Association of Cyclophosphamide (CY) and Fludarabine (FLU) Pharmacokinetics with Treatment Related Mortality (TRM) after Reduced Intensity Conditioning (RIC) Hematopoietic Cell Transplantation (HCT)

CY and FLU are chemotherapeutic agents often used in RIC regimens. RIC regimens may yield lower transplant related complications and toxicities as compared to myeloablative conditioning. However TRM at 12 months still approximates 30%. We previously reported that high F-ara-A (active metabolite of FLU) exposure was associated with more TRM at a FLU dose of 40 mg/m2, but the relationship is unknown at lower doses. Similarly, phosphoramide mustard (PM, the active metabolite of CY) may influence clinical outcomes. Thus we studied the relationship between clinical outcomes and systemic exposure of F-ara-A at lower doses and PM.

Forty adults undergoing allogeneic RIC HCT were prospectively studied from March 2013 to May 2014. All patients received FLU, CY and TBI as conditioning and cyclosporine or sirolimus, plus mycophenolate for posttransplant immunosuppression. Median age was 62 years (21-72). 22 were male and 18 female. The stem cell sources were sibling or unrelated donor PBSC (n=22, 55%), cord blood (n=13, 32.5%) or bone marrow (n=5, 12.5%). CY 50mg/kg was administered IV over 2 hrs on day -6. Pharmacokinetic samples were obtained at predose and 2, 4, 6, 21, 24 and 45 hrs after the end of infusion. PM was derivatized with diethyldithiocarbamate and measured by ultraviolet detection with HPLC. PM area under the curve (AUC_0-last), AUC_0-6 and AUC_0-24 were calculated using non-compartmental methods (Phoenix WinNonlin Professional 6.3). Fludarabine dosing was 30 mg/m2/d (n=35), 32 mg/m2/d (n=1), 35 mg/m2/d (n=2) and 40 mg/m2/day (n=2) x 5 doses depending on protocol and was administered IV over 1 hr on days -6 to -2. In addition, F-ara-A troughs were obtained at 24 hours after the start of the first infusion on day -6 and 24 hours after start of 2^nd infusion. F-ara-A quantification was performed using HPLC-UV. TRM was defined as death due to any cause other than relapse or disease progression. GVHD was staged and graded according to the standard GVHD criteria based on clinical and pathological criteria Recursive partitioning regression analysis was used to determine optimal cut points for PM and F-ara-A pharmacokinetic measures towards TRM at day 100, 6 months and 12 months and the incidence of acute graft vs host disease (GVHD). The cumulative incidence of engraftment, TRM and acute GVHD (II-IV and III-IV) was calculated using death prior to event as a competing risk. The proportional hazards model of Fine and Gray was used to assess the association of F-ara-A exposures towards TRM, acute GVHD and engraftment.

The incidence of TRM at day 100 was 13%, 20% at 6 months and 47% at 12 months. The incidence of grades 2-4 and 3-4 acute GVHD was 39% and 25%. In univariate analysis, day 100 TRM risk was significantly higher in patients with a PM AUC_0-24 ≥85 ug-hr/ml (relative risk (RR) 36%, [95% CI 9-64%]) vs. AUC_0-24 <85 ug-hr/ml (4% [0-11%] (p<0.01). Risk of TRM at 6 months was also associated with a PM AUC_0-24 ≥85 ug-hr/ml (47% [17-77%]) vs. <85 ug-hr/ml (14% [0-28%]) (p=0.02) while there was no effect of AUC_0-24 on 12 month TRM.

Patients with F-ara-A trough concentration after the second dose ≥75 ng/ml had significantly higher risk of day 100 TRM as compared to patients in whom F-ara-A was <75 ng/mL (33% [4-63%] vs 7% [0-15%])(p=0.01). F-ara-A troughs were not associated with TRM at 6 and 12 months. PM and F-ara-A exposures did not influence risks of acute GVHD. Engraftment rates were high (93%, CI 82-99%) so an insufficient number of failure events were available for evaluation.

Higher PM exposures were associated with increased risk of TRM at day 100 and 6 months. Higher F-ara-A exposures were associated with greater risk of TRM at day 100. No associations were observed towards acute GVHD. These data for the first time show that TRM is associated with high PM exposure. These data also support our earlier findings that overexposure to F-ara-A may be harmful in RIC regimens. Multivariate models adjusting for other important clinical factors will be evaluated as enrollment continues to a larger cohort. In the future predictive pharmacokinetic models may be beneficial in personalizing fludarabine and cyclophosphamide doses.