Pharmacogenomic Variation May Impact Cyclophosphamide Metabolism and Graft Status after HCT for Sickle Cell Disease

Introduction Post-transplant cyclophosphamide (PT CP) is given to deplete recipient alloreactive lymphocytes, particularly after haploidentical hematopoietic cell transplant (HCT). Several cytochrome P450s (CYPs), alcohol and aldehyde dehydrogenases (ADHs and ALDHs), and ABCB1 mediate the detoxification and clearance of cyclophosphamide (CP) and its metabolites. Furthermore, renal insufficiency, a common comorbidity in sickle cell disease, may affect CP pharmacokinetics (PK) and graft status. We hypothesize that interindividual variability affects CP PK and is associated with graft rejection in HCT patients. Methods Adult patients with SCD transplanted on one of two nonmyeloablative haploidentical protocols (NCT00977691 and NCT03077542) were analyzed. Patients received one or two 50 mg/kg PT CP infusions; samples were drawn during the first dose and analyzed utilizing uHPLC-MS/MS to simultaneously detect CP and an active metabolite, 4OH-CP. Transplant outcomes and PK were compared with eGFR. The Pharmacoscan platform was used to genotype ABCB1, ADH1A, ADH1C, ADH4, ADH5, ADH7, ALDH1A1, ALDH3A1, CYP2B6, CYP2C9, CYP3A4, and CYP3A5. Genotypes were compared to PK parameters, CP toxicity, and transplant outcomes in both univariate and multivariate analyses. Results PK parameters were not associated with graft status (n=30; Table 1), and graft status did not depend on renal function. Pharmacogenomic (PG) analysis was conducted in a subset of patients (n=9). An intronic allele in ADH5 (rs2602836) was associated with prolonged graft survival (P<0.001; Figure 1). This association withstood multivariate analysis accounting for covariates (weight, age, race, sex, albumin, eGFR). CYP3A5*3 (rs776746) was associated with decreased CP clearance and higher AUC (P=0.024), and an ADH1A variant (rs1826909) was associated with shorter half-life and lower volume of distribution (P<0.048). Conclusions ADH5 detoxifies aldophosphamide by converting it into carboxyphosphamide, and we present the first evidence that a polymorphism in this enzyme is associated with graft status. These data need to be confirmed in a larger patient cohort. Although PG variants did impact PK, there was no association between PK parameters and graft status. Additionally, renal insufficiency did not impact PK or graft status.