Large Deletion in the Factor VIII Gene Is a Predictor of Immune Tolerance Induction Failure in People with Severe and Moderately-Severe Hemophilia a and High-Responding Inhibitors

Introduction Immune tolerance induction (ITI) is a therapeutic approach to eradicate inhibitors against factor VIII (FVIII) in people with inherited hemophilia A (PwHA). The success of ITI is highly variable, ranging from 60% to 80% across studies. Beyond inhibitor titers, other predictors of ITI outcome are unknown. Deleterious variants in FVIII gene (F8) are well established risk factors for alloantibodies development in PwHA, although few studies have investigated the role of F8 variants on ITI outcome. Therefore, we aimed to comprehensively analyze the association of F8 pathogenic variants on ITI outcome in severe PwHA with high-responding inhibitors.

Methods We included severe (FVIII < 1 international units [IU]/dL) and moderately-severe (FVIII 1-2 IU/dL) unrelated PwHA and high-responding inhibitors who completed ITI from a large, admixed population of two well-characterized cohorts - the HEMFIL and the Brazilian Immune Tolerance (BrazIT) studies. We collected socio-demographic, clinical and laboratory data. ITI outcomes were defined according to previous definitions as failure, partial and total successes. Inversions of intron 1 and 22 (Inv22) were detected by polymerase chain reaction (PCR), and high-throughput sequencing approaches were used to unveil the additional F8 variants. The association between F8 pathogenic variants and ITI outcome was adjusted for inhibitor levels. To investigate whether F8 variants associated with inhibitor development were also related to ITI outcome, variants were categorized as “high-risk”, “intermediate-risk”, and “low-risk” categories. These categories for inhibitor development were then compared with the outcomes of ITI (total and partial successes and failure) using data from this study.

Results We included158 PwHA, median age 6.6 years at ITI start, 90.5% were severe (Table 1). Inv22 was the most prevalent variant (55.1%) (Table 1). In comparison with Inv22-1, the risk of ITI failure was about 9 times higher (adjusted odds ratio [adjOR] 9.29; 95% confidence interval [95% CI] 1.95-53.70) among carriers of large deletions (Table 2). Conversely, Inv22-2 was associated with favorable ITI outcome in a univariate analysis (OR 0.15; 0.01-0.84), and after adjustment (adjOR 0.32; 95% CI 0.02-1.96) (Table 2). F8 deleterious variants sorted as high-risk and intermediate-risk according to a previously published classification on inhibitor development were associated with failure and successful outcomes, respectively.

Conclusion Our study showed that F8 large deletions are independent predictors of ITI failure, and Inv22-2 is likely to be a predictor of successful ITI. We found a correspondence between variants classified as high-risk and intermediate-risk to inhibitor development with ITI failure and success, respectively. We suggest that F8 genotyping should be considered before indication of ITI, as ITI outcome can vary according to individual variant burden.