Prevalence of Anti-Human Factor VIII Binding IgG Antibodies in Sickle Cell Disease Patients before and after Hematopoietic Stem Cell Transplantation

We have reported a case of acquired hemophilia A in our series of sickle cell disease patients treated with hematopoietic stem cell transplantation (HSCT) using HLA matched sibling donors, non-myeloablative conditioning, and sirolimus immunosuppression for GVHD prevention (Lozier et al, Haemophilia 2013). Subsequently we encountered another case in the 48 patients now transplanted. The objective of our transplant regimen is to establish a stable engraftment that will correct the sickle cell disease, with minimal morbidity. As a result of the non-myeloablative conditioning, our HSCT recipients typically display mixed chimerism between donor and recipient after HSCT with a post-HSCT immune system derived from both donor and recipient. We investigated the prevalence of FVIII-binding antibodies by ELISA in HSCT recipients prior to and after HSCT, and sought to determine if prior exposure to blood products, differences in FVIII haplotypes, and/or ethnicity may predispose to FVIII inhibitors. We measured IgG antibodies that bind FVIII by ELISA in 30 of our 48 transplanted subjects and donors for whom plasma, serum, and DNA were available, before and after HSCT. Published evidence suggests 3-19% of normal blood donors have detectable FVIII binding antibodies when tested in an ELISA format (Krudysz-Amblo et al, Blood 2009; Whelan et al, Blood 2012). We were surprised to find that 22 of 30 recipients (73%) had measurable titers of ≥ 1:50 prior to HSCT. We also assessed their sibling donors as a control group and found 13 of 24 evaluable donors had titers ≥ 1:50 (54%). Measurement of antibodies to FVIII in African-American blood donors at NIH showed 8 of 13 (62%) to have ≥ 1:50 titers, in contrast to only 1 of 20 Caucasian (5%) NIH blood donors tested (P < 0.05). This suggests African-Americans are more likely to form antibodies to FVIII than Caucasians, and might explain the two cases of acquired hemophilia A in our HSCT study. This has not been reported in allogeneic HSCT studies, though 2 out of 155 autologous HSCT patients with autoimmune diseases were reported to have acquired hemophilia after HSCT (Loh et al, Blood 2007). Of 8 patients with titers <1:50, 6 had no increase in titers measured out to two years after HSCT; 2 had peak post-transplant titers of 1:50 or 1:100 at most, and none developed FVIII inhibitors. Of 22 patients with titers of ≥ 1:50 pre-transplant two developed FVIII inhibitors and 20 had unchanged or diminishing titers following HSCT (P = 0.38, for inhibitor development, NS). We investigated the possibility that transfusions with red blood cells for sickle cell disease might have sensitized patients, either by exposure to "foreign" FVIII protein or by modulation of the immune system, prior to HSCT. Of the 8 patients with titers < 1:50 pre-transplant, one recalled transfusion with > 50 units of RBCs prior to transplant (12.5%), while 9 of 22 with titers ≥ 1:50 prior to transplant (41%) recalled transfusion with > 50 units of RBCs prior to transplant (P = 0.14, NS). Six haplotypes of the FVIII protein are known, and empiric data suggests uncommon haplotypes more prevalent in African-Americans may be associated with higher rates of FVIII inhibitors in African-Americans with hemophilia A after FVIII treatment (Viel et al, NEJM 2009), and rare FVIII polymorphisms are associated with acquired hemophilia A in the Caucasian population as well (Tiede et al, Ann Hematol 2010). We ascertained FVIII haplotypes in HSCT recipients and their donors to see if mismatches between donor and recipient may be associated with FVIII antibodies or inhibitors. Of 30 patients studied, there were 27 for whom DNA data permitted us to infer haplotypes. Fourteen recipients were identical matches to their donors, of whom one (7%) had an increase in titer from baseline after transplant; none developed an inhibitor. Of 15 recipients with mismatches in FVIII haplotypes with their donors, 3 had increases from baseline titers after transplant (20%), including the two FVIII inhibitor patients (P = 0.45, NS). Our study shows that African-Americans have a higher prevalence of low-titer/subclinical anti-human FVIII antibodies compared to Caucasians which may explain subsequent FVIII inhibitors in the setting of hematopoietic stem cell HSCT. Neither greater recalled exposure to RBC transfusions pre-HSCT nor mismatch of FVIII haplotypes with donor explain the risk of inhibitor development.