Clinical and Economic Outcome Analysis of First-Line Immunosuppression in Acquired Hemophilia a

Introduction

Acquired hemophilia A (AHA) is a rare bleeding disorder in which inhibitory antibodies neutralize factor VIII (FVIII) activity leading to high morbidity and mortality from both bleeding and complications of immunosuppressive treatment (IST). Individuals with AHA are vulnerable to the adverse effects of IST as they tend to be older with multiple comorbidities. Rituximab may be less toxic than traditional IST, cyclophosphamide, but is expensive and not always accessible. Both are recommended as first-line options in high-risk patients (FVIII <1% or inhibitor >20 BU/mL) by the current international AHA treatment guidance (Tiede, Haematologica, 2020). This study compared the efficacy, safety, and economic burden of first-line IST regimens in patients with AHA in Canada.

Methods

This national retrospective cohort included adults >18 years old with AHA treated between January 2000 and November 2022 with at least 6 months of follow-up data available. Treatment groups were: steroid alone (A), steroid + rituximab (B), steroid + cyclophosphamide (C), and steroid + rituximab + cyclophosphamide (D). Kruskal Wallis and Chi-Square tests compared variables between groups, and Kaplan-Meier curves with log-rank tests compared time-to-event endpoints. Cost analysis used data from Canadian Blood Services, local hospitals, and provincial drug formularies.

Results

The study included 127 patients from 8 Hemophilia Treatment Centres; 26 in group A, 20 in B, 55 in C, 26 in D. At diagnosis, median age (71; IQR 62, 80) and Charlson Comorbidity Index (4; IQR 3, 5) did not differ between groups, but the proportion of high-risk patients did (A: 27%, B: 80%, C: 51%, D: 85%). Median time from diagnosis to last follow-up was 30.4 (IQR 14.8, 54.2) months (mo) overall, and median duration of IST was similar between groups (A and B: 2.8mo, C and D: 3.4mo).

Overall, 89% of patients achieved complete remission (FVIII level >50%, negative inhibitor, and cessation of IST; CR) with no significant difference between groups (A: 88.5%, B: 85.0%, C: 89.1%, D: 92.3%), but median time to CR was longest in group D (A: 3.0mo, B: 3.1mo, C: 3.3mo, D: 4.6mo). Median time to first negative inhibitor (A: 1.5mo, B: 1.8mo, C: 1.7mo, D: 3.1mo, p<0.01), and first normal FVIII (>50%) (A: 1.4mo, B: 1.4mo, C: 1.6mo, D: 2.8mo, p<0.01) differed significantly between groups. Median time to relapse was 1.9mo (IQR 1.2, 9.3), 85% of relapses occurred within 1 year, and most were in group A (A: 42%, B: 4%, C: 31%, D: 23%, p=0.01). There was no significant difference in median overall survival between groups (11.3 years). Death from infection was 0% (A), 25% (B), 23% (C), 33% (D) and there was only 1 death from bleeding (25%, B).

Of 395 total adverse events (AE), there were 113 non-infection grade 3/4 AE, 103 hospitalizations, and 55 infections. Non-infection grade 3/4 AE differed significantly between groups (A: 17.2%, B: 21.9%, C: 33.8%, D: 35.9%, p=0.01) and were driven by neutropenia and glucose intolerance. Incidence of infection was: A (14%), B (11%), C (15%), and D (15%). Incidence and duration of hospitalizations did not differ between groups. Overall, there were 261 bleeding events (124 ISTH major, 109 ISTH clinically-relevant non-major). Group D had the highest mean bleeds per patient (A: 2, B: 2, C: 1.8, D: 2.7), proportion of major bleeds (A: 47%, B: 40%, C: 45%, D: 57%), and number of bleeds requiring blood transfusion (A: 41%, B: 30%, C: 39%, D: 61%).

The most costly regimen (mean per patient) was group D ($384,000) followed by C ($143,000), A ($122,000), and B ($133,000). Most of the cost was from hemostatic treatment (A: $79,490, B: $95,933, C: $113,247, D: $349,020), hospitalizations (A: $41,118, B: $24,718, C: $26,952, D: $13,875), and IST (A: $57, B: $10,725, C: $1,048, D: $16,875). Bypassing agents, rFVIIa and FEIBA, comprised 94% of the hemostatic treatment costs.

Conclusion

In this Canadian cohort, all IST regimens had similar overall survival and proportions of CR. Those treated with steroids and rituximab (B) had the lowest frequency of relapse, and lower incidence of infection and non-infection grade 3/4 AE than regimens containing cyclophosphamide. To our knowledge, this is the first study to evaluate the comprehensive economic burden of AHA treatment. Overall, the majority of costs were from hemostatic treatments, not from IST, and differences between regimens may be due to proportions of high risk patients and the frequency and severity of bleeding events.