The immunogenicity of ReFacto AF (moroctocog alfa AF-CC) in previously untreated patients with haemophilia A in the United Kingdom.

INTRODUCTION: Factor VIII inhibitor development is currently the most serious complication of the treatment of haemophilia A. Differences in manufacturing and the molecular structure of brands of recombinant factor VIII have led to speculation that concentrates may differ in immunogenicity. This has led to a regulatory focus on the immunogenicity of factor VIII concentrates both before and after licensure. AIM: To investigate the immunogenicity of ReFacto AF post licensure in a real-world setting in previously untreated patients (PUPs) treated exclusively with this product until at least 50 exposure days (EDs). METHODS: The United Kingdom Haemophilia Centre Doctors' Organisation (UKHCDO) National Haemophilia Database (NHD) identified a consecutive cohort of patients with severe haemophilia A (<0.01 IU/L) whose first treatment was with ReFacto AF, monitored time to inhibitor development and described associated risk factors. RESULTS: One hundred and three boys reached 50 EDs within the study period, of whom 35 developed an inhibitor (P(t ≤ 50) = 0.33, [95% CI: 0.25-0.43]), of which 15 (P(t ≤ 50) = 0.16, [95% CI: 0.10-0.25]) were high titre. Inhibitors arose after a median (interquartile range) 11 (7-16) EDs. Inhibitors were significantly associated with high-risk mutations and non-significantly associated with non-white ethnicity. Inhibitors were negatively associated with a family history of haemophilia A. High-titre inhibitors were significantly associated with a family history of inhibitors. CONCLUSION: Inhibitor incidence in a single country population of ReFacto AF PUPs was similar to that previously described. Low- and high-titre inhibitors were detected after a similar number of EDs, contrasting with previous data, probably reflecting standardized inhibitor monitoring within the United Kingdom.

Efficacy and safety of Nuwiq® (human-cl rhFVIII) in patients with severe haemophilia A undergoing surgical procedures.

INTRODUCTION: Haemophilia A patients are at a high risk of excess bleeding during surgeries. The aim of haemostatic therapy during the perioperative period is to normalize FVIII level perioperatively and postoperatively to maintain normal haemostasis until wound healing is complete. AIMS/METHODS: To examine the efficacy of Nuwiq® (simoctocog alfa, human-cl rhFVIII), a 4th generation recombinant FVIII produced in a human cell line, for surgical prophylaxis in patients with severe haemophilia A. This analysis assessed the efficacy of Nuwiq® during surgical procedures and in the postoperative period in seven clinical studies of previously treated patients (PTPs) with severe haemophilia A. RESULTS: Thirty-six patients, aged 3-55 years, received surgical prophylaxis with Nuwiq® for 60 surgeries (28 major and 32 minor). Efficacy was evaluated for 52 surgeries (25 major and 27 minor). The success rate of Nuwiq® treatment was 98.1% (51 of 52 evaluated surgeries); haemostatic efficacy was assessed as "excellent" or "good" in all but one major surgery (assessed as "moderate"). The number of infusions ranged from 1 to 19 for minor surgeries and from 3 to 76 for major surgeries. The median (range) daily doses were 42.0 (28.2-100.9) IU kg-1 for minor surgeries and 69.3 (43.3-135.6) IU kg-1 for major surgeries. There were no serious treatment-related adverse events, and none of the patients developed FVIII inhibitors. CONCLUSIONS: The results of this pooled analysis show that Nuwiq® was efficacious in maintaining haemostasis during and after major and minor surgical procedures in PTPs with severe haemophilia A.

Pharmacokinetics, safety and efficacy of a recombinant factor IX product, trenonacog alfa in previously treated haemophilia B patients.

INTRODUCTION: Trenonacog alfa (IB1001) is a recombinant factor IX (rFIX) manufactured in Chinese hamster ovary (CHO) cells. IB1001 was evaluated in a multicentre clinical trial with haemophilia B patients. AIM: The aim was to establish IB1001 pharmacokinetic non-inferiority to comparator rFIX, safety and efficacy in previously treated patients (PTPs) with haemophilia B. METHODS: Subjects were severe or moderately severe haemophilia B adult and adolescent PTPs with no history of FIX inhibitors. RESULTS: IB1001 PK non-inferiority to comparator rFIX was demonstrated through ratio of AUC0-∞ in 32 subjects. IB1001 was well tolerated in all 76 treated subjects; the most common adverse drug reaction was headache (2.6% of subjects) and there were no reports of FIX inhibitors. Transient non-inhibitory binding FIX antibodies and anti-CHO cell protein antibodies developed in 21% and 29% of subjects respectively; no safety concerns were associated with development of these antibodies. Prophylaxis (mean duration ± SD: 17.9 ± 9.6 months, mean dose: 55.5 ± 12.9 IU/kg, median 1.0 infusion per week) was effective in preventing bleeds (median annual bleed rate: 1.52, interquartile range: 0.0-3.46). One or two IB1001 infusions resolved 84% of the bleeds, while for 84% of treatments haemostatic efficacy of IB1001 was rated excellent or good. IB1001 haemostatic efficacy for all 19 major surgeries was rated adequate or better than adequate. CONCLUSIONS: IB1001 is safe and efficacious for treatment of bleeds, routine prophylaxis and perioperative management in haemophilia B patients.

Use of the UKHCDO Database for a postmarketing surveillance study of different doses of recombinant factor VIIa in haemophilia.

INTRODUCTION: Recombinant factor VIIa (rFVIIa) is recommended in Europe at standard (3 × 90 µg kg-1 ) or high (1 × 270 µg kg-1 ) doses. When granting the license for the high dose, the European Medicines Agency (EMA) requested postmarketing surveillance for thrombosis. This was conducted by the United Kingdom National Haemophilia Database (NHD) on behalf of Novo Nordisk and the EMA. AIM: To assess the use and safety of rFVIIa utilizing prospective data collected by the NHD (1 January 2008 to 30 June 2011). RESULTS: Data were obtained from 67 haemophilia A/B patients with inhibitors treated for 1057 bleeds and 31 acquired haemophilia patients treated for 70 bleeds. Initial rFVIIa dose was categorized post hoc as low (<90 µg kg-1 ), intermediate (≥90-<180 µg kg-1 ) or high (≥180-<270 or ≥270 µg kg-1 ). For haemophilia A/B, high and lower initial rFVIIa dose was used for 38.4% and 51.4% of episodes, respectively, while for acquired haemophilia, the values were 11.4% and 77.1% respectively. Median initial doses were higher for haemophilia A/B (146.3 µg kg-1 ) than acquired haemophilia (90.5 µg kg-1 ). A single administration of rFVIIa was the most frequently used regimen for haemophilia A/B, in contrast with standard recommendations and previous reports. For acquired haemophilia, most episodes were treated with multiple doses. No adverse drug reactions or thromboembolic events were reported for any rFVIIa dose. CONCLUSION: The novel use of a national database for postmarketing surveillance has demonstrated acceptable safety for all recommended doses of rFVIIa.

The haemtrack home therapy reporting system: Design, implementation, strengths and weaknesses: A report from UK Haemophilia Centre Doctors Organisation.

INTRODUCTION: Haemtrack is an electronic home treatment diary for patients with inherited bleeding disorders, introduced in 2008. It aimed to improve the timeliness and completeness of patient-reported treatment records, to facilitate analysis of treatment and outcome trends. The system is easy to use, responsive and accessible. METHODS: The software uses Microsoft technologies with a SQL Server database and an ASP.net website front-end, running on personal computers, android and I-phones. Haemtrack interfaces with the UK Haemophilia Centre Information System and the National Haemophilia Database (NHD). Data are validated locally by Haemophilia Centres and centrally by NHD. Data collected include as follows: treatment brand, dose and batch number, time/date of bleed onset and drug administration, reasons for treatment (prophylaxis, bleed, follow-up), bleed site, severity, pain-score and outcome. RESULTS: Haemtrack was used by 90% of haemophilia treatment centres (HTCs) in 2015, registering 2683 patients using home therapy of whom 1923 used Haemtrack, entering >17 000 treatments per month. This included 68% of all UK patients with severe haemophilia A. Reporting compliance varied and 55% of patients reported ≥75% of potential usage. Centres had a median 78% compliance overall. A strategy for progressively improving compliance is in place. Age distribution and treatment intensity were similar in Haemtrack users/non-users with severe haemophilia treated prophylactically. CONCLUSION: The Haemtrack system is a valuable tool that may improve treatment compliance and optimize treatment regimen. Analysis of national treatment trends and large-scale longitudinal, within-patient analysis of changes in regimen and/or product will provide valuable insights that will guide future clinical practice.

Evaluation of the use of rotational thromboelastometry in the assessment of FXI deficency.

INTRODUCTION: The absence of a reliable clinical test to predict bleeding tendency leaves factor XI (FXI)-deficient individuals at risk of overtreatment or under treatment. AIM: To assess whether rotational thromboelastometry has value in detection of FXI deficiency and identification of bleeding tendency. METHODS: Thromboelastometry was measured in whole blood and platelet-rich plasma (PRP) samples containing corn trypsin inhibitor (CTI) from controls (n = 50) and FXI-deficient individuals (n = 93) at tissue factor (TF) 0.12 pm. The effect of tissue plasminogen activator was also assessed. For analysis, FXI-deficient individuals were divided into bleeders (n = 24) and non-bleeders (n = 44) based on experience of tonsillectomy and/or dental extraction prior to diagnosis. RESULTS: In whole blood, thromboelastometry distinguished those with major FXI deficiency (FXI:C ≤ 15 IU dL-1 ) but not partial deficiency from control populations, but did not identify bleeding phenotype. In PRP, bleeders had significantly longer clot formation time [CFT; 434 ± 179 s vs. 277 ± 70 s (mean ± SD); P < 0.05] and smaller α angle [43.8 ± 9.5° vs. 52.4 ± 5.8° (mean ± SD); P < 0.05] compared to non-bleeders. However, these parameters were found to depend on multiple additional variables and on an individual basis, ROC analysis showed test specificity for bleeding phenotype identification to be only 38.5% at 100% sensitivity: CFT (area under first derivative curve: AUC = 0.8091, P = 0.0014), α angle (AUC = 0.7804, P = 0.006). CONCLUSION: Thromboelastometry in PRP with CTI samples triggered with TF 0.12 pm was able to distinguish between bleeders and non-bleeders in FXI deficiency, but poor specificity restricts its clinical application as a test to identify bleeding phenotype. Further technical advances to the assay may allow better discrimination.

Evaluation of the use of global haemostasis assays to monitor treatment in factor XI deficiency.

INTRODUCTION: Previous guidelines recommend that FXI:C levels should be used to monitor FXI replacement in factor XI (FXI) deficiency. However, FXI:C levels do not correlate with bleeding tendency in this disorder and may not be the optimal test by which to monitor and determine further treatment in the postoperative period. AIM: To assess whether the thrombin generation assay (TGA) and rotational thromboelastometry can be used to monitor FXI replacement peri-operatively in FXI deficiency and to determine if changes in FXI:C levels correlate with changes in thrombin generation and clot formation parameters following treatment with solvent-detergent fresh frozen plasma (SD-FFP). METHODS: The TGA and rotational thromboelastometry were used to measure thrombin generation and clot formation in 11 adults with FXI deficiency who were treated with either SD-FFP (n = 8) or FXI concentrate (n = 3) as prophylaxis peri-operatively. Blood samples were taken pre- and 30 min post-treatment. RESULTS: Global haemostasis assays can be used to measure the effect of FXI replacement with SD-FFP or FXI concentrate in FXI deficiency. Both treatment types improved thrombin generation and clot formation. However, the remaining response to treatment at 24 h post SD-FFP was variable and changes in FXI:C levels were not predictive of changes in thrombin generation/thromboelastometry parameters after treatment with SD-FFP. CONCLUSION: Global haemostasis assays may provide a more reliable means of monitoring SD-FFP treatment with the potential to prevent individuals receiving unnecessary treatment, however, their clinical use in decision making needs to be tested in a larger prospective study.

The impact of sport on health status, psychological well-being and physical performance of adults with haemophilia.

BACKGROUND: There is increasing recognition that sport is important for individuals with haemophilia; however, there remains a paucity of data of the importance of this in adults, many of whom already have joint pathology related to childhood bleeds and treatment access. This multicentre, cross-sectional study presents the impact of sport on health-related quality of life (HRQoL), physical performance and clinical outcomes in adults with haemophilia. RESULTS: Fifty adults aged 35.12±14.7 with mild (n = 12), moderate (n = 10), or severe (n = 28) haemophilia A (70%) or B (30%) from four haemophilia centres across the United Kingdom participated in the study. A total of 64% were overweight/obese according to their BMI; median orthopaedic joint scores using the WFH Orthopaedic Joint Score (OJS) were 6 (range 0-48). On a VAS pain scale (range of 0-10), patients reported mean score of 5.66 ± 2.4. 36% of participants reported not doing any sport, mainly due to their physical condition. However, 64% of participants reported undertaking sporting activity including contact sports, mostly twice per week in average 4 h week(-1) . Participating in sport did not have a statistically significant impact on HRQoL; except in the domain 'sport and leisure' of the Haem-A-QoL. Patients doing more sport reported significantly better HRQoL than those doing less sport (P < 0.005). Those doing sport for more than 4 h week(-1) had a significantly better physical performance than patients doing less sport (assessed with Hep-Test-Q). Encouraging physical activity and sport in older patients with haemophilia may have a direct impact on their HRQoL; thus, education about sport activity should be incorporated into routine haemophilia care.

In vitro comparison of the effect of two factor XI (FXI) concentrates on thrombin generation in major FXI deficiency.

INTRODUCTION: Bleeding risk in factor XI (FXI) deficiency following surgery may be reduced by treatment with either of two FXI concentrates, but indications for their use are unclear and treatment has been associated with thrombosis. AIM: To quantify and compare the effects of two different FXI concentrates on thrombin generation (TG) in major FXI deficiency (FXI:C < 15 IU dL(-1) ). METHODS: Thrombin generation was measured in controls (n = 50), FXI-deficient individuals pre and post in vitro spiking with FXI concentrates (n = 10), and in ex vivo samples following treatment with FXI concentrate (n = 3). RESULTS: Thrombin generation was significantly impaired in FXI deficiency but improved following FXI replacement in vitro and in vivo. LFB Hemoleven(®) had greater effect on TG than BPL FXI concentrate in vitro (equivalent in vivo doses 10, 20 and 30 U kg(-1) ): higher endogenous thrombin potential (ETP) (P < 0.0001), peak height (P < 0.01) velocity (P < 0.0002) and shorter lag time and time to peak (both P < 0.003). Some measurements with LFB Hemoleven(®) exceeded the reference range. At lower dose (5 U kg(-1) ), BPL FXI concentrate normalized all TG parameters and LFB Hemoleven(®) normalized the ETP but exceeded the reference range with other parameters. CONCLUSION: Both FXI concentrates improve TG in vitro in major FXI deficiency but differ in dose response, and for both products, doses lower than previously recommended normalized TG in vitro. Comparison of in vitro spiked and ex vivo samples suggest that in vitro results could be used to estimate an expected in vivo response to FXI replacement.

The incidence of factor VIII inhibitors in severe haemophilia A following a major switch from full-length to B-domain-deleted factor VIII: a prospective cohort comparison.

Although it has been suggested that switching of factor VIII (FVIII) products may increase inhibitor formation this is disputed. Half of UK patients changed rFVIII brands because of national contracting in 2010, presenting an opportunity to compare inhibitor incidence of switchers with non-switchers. Centres were requested to test all the patients for inhibitors prior to the switching date and 6-monthly thereafter. Positive and negative inhibitor test data were also collected to analyse for testing bias. A total of 1198 patients with severe haemophilia A and treated with Advate, Kogenate/Helixate or Refacto AF preswitch were included in the analysis, of whom 516 switched to Refacto-AF and 682 did not switch products. Five new inhibitors were reported amongst previously treated patients (>50 exposure days) with a median titre at the time of detection of 1.25 BU mL(-1) (IQR 0.7-23.05). One inhibitor occurred in a non-switcher using Kogenate, an incidence of 1.5 per 1000 treatment-years (95% CI 0.2-10.5). Four inhibitors arose in patients who had switched from Kogenate (two) or Advate (two) to ReFacto-AF, an incidence of 7.8 per 1000 treatment-years (95% CI 2.9-20.8). These incidence rates did not differ significantly from one another (incidence rate ratio 5.3 (95% CI 0.5-260.3) or from the historical rate of 6.05 inhibitors/1000 treatment-years (95% CI 5.18-7.06). Only one inhibitor (non-switcher) persisted. Non-switchers were significantly older (P = 0.03), and used significantly less FVIII per year (P = 0.005) prior to switching. Following switching, factor usage increased similarly (P = 0.53) in both groups. Switching from FLRFVIII to Refacto-AF (BDDRFVIII) was not associated with an increased inhibitor development.

Central venous access device (CVAD) complications in Haemophilia with inhibitors undergoing immune tolerance induction: Lessons from the international immune tolerance study.

INTRODUCTION: Central venous access devices (CVADs) are frequently required as stable long-lasting venous access in children with haemophilia, especially those requiring immune tolerance induction (ITI) for inhibitors. CVAD infection is one of the most frequently reported catheter-related complications in this patient population. AIM: Detailed review of CVAD complications from the International ITI (I-ITI) study and analysis of potential risk factors for such complications. METHODS: Retrospective analysis of prospectively obtained data from the I-ITI study primarily focused on CVAD-related complications. RESULTS: A total of 115 children were recruited and 183 CVADs were placed in 99 subjects resulting in 121,206 CVAD-days observed on-study. A total of 124 CVAD infections were reported in 41 of 99 (41%) subjects with an overall infection rate of 0.94 per 1000 CVAD-days (interquartile ranges 0-1.7). A similar number of infections were observed in the two treatment arms (median: 2 and 3 in high dose and low dose respectively). Infections occurred more frequently in the presence of external catheters than with fully implanted catheters (P = 0.026). Infected patients were significantly younger at the time of CVAD insertion (median age: 22 vs. 25 months, P = 0.020). Patients with Gram-positive infections were also significantly younger than those with Gram-negative infections (median age: 17 vs. 25 months, P < 0.0001). CONCLUSION: CVAD infection was the most common complication observed in children with severe haemophilia and inhibitors in the frame of the I-ITI study. Younger age at CVAD insertion and external CVAD were associated with higher risk for infection. ITI outcome was unaffected by CVAD infections.

Purchasing factor concentrates in the 21st century through competitive tendering.

The increasing intensity of treatment, the widespread adoption of factor VIII and IX prophylaxis and increasing usage over the past decade have led to haemophilia becoming an almost uniquely expensive condition to treat. The average adult with severe haemophilia A in the UK used 250,000 IU of factor VIII in 2011/2012, at a cost in excess of £ 100,000 p.a. The cost to the end-user may be considerably higher than this for some US patients supplied by home care companies with high on-costs. This has led to a high level of administrative scrutiny of treatment and an imperative to procure clotting factor concentrates more efficiently and collectively. National procurement schemes have run successfully in various countries and will become commoner. The UK system of procurement is described. This system, following EU procurement rules, evaluated products technically and by price. The price of bioequivalent products was determined by reverse e-auction. Considerable cost reductions were achieved whilst retaining all suppliers and maintaining a degree of prescribing freedom. Elements of this system could be more widely applied.

How I treat inhibitors in haemophilia.

The management of patients with inhibitors is the greatest challenge facing haemophilia health professionals. Immune tolerance induction (ITI) can be successful in eliminating the inhibitor in the majority of patients, provided it is started soon after the inhibitor develops and the titre of the inhibitor is <10 BU at commencement of ITI. Acute bleeding is treated using one of two bypassing agents, which exhibit similar efficacy and safety. Surgery in inhibitor patients is challenging and should only be carried out in experienced centres.

Experience with Optivate®, a new high purity concentrate of factor VIII with von Willebrand factor, in patients undergoing surgery.

The efficacy and safety of Optivate(®) was assessed in 23 surgical operations, orthopaedic (12) including 5 revision arthroplasties, ophthalmic (1), ENT (1), dental (6), liver biopsy (2), and removal of portacath (1) on 15 teenagers and adults with severe haemophilia A. The preoperative dose was calculated to raise the FVIII concentration to 100 IU dL(-1). Subsequent doses were targeted to maintain at least 50 IU dL(-1). There were 11 major and 12 minor operations categorized as receiving intensive replacement therapy for ≥ 5 days or < 5 days respectively. The median preoperative dose was 50.4 (range 18.2-88.2) IU kg(-1). The median incremental recovery based on this first dose in 10 procedures (5 patients) was 2.9 (range 2.4-3.4 IU dL(-1) ) per IU kg(-1). The daily doses decreased during the first 4 days of the study. The patients in this study received 173 infusions in total. Outcome was 'good' or 'excellent' for 19 (83%) of 23 operations, 'uncertain' in three procedures because an antifibrinolytic agent was used as well and for one procedure outcome was not assessed. Tolerance was good. There were no excessive bleeds, no inhibitors and no virus transmissions.

The mutation spectrum associated with type 3 von Willebrand disease in a cohort of patients from the north west of England.

Type 3 von Willebrand disease (VWD) is a severe autosomal recessive inherited bleeding disorder. In affected individuals the underlying von Willebrand factor gene (VWF) mutations frequently remain uncharacterized. The aim of this study was to investigate the molecular basis of type 3 VWD in patients (11 Caucasians and 9 of Asian origin) attending the haemophilia centres at Central Manchester NHS Trust. A combination of DNA sequencing of VWF genomic and complementary DNA was performed to identify mutations in the patient cohort. Fifteen different VWF mutations were identified at the genomic DNA level: two gene conversion events, three nonsense, three frameshift, one missense, two splice site, one insertion-deletion and three deletion mutations. Homozygosity or compound heterozygosity for mutations was present in 15 of the 20 patients. In the remaining five individuals, heterozygosity for a single VWF mutation was identified in four cases and one patient had no detectable VWF mutation. Analysis of platelet-derived VWF RNA from these five individuals revealed heterozygosity for a deletion of exons 4 and 5 in four cases. The remaining patient was heterozygous for a three base deletion which had already been identified at the DNA level. Overall the observed VWF genotype explained the phenotype in 18 of the 20 patients investigated. In genetic studies in type 3 VWD, if VWF mutations are not detected at the DNA level, RNA analysis should be performed to search for intronic mutations, heterozygous deletions or aberrant splicing/post-transcriptional events. However, this may still not explain all cases of previously phenotypically diagnosed type 3 VWD.

Phase 1, single-dose escalating study of marzeptacog alfa (activated), a recombinant factor VIIa variant, in patients with severe hemophilia.

Essentials Marzeptacog alfa (activated) [MarzAA] is a novel variant of activated human factor VII. A phase 1 dose escalation trial of MarzAA was conducted in subjects with severe hemophilia. MarzAA was safe and tolerated at intravenous doses up to 30 µg kg-1 Data observed support further trials for hemophilia patients with inhibitors to factors VIII/IX. SUMMARY: Background Marzeptacog alfa (activated) (MarzAA), a new recombinant activated human factor VII (rFVIIa) variant with four amino acid substitutions, was developed to provide increased procoagulant activity and a longer duration of action in people with hemophilia. Objectives To investigate the safety, tolerability, immunogenicity, pharmacokinetics (PK) and pharmacodynamics (PD) of single ascending intravenous bolus doses of MarzAA in non-bleeding patients with congenital hemophilia A or B with or without inhibitors. Methods This international, phase 1, open-label study (NCT01439971) enrolled males aged 18-64 years with severe hemophilia A or B, with or without FVIII or FIX inhibitors. Subjects were assigned to single-dose MarzAA cohorts (0.5, 4.5, 9, 18 or 30 µg kg-1 ). Blood sampling was performed predose and postdose, and subjects were monitored for 60 days postdose. Safety endpoints included adverse events, vital sign changes, electrocardiograms, laboratory abnormalities, and immunogenicity; secondary endpoints included evaluation of PK and PD. Results Overall, in 25 patients, MarzAA was well tolerated at all dose levels tested, and was not associated with dose-limiting toxicity. No treatment-emergent severe or serious adverse events occurred. MarzAA showed linear dose-response PK across the 4.5-30 µg kg-1 dose range, with a terminal half-life of ⁓ 3.5 h. Dose-dependent shortening of the activated partial thromboplastin time and prothrombin time, and evidence of an increase in peak thrombin as determined with a thrombin generation assay, were observed at all doses. Conclusions MarzAA was tolerated at doses up to 30 µg kg-1 . The safety profile and pharmacological effects observed support further clinical trials for the treatment of hemophilic patients with inhibitors.

The incidence of factor VIII and factor IX inhibitors in the hemophilia population of the UK and their effect on subsequent mortality, 1977-99.

BACKGROUND: Previous studies of the development of inhibitors and their impact on mortality have been small. OBJECTIVES: To examine the development of inhibitors in people with hemophilia in the UK and their effect on subsequent mortality. PATIENTS: 6078 males with hemophilia A and 1172 males with hemophilia B registered in the UK Haemophilia Centre Doctors' Organisation database, 1977-98. RESULTS: In severe hemophilia A inhibitors developed at rates of 34.4, 5.2 and 3.8 per 1000 years at ages <5, 5-14 and 15+years; cumulative risks at ages 5 and 75 were 16% and 36%. In hemophilia A the rate of inhibitor development decreased during 1977-90, but increased during the 1990s. In severe hemophilia B inhibitors developed at rates of 13.3 and 0.2 per 1000 years at ages <5 and 5+ and cumulative risks at ages 5 and 75 were 6% and 8%. With HIV, inhibitor development did not increase mortality. In severe hemophilia without HIV, inhibitor development doubled mortality during 1977-92, but during 1993-99 mortality was identical with and without inhibitors. In severe hemophilia without HIV but with inhibitors, mortality from causes involving bleeding decreased during 1977-99 (P = 0.001) as did mortality involving intracranial hemorrhage (P = 0.007). CONCLUSIONS: These data provide estimates of the rate of inhibitor development in hemophilia A and hemophilia B, and they show that the rate of inhibitor development has varied over time, although the reasons for this remain unclear. They also show that in severe hemophilia the substantial increase in mortality previously associated with inhibitors is no longer present.