Prognostic factors for remission of and survival in acquired hemophilia A (AHA): results from the GTH-AH 01/2010 study

Abstract

Acquired hemophilia A (AHA) is caused by autoantibodies against factor VIII (FVIII). Immunosuppressive treatment (IST) results in remission of disease in 60% to 80% of patients over a period of days to months. IST is associated with frequent adverse events, including infections as a leading cause of death. Predictors of time to remission could help guide IST intensity but have not been established. We analyzed prognostic factors in 102 prospectively enrolled patients treated with a uniform IST protocol. Partial remission (PR; defined as no active bleeding, FVIII restored >50 IU/dL, hemostatic treatment stopped >24 hours) was achieved by 83% of patients after a median of 31 days (range 7-362). Patients with baseline FVIII <1 IU/dL achieved PR less often and later (77%, 43 days) than patients with ≥1 IU/dL (89%, 24 days). After adjustment for other baseline characteristics, low FVIII remained associated with a lower rate of PR (hazard ratio 0.52, 95% confidence interval 0.33-0.81, P < .01). In contrast, PR achieved on steroids alone within ≤21 days was more common in patients with FVIII ≥1 IU/dL and inhibitor concentration <20 BU/mL (odds ratio 11.2, P < .0001). Low FVIII was also associated with a lower rate of complete remission and decreased survival. In conclusion, presenting FVIII and inhibitor concentration are potentially useful to tailor IST in AHA.

© 2015 by The American Society of Hematology.

Figures

Figure 1

Course and outcome of immunosuppressive therapy. *All 18 patients from the 2 sites not following the IST protocol were excluded regardless of their actual therapy. **Patients not yet achieving PR but continuously improving their FVIII activity until day 21 (while on prednisolone) or day 42 (while on cyclophosphamide and prednisolone) remained on the current treatment as per protocol. CTX, cyclophosphamide; P, prednisolone.

Figure 2

Pathways to complete remission. Arrows are proportional to patient numbers.

Figure 3

Adjusted time to PR and CR according to baseline FVIII activity. Cox proportional hazard model of time to end point with age, gender, underlying disorder, WHO-PS, inhibitor concentration, and baseline FVIII as categorical covariates. (A) PR (log rank P < .01). (B) Complete remission (log rank P < .05). Closed blue line, FVIII activity ≥1 IU/dL (n = 55); dashed red line, FVIII activity <1 IU/dL (n = 47).

Figure 4

Adjusted OS according to baseline characteristics. Cox proportional hazard model of OS with age, gender, underlying disorder, WHO-PS, inhibitor concentration, and baseline FVIII activity as categorical covariates. Patterns were drawn according to (A) baseline FVIII (<1 IU/dL n = 47 vs ≥1 IU/dL n = 55, P < .05), (B) malignancy status (present n = 13 vs absent n = 89, P < .05), and (C) WHO-PS (≤2 n = 64 vs >2 n = 38, P < .01).