Optimized Cyclophosphamide (CY) Dosing in Combination with Fludarabine, TBI and ATG As Conditioning for Unrelated Donor Bone Marrow Transplantation (BMT) in Severe Aplastic Anemia (SAA): A Phase I/II Study from the Blood and Marrow Transplant Clinical Trials Network (BMT CTN)

Objective: To determine toxicity and efficacy of adding fludarabine (FLU) to a standard preparative regimen of low-dose total body irradiation (TBI), anti-thymocyte globulin (ATG) and CY (Deeg J et al, Blood 108:1485, 2006), with de-escalation of the CY dose.

Patients and Methods: Between May 2006 and December 2013, the BMT CTN (sponsored by the NHLBI and NCI) conducted a Phase I/II trial of unrelated donor BMT in SAA (BMT CTN 0301; NCT 00326417). Patients were eligible if they were aged < 65 years, with adequate organ function, and an available unrelated marrow donor matched at 7 or 8 of 8 HLA-A, B, C, and DRB1 loci. The trial accrued 97 patients, although analyses are limited to 96; 1 patient withdrew consent prior to BMT. All patients received TBI 200 cGy (day -1), ATG (either thymoglobulin 3 mg/kg IV or ATGAM 30 mg/kg IV daily x 3, days –4 to –2), FLU (30 mg/m² IV daily x 4, days –5 to –2). The Phase I portion of the trial tested four CY dose levels: 150 mg/kg (days –4 to –2); 100 mg/kg (days –3 to –2); 50 mg/kg (day –2); and 0 mg/kg. The Phase I design allowed enrollment of up to six patients at each CY dose level unless toxicity or graft failure boundaries were crossed. In the Phase II portion, patients were enrolled onto the optimal CY dose level, using adaptive Bayesian criteria to rank desirability of CY doses based on Day 100 outcomes of engraftment and early death. The primary endpoint of the study was determination of the optimal CY dose based on Day 100 assessments of graft failure (primary: absolute neutrophil count <0.5 x 10⁹/L and secondary: sustained decline in absolute neutrophil count to <0.5 x 10⁹/L), major (grade 3 or higher) regimen-related toxicity (RRT) and early death. Early stopping guidelines were based on a composite endpoint of graft failure and treatment-related mortality through day 100.

Results: Twenty-one patients accrued to the Phase I portion of the trial and all CY dose levels were tested. CY dose level 0 mg/kg was closed after 3 of the first 3 patients developed secondary graft failure. The Phase II portion of the trial opened with CY dose level 150 mg/kg. However, this dose level was closed for excess toxicity (7 of 14 patients died; organ failure n=4, ARDS n=2 and infection n=1; Tolar et al, Biol Blood Marrow Transplant, 2012). Patients were subsequently assigned to CY dose level 100 mg/kg (CY DL 100; n=41) or CY dose level 50 mg/kg (CY DL 50; n=38), depending on Bayesian assessment of criteria noted above except that the last 20 patients were preferentially assigned to CY DL 50 to balance accrual to the two dose levels. Approximately half of patients were male and 79% Caucasian. The median age was 20.6 years. The median age for CY DL 50 and DL 100 was 24.5 (0.5-65) and 17.6 (1.9-63) years, respectively. The number of transplants with a mismatched (i.e.7/8) donor was 7 (18%) in the CY DL 50 group and 14 (34%) in the CY DL 100 group. All patients reached their Day 100 endpoint. The cumulative incidence of grade 2-4 acute GVHD was comparable between the two CY dose levels: 24% (DL 50) vs 27% (DL 100). All deaths before Day 100 were due to primary or secondary graft failure (CY DL 50; n=1 and CY DL 100; n=2),. Table 1 shows individual Day 100 outcomes for CY DL 50 and CY DL 100. Median follow up is 15 (3-27) and 24 (12-50) months for CY DL 50 and 100, respectively. All surviving patients on CY dose level 100 mg/kg have achieved a minimum follow up of 12 months, as opposed to only 75% of surviving patients on CY dose level 50 mg/kg.

Conclusion: Results of the adaptive Bayesian dose selection suggest that the most desirable CY dose is 50 mg/kg followed by CY dose, 100 mg/kg. However, interpretation of data needs to take into account the potential imbalance of donor-recipient HLA disparity in the CY 50 mg/kg dose level. CY dose 150 mg/kg and 0 mg/kg should be avoided.