Abstract
Reduced-intensity conditioning (RIC) regimens have expanded allogeneic hematopoietic stem cell transplantation (allo-HSCT) to patients (pts) with acute myeloid leukemia (AML) and myelodysplastic syndrome (MDS) who are unfit for myeloablative regimens. Several RIC regimens are typically used. We and others have previously shown that fludarabine-melphalan (FM) has a lower risk of relapse compared to fludarabine-busulfan (FB).
Herein, we re-examine these findings with a larger cohort and longer follow-up. We also compare the composite endpoint graft-versus-host disease (GVHD)-free, relapse-free survival (GRFS), and elucidate factors that predict for relapse and NRM.
After IRB approval, we conducted a study of pts (age ≥18) diagnosed with AML or MDS who underwent allo-HSCT from January 1st, 2008 to March 1st 2017 at Mayo Clinic Rochester. Only pts who received FM (melphalan 140 mg/m2) or FB (busulfan 0.8 mg/kg IV for 10 doses with therapeutic AUC target of 900-1500 mcmol/L[min]) were included. Grade III-IV acute GVHD, systemic therapy-requiring chronic GVHD, relapse, or death counted as events for GRFS. Statistical analyses were performed using JMP 13 and EZR 1.36.
We identified 184 pts who underwent RIC allo-HSCT: 135 with AML and 49 with MDS. The median age for the entire cohort was 61 (range, 18-74). FM was used in 123 pts, while 61 received FB. Baseline characteristics were comparable (Table 1a). The median follow-up was 5 years (95% CI, 4-5.6).
In line with previous findings, the cumulative incidence of relapse (CI-R) was significantly higher with FB compared to FM: 3-year CI-R was 31.7% for FB vs. 17.3% for FM (p= .04)(Figure 1a). On the other hand, there was a nonsignificant trend towards an increased CI-NRM in the FM group; at Day +100, CI-NRM was 6.5% for FM vs. 1.6% for FB (p= .24). Acute and chronic GHVD were comparable (Table 1b).
The median GRFS for the entire cohort was 7.3 months (95% CI, 6.1-8.4); there was no difference between the two regimens (p=.94). Overall survival (OS) was comparable (p= .61), as were causes of death (Figure 1b). In a multivariate Fine-Gray proportional hazard model for relapse vs. NRM, which included factors in Table 2; disease status (CR1 vs. CR2 or active disease) predicted relapse (HR 0.42, 95% CI 0.21-0.84, p= .01). There was also a trend towards a decreased risk of relapse with FM (HR 0.54, 95% CI 0.28-1.03, p= .06). On the other hand, age was predictive of NRM (HR 1.05 per year, 95% CI 1.01-1.09, p= .01).
In the present follow-up study, there were no significant differences in GRFS or OS between FM and FB regimens for RIC allo-HSCT in AML/MDS, despite FB having a significantly higher risk of relapse. Likely, the OS was balanced by continuing NRM after FM conditioning, particularly in older pts. Further studies, ideally prospective, are required to elucidate the optimal RIC regimen matched to disease biology and pt fitness.