UKHCDO Acquired Haemophilia Study: A Complete National Cohort.

The literature on acquired haemophilia A (AH) predominantly consists of tertiary referral centre patients. This may have introduced referral and reporting bias with younger and more severely affected patients over-represented. This study reports all patients with AH presenting between May 2001 and April 2003 from 255 out of the 256 haematology departments in the UK. A total of 173 patients were reported, incidence 1.5/million/year. Age related incidences for the 0–15, 16–64, 65–89 and >89 year age groups were 0.04, 0.28, 7.0 and 10.3/million/year Compared to previous cohorts, patients were older, more likely to have malignancy (15% of patients) and less likely to be post partum (1.7% of patients, 1:350 000 UK births). The bleeding phenotype was variable, 33% required no haemostatic treatment and 8% had fatal bleeds. Fatal bleeds occurred between 1 and 146 days. Early fatal bleeds were gastrointestinal and lung haemorrhage (days 1–4) and later fatal bleeds were intracranial and retroperitoneal. Immunosuppression was at the discretion of the local clinician. Treatment groups were similar in age and sex and presenting FVIII levels and inhibitor titres. Complete remission (CR) (defined as FVIII normal, inhibitor undetectable and off immunosuppression) was achieved in 77% of steroid treated patients (n=41) after a median of 53 days, 76% of steroid and cyclophosphamide treated patients (n=88) (median 46 days) and 33% of cyclophosphamide treated patients (n=9) (median 210 days). Remission was not related to age or presenting FVIII level or inhibitor titre. Death occurred in 50% of steroid treated patients after a median of 79 days, 41% of steroid and cyclophosphamide treated patients (median 78 days) and 33% of cyclophosphamide treated patients (median 232 days). Survival was not related to presenting FVIII level or inhibitor titre but was higher in younger patients. Intravenous immunoglobulin (IVIG) was ineffective. CR was attained in 73% of patients treated with IVIG (n=48) compared with 78% not treated with IVIG (n=78). Relapse occurred in 22% of patients treated with steroids and 24% of patients treated with steroids and cyclophosphamide between 1 week and 3 years (median 3 months) after stopping immunosuppression. Second CR was induced in 63% of patients with a further 19% requiring maintenance immunosuppression to sustain remission. These data, which are unbiased with regard to referral and reporting practice, from the largest and most representative cohort of AH patients so far collected, do not support the commonly held views that steroids and cyclophosphamide lead to better outcomes than steroids alone, that IVIG is a useful adjunct therapy or that the FVIII or inhibitor levels at diagnosis should guide therapeutic decisions. The high relapse rate and variable bleeding phenotype, ranging from fatal to requiring no haemostatic therapy, are highlighted. Treatment guidelines based on data from referral centre patients may not be applicable to patients with AH presenting at other centres who are likely to be older, less able to tolerate immunosuppression and to have milder bleeding. The death rate in AH is high but not due to bleeding in most cases. Studies should investigate both the efficacy and toxicity of treatment regimens.