Outcomes of Relapse of AML Post Allogeneic Transplantation in the Era of Hypomethylating Agents: An Analysis of 162 Allogeneic Transplants from a Single Center

Allogeneic hematopoietic cell transplantation (allo-HCT) can be curative in many patients with intermediate and high risk AML. However, patients with AML who relapse post allo-HCT typically have a dismal prognosis with limited therapeutic options. The activity and minimal toxicity associated with hypomethylating agents (hypomeths) makes them potentially useful in the management of post allo-HCT relapsed AML patients. However, their use in this setting has not been well studied. We assessed 162 consecutive patients who underwent a first allo-HCT for AML at our center during Feb. 2005 - May 2013. Patient characteristics were: median age, 53 (18-74); M=75, F=87; MRD=59, MUD=67, haplo=36; ablative conditioning=98, RICT/NST=64; PBSC=142, BM=20; CIBMTR risk category- high=53, intermediate=28, low=74, unknown=7. Post-relapse therapy (PRT) was determined by the attending physician and patient preference. Patient characteristics, PRT, GVHD and survival were prospectively collected. With a median follow-up for surviving patients of 22.4 months, relapse of AML post-allo-HCT occurred in 55 patients (22 MRD, 24 MUD, 9 haplo; cytogenetic risk 23 poor, 31 intermediate, 1 unknown) at a median of 113 days (25-1106) post allo-HCT. Thirty-four patients (62%) were treated with a hypometh post-relapse (17 azacitidine, 10 decitabine, 7 both) (12 hypometh alone, 11 combined with chemotherapy, 11 sequentially with chemotherapy). Median number of hypometh cycles was 2 (1-9). Of the 23 patients that received at least 2 hypometh cycles, 14 achieved CR or CRi. DLI was administered in 15 of the 55 relapsed patients and 16 patients received a second allo-HCT. Estimated Kaplan-Meier probability of post-relapse survival (PRS) at 6, 12 and 24 months for all patients was 53%, 36% and 19% and was not significantly different for patients who developed any versus no PRT GVHD. However, PRS at 6 and 12 months was 62% (95% CI 43-76%) and 38% (95% CI 22-54%) in patients who received post-relapse hypomeths versus 38% (95% CI 18-58%) and 33% (95% CI 15-53%) in patients who did not (p=0.063 Gehan’s test). There was no difference in PRS at 24 months. PRS > 1 year was achieved in 19 of 55 patients (9 received a second allo-HCT, 9 received DLI, 12 received a hypometh). At the time of writing, 6 patients are alive and in CR at a median of 49 months (10-72) post-relapse. Of the 6 patients (5 MRD, 1 MUD; cytogenetic risk 2 poor, 4 intermediate), 4 received hypomeths, 3 received a second transplant, 1 received DLI, and 4 have active GVHD. 

Relapse post allo-HCT remains a major obstacle to long-term survival in patients with AML. The use of hypomeths may be associated with a prolongation of early PRS although it does not appear to increase PRS beyond one year. PRS beyond one year can be achieved without a second allogeneic transplant in a minority of relapsing patients. Combination of therapy with hypomeths and novel agents may be necessary to impact long-term outcome.