In hemophilia A, up to 25% of new antifactor VIII (FVIII) inhibitory antibodies (inhibitors) occur in patients with mild or moderate disease [1]. Once the inhibitor develops, options for management include observation, immune modulation, and immune tolerance induction (ITI). Currently, there is little data to guide a clinician's management decisions. In a case series, eight of the 26 subjects with mild or moderate hemophilia complicated by an inhibitor underwent ITI; two were successful, two were unsuccessful, and four were partially successful [2]. In a systematic review of the literature, 12 of the 16 patients with mild or moderate hemophilia responded to rituximab for treatment to eradicate the inhibitor [3]. To increase our understanding of treatment options for inhibitor eradication in patients with mild or moderate hemophilia A complicated by an inhibitor, a secondary analysis of clinical and treatment characteristics in a cohort of 36 patients with mild or moderate hemophilia A and inhibitor was undertaken. In multivariate analyses, rituximab alone (n = 6) and other immune-modulating treatments alone (n = 2) were significantly associated with an increased likelihood of inhibitor clearance [hazard ratio (HR) = 4.4 (95% CI = 1.06–20.03) and 10.21 (95% CI = 1.17–78.28), respectively], whereas ITI alone (n = 9) was not [HR = 1.35 (95% CI = 0.44–4.07)].

Data were originally collected as part of a case–control study [4]. For this analysis, complete data were available on 32 subjects [missing the peak inhibitor titer (n = 1), effect of inhibitor on FVIII coagulant activity (n = 2), and initial inhibitor titer (n = 1)] (Supporting Information Table 1). The onset of inhibitor was after the year 2000 in 66% of subjects, but as early as 1970s in one subject [4]. The median age at inhibitor onset was 31.0 years [interquartile range (IQR), 41 years]. The majority of subjects were non-Hispanic white (83.3%). Subjects had a median baseline FVIII coagulant activity of 4.0% (IQR = 9%) and a median peak inhibitor titer of 17.0 Bethesda unit (BU)/ml (IQR = 85.7 BU/ml).

Table I. Subject Characteristics According to Inhibitor Status
CharacteristicCleared (N = 25)Persistent (N = 11)P value
  • IQR, interquartile range; FVIII, factor VIII; BU, Bethesda unit.

  • a

    n = 24;

  • b

    n = 10;

  • c

    Fisher exact test;

  • d

    FVIII after inhibitor onset unknown in two subjects.

Age at inhibitor onset (years)30.0 (IQR 47)32.0 (IQR 31)0.86
White race23 (92.0%)9 (81.8%)0.57
Non-Hispanic ethnicity23 (92.0%)11 (100%)1.00
Baseline FVIII (%)4.0 (IQR 9.0)4.0 (IQR 9.2)0.79c
Initial inhibitor titer (BU/ml)4.4 (IQR 9.9)a4.8 (IQR 17.7)0.97
Peak inhibitor titer (BU/ml)14.9 (IQR 39.4)49.0 (IQR 73.6)b0.12
High-titer inhibitor (>5 BU/ml)18 (72.0%)9 (81.8%)0.68c
Received eradication treatment17 (68.0%)6 (54.6%)0.47c
FVIII reduced after inhibitor onset22 (88.0%)9 (81.2%)d0.54c
Duration of inhibitor (months)15.9 (IQR 35.1)93.8 (IQR 168.4)0.22

Characteristics of subjects with cleared versus persistent inhibitors are shown in Table I. The median duration of the inhibitor among subjects whose inhibitor cleared was 15.9 months (IQR = 35.1 months), whereas the median duration of follow-up of subjects with a persistent inhibitor was 93.8 months (IQR = 168.4 months). None of the differences between the two groups in the distribution of studied characteristics reached statistical significance.

Among the eight subjects whose inhibitor spontaneously cleared, the median peak inhibitor titer was 4.4 BU/ml (IQR = 12.4 BU/ml) and only 37.5% of these subjects had a high-titer inhibitor (>5 BU/ml). In contrast, among the 17 subjects whose inhibitor cleared following treatment, the median peak inhibitor titer was 16.0 BU/ml (IQR = 40.8 BU/ml; P = 0.07) and 88.2% (P = 0.002) had a high-titer inhibitor. The median time to inhibitor clearance was 12.9 months (IQR = 24.1 months) and 43.8 months (IQR = 61.7 months; P = 0.15) in those who did or did not receive treatment, respectively.

The impact of receiving any treatment to eradicate the inhibitor on the probability of having a persistent inhibitor over time is shown in Fig. 1a. Although subjects who received any treatment appeared to clear the inhibitor sooner, the results were not statistically significantly different (P = 0.26). The effects of specific treatments, ITI, rituximab, and various other treatments (prednisone, cyclophosphamide, cyclosporine, and IVIg) alone or in combination, on the probability of having a persistent inhibitor are also shown in Fig. 1b–d. Although none of the differences in outcomes were statistically significant, the group that received ITI appeared to be more likely to have a persistent inhibitor when compared with all other subjects.

thumbnail image

Figure 1. Kaplan-Meier survival curves for any treatment versus no treatment (a), and any rituximab (b), any immune tolerance induction (ITI) (c), or any other treatments (d) alone or in combination.

Download figure to PowerPoint

The effect of any ITI, rituximab, or other treatments on the duration of inhibitor after adjustment for age, baseline FVIII coagulant activity, race, and peak inhibitor titer was evaluated in 35 subjects. When comparing any use of ITI, rituximab, or other treatments with those that did not receive ITI, rituximab, or other treatments, respectively, none were statistically significant and only ITI trended toward having a lower likelihood of inhibitor clearance at any given point in time (HR = 0.57, 95% CI = 0.20–1.44).

The effect of rituximab, ITI, other treatments alone, or combinations of treatments after adjustment for age, race, baseline FVIII coagulant activity, peak inhibitor titer, and all other treatment parameters was evaluated in 35 subjects (Table II). Rituximab alone and other treatments alone demonstrated a positive effect consistent with an increased likelihood of inhibitor clearance at any given point in time. There was no apparent effect of ITI alone on inhibitor clearance. Combinations of treatment used in six subjects (ITI + rituximab in two, rituximab + other in three, and ITI + other in one) was associated with a HR of 0.30 (P = 0.23, 95% CI = 0.03–1.51). Age, baseline FVIII coagulant activity, and peak inhibitor titer did not affect the likelihood of clearance of inhibitor (HR = 0.99 for each). Being of white race trended toward greater likelihood of clearance of inhibitor at any given point in time (HR = 1.6, 95% CI = 0.45–8.82) however confidence intervals were wide reflective of the small number of nonwhite subjects enrolled in the study.

Table II. Effect of Each Treatment on the Likelihood of Inhibitor Clearance
ParameterNParameter EstimateStandard Errorχ2P valueHazard Ratio95% CI
  1. a

    Adjusted for race, age, baseline factor VIII, maximum inhibitor titer and all other treatment parameters.

Immune tolerance alone90.300.590.260.61 1.350.44–4.07
Rituximab alone61.490.763.810.05 4.441.06–20.03
Other treatments alone22.321.124.270.0410.211.17–78.28
Combination treatment6−1.190.991.420.23 0.300.03–1.51

Among those that received ITI alone, the median time to the start of ITI was 1.3 months (IQR = 2.7 months), the median age at start of ITI was 8.2 years (IQR = 10 years), and the median peak inhibitor titer was 13 BU/ml (IQR = 33.7 BU/ml). Among the seven subjects in whom ITI was started within 3 months after onset of the inhibitor, the peak titer is likely a close approximation of the pre-ITI titer (median peak titer 7 BU/ml, IQR = 47.2 BU/ml; <10 BU/ml in four subjects and >10 BU/ml in three subjects). In the remaining two subjects who started ITI after 6 months, it is possible that their pre-ITI inhibitor titer was lower than the peak inhibitor titer of 17 and 39 BU/ml.

In this cohort, R593C and N1922S were the most common FVIII gene (F8) mutations found. Among subjects with R593C, four of the six patients cleared their inhibitor after a median of 20.0 months; although one subject classified as having a persistent inhibitor was a relatively recent inhibitor (recorded follow-up of only 0.5 months) and may not have had adequate time to clear spontaneously. Five of six patients with N1922S cleared their inhibitor titer after a median of 9.2 months and the one classified with persistent inhibitor had only 2.8 months of follow-up. There were three subjects with R2150H F8 mutation and none cleared their inhibitor after 93.9, 117.5, and 191.8 months of follow-up.

This study has several limitations. First, it is a retrospective secondary analysis of subjects enrolled from 16 sites for another study of risk factors for inhibitor development [4]. Efforts were made to enroll all eligible subjects; however, 17 other subjects who had been identified as eligible for inclusion were not enrolled either because the hemophilia treatment center (HTC) declined to participate or the subject at a participating HTC declined enrollment. These 17 subjects were similar to those enrolled with regard to baseline FVIII coagulant activity and race. Second, there was limited temporal information, making it difficult to fully assess the direct effect of treatments on inhibitor eradication. Third, the sample size is small leading to a lack of power to detect weak associations. Fourth, the observational nature of the study leads to confounding by indication. Subjects perceived to be at lower risk were observed and those perceived to be at higher risk were treated thus biasing the results against seeing a benefit of treatment.

Despite these limitations, given the rarity of this condition, this cohort provides an opportunity to explore and generate hypotheses that should be tested in subsequent studies. It appears that those subjects who spontaneously cleared their inhibitor tended to have a lower inhibitor titer compared with those that received treatment and ITI appears less beneficial than rituximab or other immune-modulating agents alone. The HR associated with other treatments alone is likely to be an overestimate since it is based on two subjects; however, our overall results suggest that immune modification may be more effective than ITI in eradicating inhibitors in patients with nonsevere hemophilia A. These results are consistent with the poor response rate to ITI reported by Hay et al. [2] and the favorable response rate to rituximab reported by Franchini et al. [3] and do not appear to be the result of a delay between inhibitor onset and ITI initiation or that those that received ITI represented a poor risk group. The poor response rate to combination treatment most likely represents confounding by indication, that is, patients with worse disease or a more persistent inhibitor, were more likely to have received several treatments either simultaneously or sequentially. Other single case reports or small case series have reported a positive experience with combination treatments that have included immune suppressive agents such as corticosteroids and/or cyclophosphamide; however, case reports are prone to reporting bias [5–7].

In this cohort, there were only three F8 mutations that occurred in three or more subjects. Therefore, it was difficult to examine associations between genotype and inhibitor development. The mutation R2150H appears to be associated with a greater likelihood of having a persistent inhibitor. Santagastino et al. [8] reported a case of two brothers with R2150H mutation and an inhibitor that persisted for decades even in the absence of FVIII infusions. Despite numerous publications describing subjects with the R593C mutation, the responses to treatment and inhibitor duration in these cases are not clearly reported.

In conclusion, ∼22% of subjects with mild/moderate hemophilia A complicated by an inhibitor will spontaneously clear the inhibitor. Such spontaneous clearance appears to be more likely to occur among those with a low-titer inhibitor. However, it is possible that care providers are more likely to observe rather than treat patients with a low-titer inhibitor thereby allowing spontaneous clearance to occur. Conversely, care providers may be more likely to quickly treat patients with high-titer inhibitors thereby not only masking spontaneous clearance but also incorrectly attributing the clearance to the treatment initiated. Our findings suggest that when treatment is used, strategies that modulate the immune system have greater benefit than ITI and deserve closer study. However, a single approach is unlikely to be appropriate for all patients. It is possible that a combination F8 mutation along with other clinical characteristics, such as peak inhibitor titer, could better stratify those into who should be treated and who should be observed but further studies are needed.

Materials and Methods

This study is a secondary analysis of the cohort of case subjects enrolled as part of a retrospective case–control study approved by Institutional Review Boards at 16 participating sites. Subjects were enrolled during an 18-month period from July 2007 to December 2008 [4].

Case selection

Subjects were defined as individuals with mild or moderate hemophilia A (FVIII = 1–40%) based on local FVIII testing and a history of either two inhibitor titers > 1 BU/ml or one such inhibitor titer followed by the initiation of ITI (reference). Subjects were initially identified through the Centers for Disease Control and Prevention Universal Data Collection dataset. HTCs with potential subjects were invited to participate. Sixteen HTCs agreed to participate. Participating HTCs also reviewed their clinical data base and all patients that met eligibility criteria were invited to participate.

Data collection

After written informed consent was obtained, staff at participating HTC completed a standard data collection form for each subject. Data collected included the following: age at inhibitor onset, self-identified race (Black or African-American, White, Asian, or other); ethnicity (Hispanic or non-Hispanic); FVIII genotype if known; first inhibitor titer, peak inhibitor titer, effect of inhibitor on FVIII coagulant activity; the type of treatment [rituximab, ITI, or other treatments (cyclophosphamide, prednisone, or IVIg)] used to eradicate the inhibitor and the duration of the inhibitor (time from first inhibitor onset to a consistently negative inhibitor titer of <0.6 BU/ml). Subjects with a consistently negative inhibitor titer on all follow-up visits were classified as having cleared their inhibitor whereas all others were classified as having a persistent inhibitor. The duration of negative inhibitor titer or number of negative inhibitor titers was not recorded. An inhibitor was considered high-titer if it was >5 BU/ml [9].

Laboratory materials and methods

Blood was collected via venipuncture into two 4.5 ml vacuum-sealed tubes containing 3.2% sodium citrate. Red cells and buffy coat were collected for DNA extraction. Samples were either stored at 2–8°C and shipped to the Molecular and Hemostasis Laboratory at the Centers for Disease Control and Prevention (CDC) within 24 h or stored at −70°C until shipment after 24 h.

FVIII genotyping for all subjects was performed by the Molecular and Hemostasis Laboratory at the CDC except in the case of two subjects for whom historical FVIII genotype data were available. All exons, intron–exon junction regions, and the 3′ untranslated regions of F8 were sequenced in both directions. The VariantSEQr™ protocol was used for resequencing on a 3730 DNA Analyzer from Applied Biosystems. The PCR primers and M13 sequencing primers are described at with a few modifications to the PCR primers to enhance throughput and reproducibility. Data were analyzed with SeqScape®. Inversions of intron 22 and intron 1 in F8 were examined by PCR [10]. Mutations were assigned names based on both Human Genome Variation Society (HGVS) protein amino acid changes using Mutalyzer sequence variation nomenclature checker at from reference sequences NM_000132.3 and NP_000123.1 and traditional mature protein amino acid changes.

Data analysis

Continuous variables were reported as median and range. Wilcoxon rank test was used to compare medians. Categorical variables were reported as frequencies and were compared using the χ2 distribution. For sparse data, Fisher exact test was used to calculate P-values. Survival curves were created using the Kaplan–Meier method. Subjects who did not clear their inhibitor titer were right censored at the date of last positive inhibitor titer. Survival curves were compared using the Log-rank test. The influence of treatment to eradicate the inhibitor on the duration of inhibitor with adjustment for clinically relevant covariates (age, race, baseline FVIII coagulant activity, and peak inhibitor titer) was evaluated using Cox proportional hazards model using the Firth modification [11]. Age, baseline FVIII coagulant activity, and peak inhibitor titer were included in the model since: (1) older patients (> median 30 years) were more likely to receive rituximab and younger patients were more likely to receive ITI (P = 0.03); and (2) subjects who received rituximab alone trended toward having a higher median FVIII coagulant activity than those who received ITI alone (5.0 vs. 2.0%, P = 0.26) and a higherpeak inhibitor titer (26.5 vs. 6.5 BU/ml, P = 0.39). Race was included because of the well documented effect of race on risk of inhibitor development. All analyses were performed using SAS version 9.2 (Cary, NC).


The authors wish to acknowledge the following investigators and institutions for their enrollment of subjects in the study: Miguel A. Escobar, MD, University of Texas Health Science Center and The Gulf States Hemophilia and Thrombophilia Center, Houston, TX; Cynthia Gauger, MD, Nemours Children's Clinic, Jacksonville, FL; Mark T. Reding, MD, University of Minnesota, Minneapolis, MN; Alice J. Cohen, MD,, Newark Beth Israel Med. Ctr., Newark, NJ; Nigel S. Key, MBChB,, University of North Carolina, Chapel Hill, NC; Arthur R. Thompson, M.D., Ph.D., Puget Sound Blood Center, Seattle, WA; Marilyn J. Manco-Johnson, MD,, University of Colorado Denver and The Children's Hospital, Aurora, CO. The authors thank the Hemophilia Treatment Center Network for their stewardship of the UDC project and review of the manuscript, all of the nursing and research staff for the hard work in completing the regulatory documents, enrolling subjects, and completing the data collection, and the CDC laboratory staff for their work on the Bethesda assay and FVIII genotyping. These results were presented in part at the XXIII Congress of The International Society of Thrombosis and Haemostasis. The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention.

Author contributions: C.K. designed the research, analyzed the data,and wrote the paper, G.A., J.H., R.KJ, R.P., C.W., and G.Y. performed the research and critically revised the manuscript, and J.M.S. designed the research, analyzed the data and critically revised the manuscript.


  1. Top of page
  2. Supporting Information
  • 1
    Darby SC,Keeling DM,Spooner RJ, et al. The incidence of factor VIII and factor IX inhibitors in the hemophilia population of the UK and their effect on subsequent mortality, 1977–99. J Thromb Haemost 2004; 2: 10471054.
  • 2
    Hay CR,Ludlam CA,Colvin BT, et al. Factor VIII inhibitors in mild and moderate-severity haemophilia A. UK Haemophilia Centre Directors Organisation. Thromb Haemost 1998; 79: 762766.
  • 3
    Franchini M,Mengoli C,Lippi G, et al. Immune tolerance with rituximab in congenital haemophilia with inhibitors: A systematic literature review based on individual patients' analysis. Haemophilia 2008; 14: 903912.
  • 4
    Kempton CL,Soucie JM,Miller CH, et al. In non-severe hemophilia A the risk of inhibitor after intensive factor treatment is greater in older patients: a case-control study. J Thromb Haemost 2010; 8: 22242231.
  • 5
    Robbins D,Kulkarni R,Gera R, et al. Successful treatment of high titer inhibitors in mild hemophilia A with avoidance of factor VIII and immunosuppressive therapy. Am J Hematol 2001; 68: 184188.
  • 6
    Puetz JJ,Bouhasin JD. High-titre factor VIII inhibitor in two children with mild haemophilia A. Haemophilia 2001; 7: 215219.
  • 7
    Franchini M,Girelli D,Olivieri O, et al. Tyr2105Cys mutation in exon 22 of FVIII gene is a risk factor for the development of inhibitors in patientswith mild/moderate haemophilia A. Haemophilia 2006; 12: 448451.
  • 8
    Santagostino E,Gringeri A,Tagliavacca L, et al. Inhibitors to factor VIII in a family with mild hemophilia: Molecular characterization and response to factor VIII and desmopressin. Thromb Haemost 1995; 74: 619621.
  • 9
    White GCII,Rosendaal F,Aledort LM, et al. Definitions in hemophilia. Recommendation of the scientific subcommittee on factor VIII and factor IX of the scientific and standardization committee of the International Society on Thrombosis and Haemostasis. Thromb Haemost 2001; 85: 560.
  • 10
    Bagnall RD,Giannelli F,Green PM. Int22h-related inversions causing hemophilia A: A novel insight into their origin and a new more discriminant PCR test for their detection. J Thromb Haemost 2006; 4: 591598.
  • 11
    Heinze G,Schemper M. A solution to the problem of monotone likelihood in Cox regression. Biometrics 2001; 57: 114119.

Christine L. Kempton*, Geoff Allen†, Jeffrey Hord‡, Rebecca Kruse-Jarres§, Rajiv K. Pruthi¶, Christopher Walsh**, Guy Young††, J. Michael Soucie‡‡, * Emory University, Atlanta, Georgia, † Brown University, Providence, RI, ‡ Akron Children's Hospital, Akron, Ohio, § Tulane University, New Orleans, Louisiana, ¶ Mayo Clinic, Rochester, Minnesota, ** Mount Sinai Medical Center, New York, New York, †† Children's Hospital Los Angeles, Los Angeles, California, ‡‡ Division of Blood Disorders, Centers for Disease Control and Prevention, Atlanta, Georgia.

Supporting Information

  1. Top of page
  2. Supporting Information

Additional Supporting Information may be found in the online version of this article.

AJH_23269_sm_SuppTab1.docx19KSupporting Information Table 1. Clinical characteristics of all subjects

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.