detection of FVIII-specific T cells in ELISPOT

3 Detection of FVIII-specific T cells was performed in ELISPOT

using Eli-spot kits for IFN-c, IL-5, IL-13 and IL-2 (Diaclone,
10 0 Besançon, France) according to the manufacturerÕs recom-
0 24 48 72 IFN-γ IL-5 IL-13 IL-2 mendations. In each well of a MultiscreenTP Immobilon-P
h
membrane plate (Millipore, Bedford, MA, USA) precoated
D E with cytokine antibody, 10 000 T cells were incubated with
100 15 30 000 autologous and 10 000 heterologous lymphoblastoid
Mi1 Mi1
Stimulation index

12 cells in the presence or in the absence of recombinant FVIII

(10 lg mL)1). Each test was performed at least in duplicate.
FVIII (%)

9
6
3
10 0 2000 (Bio-Sys Gmbh, Karben, Germany). When the mean
0 24 48 72
h numbers of spots in the presence and in the absence of FVIII
Fig. 1. FVIII-specific T cells in blood of successfully desensitized hemo-
philia A patient Mi1. (A) Time schedule of collection of CD4+ T lym-
phocytes derived from patient Mi1. The arrows indicate the time-points at
which T lymphocytes were collected. Basal FVIII levels were measured
either 24 h after administration of FVIII concentrates during ITI (open
circle) or 72 h after administration of FVIII (open square) after successful
ITI. (B) Recovery and pharmacokinetics of FVIII measured following
administration of 30 IU FVIII per kg about 1 month (open star in panel
A) before the analysis of FVIII-specific T cells indicated by the open arrow
in panel A. (C) Elispot analysis of CD4+ T-cell microcultures derived
from patient Mi1 at the time-point indicated by the open arrow in panel A.
Results are expressed as the ratio (stimulation index) of the mean number of
spots recorded in wells incubated with FVIII to that in control wells (white
circle, the difference between the number of spots in the presence and the
absence of FVIII is between 15 and 30; black circle, the difference is higher
than 30). Results of microcultures are not represented when the difference
between the mean number of spots in the presence and in the absence of
FVIII is lower than 15. Twenty microcultures were tested in each experi-
ment. (D) Recovery and pharmacokinetics of FVIII measured following
administration of 30 IU FVIII per kg about 1 month (filled star in panel A)
before the analysis of FVIII-specific T cells indicated by the filled arrow
in panel A. (E) Elispot analysis of 20 microcultures of patient Mi1 CD4+
T cells collected after successful ITI and a subsequent FVIII continuous
infusion (filled arrow in panel A). Results are expressed as in panel C.
mild hemophilia A (Mi2, Mi3, Mi4) without inhibitor and
treated with FVIII, and four subjects (Ct1, Ct2, Ct3 and Ct4)
without hemophilia A were included as controls in the study.
The characteristics of all patients with hemophilia A
included in the study are summarized in Table 1.
After 24 h incubation, the spots corresponding to cells
producing each cytokine were detected according to the
manufacturerÕs recommendations and read with a BioReader
IFN-γ IL-5 IL-13 IL-2
differed by at least 15, the difference was considered as
potentially significant and the results were expressed as the
ratio (stimulation index) between the number of spots in the
presence and in the absence of FVIII. A stimulation index
higher than 3 was taken as cut-off value to identify a specific T-
cell response [22].
Detection of anti-FVIII antibodies in ELISA
Anti-FVIII IgG antibodies were detected in ELISA with the
factor VIII antibody screen (GTI Diagnostics, Waukesha, WI,
USA). According to the manufacturerÕs recommendations, test
results showing optical density values greater than the average
of the kit cut-off control wells were regarded as positive results.
Test results showing optical density values less than or equal to
the average of the kit cut-off control wells were regarded as
negative results.
Results
To determine the fate of FVIII-specific T cells following ITI, we
studied microcultures of CD4+ T cells derived from five
hemophilia A patients who had developed inhibitor and had
been successfully desensitized prior to collection of the cells. All
microcultures were first stimulated twice with autologous
dendritic cells loaded with FVIII so that the frequency of
FVIII-specific T cells could be increased if FVIII-specific

 2011 International Society on Thrombosis and Haemostasis

1166 B. Pautard et al

precursor cells were present at the initiation of the experiment. A

9
B
9

Stimulation index
Stimulation index
All microcultures were then tested in ELISPOT for the Mo1 Se1
presence of FVIII-specific T cells. 6 6
Several microcultures of patient Mi1 produced IL-5 or IL-
2, with a stimulation index higher than 3 (Fig. 1C), indicating 3 3
the presence of FVIII-specific T cells 1 year after complete
0 0
elimination of the inhibitor (open arrow in Fig. 1A). Inter- IFN-γ IL-5 IL-13 IL-2 IFN-γ IL-5 IL-13 IL-2
estingly, Mi1 remained tolerant to FVIII even after a 10-day
C D
continuous infusion of FVIII 18 months after elimination of 9 9

Stimulation index

Stimulation index
Se2 Se3
the inhibitor. After this treatment episode, blood was
6 6
collected again (filled arrow in Fig. 1A) and several micro-
cultures from CD4+ T cells responded to FVIII specifically 3 3
by producing IL-5, IL-13 or IL-2 (Fig. 1E and Tables S1-S4),
demonstrating the long-term persistence of FVIII-specific T 0 0
IFN-γ IL-5 IL-13 IL-2 IFN-γ IL-5 IL-13 IL-2
cells after ITI. Out of these 20 microcultures, six produced at
least one interleukin with a stimulation index higher than 3,
Fig. 2. Absence of detectable FVIII-specific T cells following ITI in blood
indicating a precursor frequency ‡ 1/333 000 among Mi1 of successfully desensitized hemophilia A patients Mo1, Se1, Se2 and
CD4+ T cells. Several FVIII-specific T-cell clones could be Se3. Elispot analysis of 20 microcultures of patients Mo1 (A), Se1 (B),
derived from the oligoclonal cell lines analyzed in Fig. 1(C,E) Se2 (C) and Se3 (D). The CD4+ T cells were collected after successful ITI.
(data not shown). Twenty microcultures were tested in each experiment. Results were
Figure 1(B,D) shows the recovery and pharmacokinetics of expressed as in Fig. 1(B).
FVIII following administration of 30 IU FVIII per kg about
1 month before each of the two analyses of patient Mi1 FVIII-
specific T cells. After subtraction of FVIII basal values, the A
15
FVIII recovery was 66 and 42 IU dL)1, respectively, and the
Stimulation index
Mi2
12 Mi3
half-life was about 8 h at both time-points. The FVIII recovery Mi4
9
higher than 66% of the expected value and the FVIII half-life
6
longer than 7 h, indicated fully successful ITI.
3
To further ensure that anti-FVIII antibodies had been
0
completely eliminated following ITI, the presence of non- IFN-γ IL-5 IL-13 IL-2
inhibitory antibodies was evaluated by ELISA. No significant
B
IgG binding to FVIII was observed with two Mi1 samples 15 Ct 1
Stimulation index

collected during the tolerant period (Table S5). Similarly, no 12 Ct 2

IgG antibodies from the control subject Ct1 did bind to FVIII Ct 3
9 Ct 4
in this assay (Table S5). By contrast, the plasma of patient Se4
6
with severe hemophilia A and a high titer inhibitor showed a
3
high IgG binding (Table S5). These observations indicate that
the anti-FVIII antibodies of Mi1 had not switched to a non- 0
IFN-γ IL-5 IL-13 IL-2
inhibitory pattern and had been fully eliminated following ITI.
C 15
By contrast, no FVIII-specific T cells were detected following Se4
Stimulation index

successful ITI in the blood of one moderate hemophilia A and 12

of three severe hemophilia A patients (Fig. 2A–D). Similarly, 9
T-cell microcultures from three mild hemophilia A patients 6
who never produced antibodies following substitution therapy
3
did not show any clear cut interleukin secretion in the presence
0
of FVIII (Fig. 3A). IFN-γ IL-5 IL-13 IL-2
When oligoclonal T cells from four non-hemophiliac indi-
viduals were exposed to FVIII, no high interleukin production Fig. 3. FVIII-specific T cells in patients with haemophilia A with or
without inhibitor and in control individuals. (A–E) Elispot analysis of
was detected (Fig. 3B). A few clones appeared to produce microcultures of mild/moderate hemophilia A patients who never devel-
interleukin following stimulation with FVIII but the stimula- oped inhibitor (A), of non-hemophiliac individuals (B) and of a patient
tion indexes were always lower than 3 (Fig. 3C) and no FVIII- with severe hemophilia A and a high titer inhibitor (C).
specific T-cell clones could be expanded from those cultures
(data not shown).
In a control experiment, a high number of FVIII-specific T- frequency of FVIII-specific T cells in the latter patient was of
cell clones were also identified in a patient with severe the same order of magnitude as that measured in a patient with
hemophilia A and a high titer inhibitor (Fig. 3C). The mild hemophilia A and inhibitor [14].

 2011 International Society on Thrombosis and Haemostasis


Discussion
Limited information is available regarding the mechanism by
which prolonged administration of FVIII restores tolerance to
FVIII in patients with inhibitor. To determine whether ITI
leads to the deletion of FVIII-specific T cells, we first expanded
FVIII-specific T cells by coculture with dendritic cells loaded
with FVIII and then detected the cells releasing interleukins
following stimulation with FVIII. A significant number of
FVIII-specific T cells were detected in one out of five
successfully desensitized patients, indicating that elimination
of FVIII-specific T cells is not a prerequisite for the restoration
of tolerance of FVIII.
A critical parameter for the detection of FVIII-specific T
cells is the specificity of the assay. The ELISPOT methodology
was selected for the detection of the specific T cells. This
method has the advantage of a high sensitivity as interleukins
are captured and detected in the vicinity of the cells releasing
them. Two complementary criteria were exploited to guarantee
the specificity of the assay. First, we calculated the ratio
between the number of interleukin spots in the presence and in
the absence of stimulation with FVIII [22]. Second, as this ratio
can vary largely when the number of spots is low, we only took
into consideration wells in which more than 15 spots were
detectable after stimulation with FVIII.
T-cell activity characterized by a ratio > 3 is likely to
correspond to antigen-specific T cells, as indicated by our
observations that FVIII-specific T cells can frequently be
cloned and expanded from such microcultures. By contrast, the
significance of T-cell activity corresponding to a ratio between
1 and 3 is unclear. Significantly, the rescreening of several
microcultures with a ratio lower than 3 never confirmed a
FVIII-specific T-cell response (data not shown) and FVIII-
specific T-cell clones could never be derived from such wells. It
is unknown whether this type of low response corresponds to a
low T-cell sensitivity to FVIII, to a T-cell defect in the ability to
proliferate or to secrete interleukin following stimulation with
FVIII, or rather represents a false positive result due to random
variation in the number of background interleukin spots.
The detection of FVIII-specific T cells was performed after
expansion of FVIII-specific T cells. Such a procedure has the
advantage that it increases the precursor frequency of specific T
cells, which improves the reliability of the ELISPOT assay. The
disadvantage of this strategy is that it may modify the
phenotype of the T cells, notably their pattern of interleukin
secretion. Another drawback of the amplification strategy is
that only T cells able to divide can be detected. T cells with a
limited growth potential or a short lifespan may be lost during
the amplification phase. The T-cell expansion method used in
this study has been exploited previously to characterize FVIII-
specific T cells in a patient with mild hemophilia A due to a
point mutation (Arg2150His) who developed a high titer
inhibitor to normal exogenous FVIII but not to his own FVIII
[14,23]. After stimulation with normal recombinant FVIII
several FVIII-specific T-cell clones have been established. All
T-cell clones recognized a single stretch of amino acids
 2011 International Society on Thrombosis and Haemostasis
FVIII-specific T cells in ITI 1167
overlapping the mutation site in the patient FVIII molecule
[14]. The correlation between the specificity of the T-cell clones
and the mutation in the patient FVIII gene strongly suggests
that this method allows the study of T cells specifically
associated with the immune response to FVIII.
In contrast to our results, Reding et al. [13] have identified T
cells proliferating in response to FVIII peptides in hemophilia
A patients without inhibitor and in healthy individuals.
Kamaté et al. have observed a proliferative response after
stimulation with native FVIII in peripheral blood mononuclear
cells (PBMC) in 4 out of 13 normal donors. In one case, this
proliferative response was observed only after depletion of
CD25hi CD4+ T cells. [17]. In this study, the number of
responding donors was significantly lower than in the study by
Reding et al., which may be due to technical differences. As
suggested by Kamaté et al., it is plausible that peptide may be
more easily presented by antigen presenting cells than the
native FVIII molecule. Reding et al. also tested the immune
response of the donors at multiple time-points. The responses
to some peptides were detected at some time-points but not at
others, suggesting that the responses were transitory or that the
frequency of precursors with a given specificity was too low for
such T cells to be present in each experiment [13]. Altogether,
these observations leave open the possibility that, although we
could not expand and clone FVIII-specific T cells from the
blood of hemophilia A patients without inhibitor or from
normal donors, other techniques may have allowed the
detection of FVIII-specific T cells in some of these subjects.
Using CD8-depleted peripheral blood mononuclear cells,
Hu et al. [24] have analyzed the cytokines produced by CD4+
blast T cells responding to native FVIII. In normal donors,
FVIII induced either IFN-c and TGF-b or TGF-b alone. In
patients with hemophilia A without inhibitor, FVIII induced
IFN-c and TGF-b. By contrast, in patients with inhibitor,
FVIII induced IFN-c and IL-4 but no TGF-b [24]. These
results suggested that Th2 cells may support inhibitor forma-
tion whereas FVIII-specific cells expressing TGF-b may
prevent antibody synthesis. In agreement with these observa-
tions, Ettinger et al. [25] have recently succeeded in establishing
T-cell clones specific for a FVIII peptide from one patient with
hemophilia A without inhibitor. These T cells were selected and
expanded with a peptide overlapping the mutation site in the
patientÕs FVIII gene. Interestingly, the T-cell clones derived
from this subject produced Th1 cytokines. By contrast, FVIII-
specific T cells derived from his brother, who had developed an
inhibitor, produced IL-17 at the beginning of the immune
response to FVIII and Th2 cytokines at later time-points [25].
In our study, the identification of FVIII-specific T cells was
based on the detection of Th1 cytokines (IFN-c and IL-2) and
of Th2 cytokines (IL-5 and IL-13). Suppressive cytokines such
as IL-10 and TGF-b could not be detected for technical
reasons. The secretion of IL-17 was not evaluated because the
role of this cytokine was not known when our experiments were
carried out [25]. The CD4+ T cells from the patient with severe
hemophilia A and inhibitor (Se4) produced IFN-c, IL-5 and
IL-2. The FVIII-specific T cells present in one patient after
1168 B. Pautard et al
successful ITI (Mi1) produced IL-5, IL-13 and IL-2. However,
due to the limited size of the study, it is unclear whether the
difference in the pattern of interleukins secreted by the FVIII-
specific T cells is significant.
The persistence of FVIII-specific T cells in a successfully
desensitized patient may highlight the importance of other
elements of the immune system modulating tolerance of
FVIII despite the presence of specific T cells. B cells may
play an important role in such a process, as suggested by
previous observations in patients with mild/moderate hemo-
philia A, who developed an immune response restricted to
exogenous FVIII [9,10]. The T cells that were characterized
in these patients recognized exogenous FVIII but not self
FVIII. In principle, such T cells could provide help to any B
cells recognizing the exogenous (wild-type) FVIII molecule,
including B cells recognizing antigenic determinants expressed
both on the patientÕs FVIII molecule and exogenous FVIII.
The observation that these patientsÕ antibody responses
remained focused on exogenous FVIII implies that an
important tolerance mechanism controls B cells recognizing
self FVIII.
The persistence of highly reactive FVIII-specific T cells in a
successfully desensitized hemophilia A patient suggests that
other mechanisms than T-cell deletion may be at play during
ITI. As suggested by previous observations, B cells may also
be important targets for ITI [7,9]. In successfully desensitized
patients, anti-idiotypic antibodies were reported to neutralize
anti-FVIII antibodies [7]. In addition, animal experiments
suggested that administration of high doses of FVIII can lead
to selective B-cell elimination [9]. The importance of regula-
tory T cells (Treg) in the immune response to FVIII has been
highlighted in different models. Hu et al. [24] have shown
that patients with hemophilia A without inhibitor and some
normal donors have T cells responding to FVIII by
producing TGF-b. Several animal models have also high-
lighted the important role of Treg in the modulation of the
immune response to FVIII [4,26]. Unfortunately, our exper-
imental approach to detecting FVIII-specific T cells did not
allow determining whether the FVIII-specific T cells identi-
fied in one successfully desensitized patient had regulatory
properties that could have contributed to the success of ITI.
It is also unknown whether the strategy that we exploited to
expand the FVIII-specific T cells would allow the expansion
of regulatory T cells.
Interestingly, two of the successfully desensitized patients
included in this study had transiently received glucocorticoids
during the desensitization protocol. Mi1, in whom FVIII-
specific T cells were detected, had received glucocorticoids
when the antibody titer increased after an initial drop
(Fig. 1A). Mo1 was also treated with glucocorticoids when
the antibody had nearly completely disappeared but the
residual FVIII after 24 h remained abnormally low (Fig. 2).
Due to the large variability in the evolution of antibody
response in patients undergoing desensitization, it cannot be
determined whether glucocorticoids had a decisive impact on
the outcome of the therapy. Nevertheless, the data indicate that
glucocorticoids do not prevent successful ITI, in agreement
with previous observations [27]. In studies evaluating ITI
protocols, transient glucocorticoid administration may there-
fore act as an add-on immunosuppressive treatment for
partially successful ITI.
The FVIII-specific T cells persisting after successful ITI
were detected in one of the two patients with mild/moderate
hemophilia A transiently treated with glucocorticoids during
the course of ITI. It remains to be determined whether the
survival of FVIII-specific T cells may be linked to the
treatment with glucocorticoids. Indeed, although glucocortic-
oids induce immunosuppression by inhibition of T-cell
activation, they can also contribute to T-cell survival. The
latter effect is mediated by downregulation of FasL expres-
sion, which results in the inhibition of activation-induced T-
cell apoptosis [28].
ITI is frequently unsuccessful in mild/moderate hemophilia
A patients [29]. The persisting T cells after ITI were detected in
a patient with mild hemophilia A, raising the question of
whether this patient was stably desensitized. Several observa-
tions suggest that this patient was indeed successfully desen-
sitized.
All desensitization criteria (Bethesda titer < 0.5 BU, normal
FVIII recovery and pharmacokinetics, and normalization of
basal FVIII) were met. Those parameters remained stable for
more than 1 year. In addition, the patient remained tolerant of
FVIII after a continuous FVIII infusion for a surgical
intervention, which is a high-risk period for regaining an
inhibitor antibody. T cells were analyzed 1 year after the
complete elimination of the inhibitor antibodies and a few
weeks after the continuous FVIII infusion for the surgical
intervention. In both cases, FVIII-specific T cells were detected
although the inhibitor titer was negative.
In conclusion, the analysis of the cellular response following
successful ITI demonstrated that successful tolerance induction
may occur without deletion of FVIII-specific T cells.
Addendum
R. LavendÕhomme and M. Jacquemin were responsible for
in vitro studies. B. Pautard, R. dÕOiron, V. Li Thiao Te, J.-
M. Saint-Remy, M. Jacquemin, A. L, J.-M. Saint-Remy
and K. Peerlinck contributed to the design of the study. All
authors contributed to the redaction of the manuscript. B.
Pautard, R. dÕOiron and K. Peerlinck performed clinical
studies.
Acknowledgements
We thank M. Stapleton (GTI Diagnostics) for help in the
detection of anti-FVIII antibodies by ELISA and Jos Vermylen
(University of Leuven) for discussions and comments on the
manuscript. This study was supported by grants G.0231.05 and
G.0275.05 from the Flemish Research Foundation, by the
KULeuven Wyeth and Baxter chairs for hemophilia and by the
ÔExcellentie Financiering KULeuvenÕ EF/05/013.
 2011 International Society on Thrombosis and Haemostasis

Disclosure of Conflict of Interests
The authors state that they have no conflict of interest.
Supporting Information
Additional Supporting Information may be found in the online
version of this article:
Table S1. ELISPOT analysis of IFN-c production by FVIII-
specific T cells in blood of successfully desensitized hemophilia
A patient Mi1.
Table S2. ELISPOT analysis of IL-5 production by FVIII-
specific T cells in blood of successfully desensitized hemophilia
A patient Mi1.
Table S3. ELISPOT analysis of IL-13 production by FVIII-
specific T cells in blood of successfully desensitized hemophilia
A patient Mi1.
Table S4. ELISPOT analysis of IL-2 production by FVIII-
specific T cells in blood of successfully desensitized hemophilia
A patient Mi1.
Table S5. Detection of anti-FVIII antibodies by ELISA.
Please note: Wiley-Blackwell are not responsible for the
content or functionality of any supporting materials supplied
by the authors. Any queries (other than missing material)
should be directed to the corresponding author for the article.
References
1 Brackmann HH, Gormsen J. Massive FVIII infusion in haemophiliac
with factor VIII inhibitor, high responder. Lancet. 1977; 2: 933.
2 Waters B, Lillicrap D. The molecular mechanisms of immunomodu-
lation and tolerance induction to factor VIII. J Thromb Haemost 2009;
7: 1446–56.
3 Peng B, Ye P, Rawlings DJ, Ochs HD, Miao CH. Anti-CD3 anti-
bodies modulate anti-factor VIII immune responses in hemophilia A
mice after factor VIII plasmid-mediated gene therapy. Blood 2009; 114:
4373–82.
4 Miao CH, Harmeling BR, Ziegler SF, Yen BC, Torgerson T, Chen L,
Yau RJ, Peng B, Thompson AR, Ochs HD, Rawlings DJ.
CD4+FOXP3+ regulatory T cells confer long-term regulation of
factor VIII-specific immune responses in plasmid-mediated gene
therapy-treated hemophilia mice. Blood 2009; 114: 4034–44.
5 Finn JD, Ozelo MC, Sabatino DE, Franck HW, Merricks EP, Crudele
JM, Zhou S, Kazazian HH, Lillicrap D, Nichols TC, Arruda VR.
Eradication of neutralizing antibodies to factor VIII in canine hemo-
philia A after liver gene therapy. Blood 2010; 116: 5842–8.
6 Zhang AH, Skupsky J, Scott DW. Effect of B-cell depletion using anti-
CD20 therapy on inhibitory antibody formation to human FVIII in
hemophilia A mice. Blood 2011; 117: 2223–6.
7 Gilles JG, Desqueper B, Lenk H, Vermylen J, Saint-Remy JM. Neu-
tralizing antiidiotypic antibodies to factor VIII inhibitors after desen-
sitization in patients with hemophilia A. J Clin Invest 1996; 97: 1382–8.
8 Gilles JG, Grailly SC, De Maeyer M, Jacquemin MG, VanderElst LP,
Saint-Remy JM. In vivo neutralization of a C2 domain-specific human
anti-Factor VIII inhibitor by an anti-idiotypic antibody. Blood 2004;
103: 2617–23.
9 Hausl C, Ahmad RU, Sasgary M, Doering CB, Lollar P, Richter G,
Schwarz HP, Turecek PL, Reipert BM. High dose factor VIII inhibits
factor VIII-specific memory B cells in hemophilia A with factor VIII
inhibitors. Blood 2005; 106: 3415–22.
 2011 International Society on Thrombosis and Haemostasis
FVIII-specific T cells in ITI 1169
10 Berntorp E, Astermark J, Carlborg E. Immune tolerance induction
and the treatment of hemophilia. Malmö protocol update. Haemato-
logica 2000; 85 (10 Suppl): 48–50.
11 Franchini M, Mengoli C, Lippi G, Targher G, Montagnana M,
Salvagno GL, Zaffanello M, Cruciani M. Immune tolerance with rit-
uximab in congenital haemophilia with inhibitors: a systematic litera-
ture review based on individual patientsÕ analysis. Haemophilia 2008;
14: 903–12.
12 Singer ST, Addiego JE, Reason DC, Lucas AH. T lymphocyte pro-
liferative responses induced by recombinant Factor VIII in hemophilia
A patients with inhibitors. Thromb Haemost 1996; 76: 17–22.
13 Reding MT, Wu H, Krampf M, Okita DK, Diethelm-Okita BM,
Christie BA, Key NS, Conti-Fine BM. Sensitization of CD4+ T cells
to coagulation Factor VIII: response in congenital and acquired
hemophilia patients and in healthy subjects. Thromb Haemost 2000; 84:
643–52.
14 Jacquemin M, Vantomme V, Buhot C, LavendÕhomme R, Burny W,
Demotte N, Chaux P, Peerlinck K, Vermylen J, Maillere B, van der
Bruggen P, Saint-Remy JM. CD4+ T-cell clones specific for wild-type
factor VIII: a molecular mechanism responsible for a higher incidence
of inhibitor formation in mild/moderate hemophilia A. Blood 2003;
101: 1351–8.
15 Jones TD, Phillips WJ, Smith BJ, Bamford CA, Nayee PD, Baglin TP,
Gaston JS, Baker MP. Identification and removal of a promiscuous
CD4+ T cell epitope from the C1 domain of factor VIII. J Thromb
Haemost 2005; 3: 991.
16 James EA, Kwok WW, Ettinger RA, Thompson AR, Pratt KP. T-cell
responses over time in a mild hemophilia A inhibitor subject: epitope
identification and transient immunogenicity of the corresponding self-
peptide. J Thromb Haemost 2007; 5: 2399–407.
17 Kamaté C, Lenting PJ, van den Berg HM, Mutis T. Depletion of
CD4+/CD25high regulatory T cells may enhance or uncover factor
VIII-specific T-cell responses in healthy individuals. J Thromb Hae-
most 2007; 5: 611–3.
18 Ragni MV, Bontempo FA, Lewis JH. Disappearance of inhibitor to
factor VIII in HIV-infected hemophiliacs with progression to AIDS or
severe ARC. Transfusion. 1989; 29: 447–9.
19 Bray GL, Kroner BL, Arkin S, Aledort LW, Hilgartner MW, Eyster
ME, Ragni MV, Goedert JJ. Loss of high-responder inhibitors in
patients with severe hemophilia A and human immunodeficiency virus
type 1 infection: a report from the Multi-Center Hemophilia Cohort
Study. Am J Hematol 1993; 42: 375–9.
20 Webb S, Morris C, Sprent J. Extrathymic tolerance of mature T cells:
clonal elimination as a consequence of immunity. Cell 1990; 63: 1249–
56.
21 Rocha B, von Boehmer H. Peripheral selection of the T cell repertoire.
Science 1991; 251: 1225–8.
22 Arif S, Tree TI, Astill TP, Tremble JM, Bishop AJ, Dayan CM, Roep
BO, Peakman M. Autoreactive T cell responses show proinflammatory
polarization in diabetes but a regulatory phenotype in health. J Clin
Invest 2004; 113: 451–63.
23 Peerlinck K, Jacquemin MG, Arnout J, Hoylaerts MF, Gilles JG,
LavendÕhomme R, Johnson KM, Freson K, Scandella D, Saint-Remy
JM, Vermylen J. Antifactor VIII antibody inhibiting allogeneic but not
autologous factor VIII in patients with mild hemophilia A. Blood 1999;
93: 2267–73.
24 Hu G, Guo D, Key NS, Conti-Fine BM. Cytokine production
by CD4+ T cells specific for coagulation factor VIII in healthy
subjects and haemophilia A patients. Thromb Haemost 2007; 97: 788–
94.
25 Ettinger RA, James EA, Kwok WW, Thompson AR, Pratt KP.
Lineages of human T-cell clones, including T helper 17/T helper 1 cells,
isolated at different stages of anti-factor VIII immune responses. Blood
2009; 114: 1423–8.
26 Miao CH. Immunomodulation for inhibitors in hemophilia A: the
important role of Treg cells. Expert Rev Hematol 2010; 3: 469–83.
1170 B. Pautard et al
27 Aznar JA, Jorquera JI, Peiro A, Garcia I. The importance of corticoids
added to continued treatment with Factor VIII concentrates in the
suppression of inhibitors in haemophilia A. Thromb Haemost 1984; 51:
217–21.
28 Zacharchuk CM, Mercep M, Chakraborti P, Simons SS Jr, Ashwell
JD. Programmed T lymphocyte death: cell activation- and steroid-
induced pathways are mutually antagonistic. J Immunol 1990; 145:
4037–45.
29 Hay CR, Ludlam CA, Colvin BT, Hill FG, Preston FE, Wasseem N,
Bagnall R, Peake IR, Berntorp E, Mauser Bunschoten EP,
Fijnvandraat K, Kasper CK, White G, Santagostino E. Factor VIII
inhibitors in mild and moderate-severity haemophilia A. UK Hae-
mophilia Centre Directors Organisation. Thromb Haemost 1998; 79:
762–6.
 2011 International Society on Thrombosis and Haemostasis