Fludarabine in Combination with Busulfan As Pretransplant Conditioning for Platelet Gene Therapy in Murine Hemophilia a with Inhibitors

Treatment of hemophilia A (HA) with inhibitors is costly, inconvenient, and incompletely effective. We have demonstrated that syngeneic transplantation of hematopoietic stem cells (HSCs) that are genetically modified to express FVIII in platelets restores hemostasis in HA mice with pre-existing anti-FVIII immunity under total body irradiation (TBI) preconditioning, a proof-of principle that platelet gene therapy could be a promising strategy for HA patients. Currently, Fludarabine along with busulfan (FB) has been shown to be highly effective as a preconditioning regimen for allogeneic HSC transplantation (HSCT) in the clinic. Thus, in this study, we aimed to evaluate if FB works as a clinically translatable preconditioning approach for platelet gene therapy of HA with pre-existing inhibitors. Fludarabine was intraperitoneally administered at 100mg/kg/day on Days -7 through -4, and busulfan was given at 25mg/kg/day on Days -2 and -1 into rhFVIII-primed FVIII^null mice. A TBI regimen of 6.6Gy was used as a parallel control. Platelet FVIII expression was introduced by transplantation of 2bF8 lentivirus (2bF8LV)-transduced HSCs. The peripheral blood recovery was monitored by blood count, and cell subset reconstitution was analyzed by flow cytometry. Animals were further analyzed by quantitative PCR (qPCR), FVIII assays, inhibitor assays, and tail bleeding test.

Blood counts demonstrated that platelet numbers and hemoglobin levels in the FB group were similar to those in the TBI group after HSCT. The recovery of leukocytes in the recipients with FB conditioning was greater than that in the animals conditioned with TBI at the early stage (4-12 weeks after HSCT). The engraftments in the two groups were comparable in terms of donor- derived leukocyte chimerism during the study course. The percentage of leukocyte chimerism at each time point in the FB group (ranging from 68.6±12.3% to 83.5±8.5%, n=7) was not statistically significantly different than that in the TBI group (ranging from 66.5±15.8% to 74.5±14.8%, n=5-6). All recipients achieved more than 55% donor-derived leukocytes at 20 weeks after HSCT. When we looked into the cell subsets in the recipients, we found that FB preconditioning is favorable for donor-derived myeloid cell reconstitution (the average engraftment ranging from 81.3-85.4% in the FB group, 52.0-62.3% in the TBI group).

qPCR results confirmed that HSCs were genetically modified by 2bF8LV in the recipients with either FB or TBI conditioning. The average copy number of the 2bF8 cassette per cell was 0.93±0.17 (n=7) in the FB group, which was not significantly different compared to the TBI group (0.79±0.11, n=6). The average platelet-FVIII levels at each time point ranged from 3.87-5.98 mU/10⁸ platelets in the FB group and 2.21-10.47 mU/10⁸ platelets in the TBI group, respectively. There was no significant difference between the two groups. To assess whether the bleeding phenotype was rescued in HA mice after receiving 2bF8 LV-transduced HSCs, we used a 6-hour (hr) tail bleeding test. All 2bF8 LV-transduced recipients stopped bleeding within 6 hrs with a remaining hemoglobin level of 62.7±20.6%, which was significantly higher than the FVIII^null control group (37.6±4.5%). Of note, none of the HA control mice stopped bleeding within 6 hrs.

Notably, we found that both anti-FVIII inhibitor and total IgG titers declined with time in recipients under either FB or TBI preconditioning. When reduction of the inhibitor titer over time was calculated as a half-life (t_1/2), there was no significant difference between the FB group and the TBI group. When the inhibitor titer dropped to undetectable, we further challenged recipients with 50 U/kg rhF8 weekly for 4 weeks. There were no detectable anti-FVIII inhibitors in the 2bF8LV-transduced recipients from either the FB or TBI group. In contrast, all the untransduced transplanted HA control mice produced various levels of inhibitors under the same challenge. These data indicate that 2bF8LV-transduced primed-HA mice under either FB or TBI conditioning may induce immune tolerance.

In conclusion, fludarabine along with busulfan preconditioning successfully introduced therapeutic levels of platelet FVIII expression without evoking an anti-FVIII memory response in the transduced recipients with pre-existing immunity. Our data suggest that this approach may be a promising and clinically translatable strategy for gene therapy of hemophilia A.