Abstract
Allogeneic transplant remains the best postremission therapy for patients with nonfavorable risk acute myeloid leukemia (AML). However, some patients are ineligible because of psychosocial barriers, such as lack of appropriate caregiver support. We hypothesized that immune checkpoint inhibition after autologous transplant might represent effective postremission therapy in such patients. We conducted a phase 2 study of autologous transplantation followed by administration of pembrolizumab (8 cycles starting day +1). Twenty patients with nonfavorable AML in complete remission were treated (median age, 64 years; CR1, 80%); 55% were non-White and adverse-risk AML was present in 40%. Treatment was well tolerated, with only 1 nonrelapse death. Immune-related adverse events occurred in 9 patients. After a median follow-up of 80 months, 14 patients remain alive, with 10 patients in continuous remission. The estimated 2-year LFS was 48.4%, which met the primary end point of 2-year LFS >25%; the 2-year overall survival (OS), nonrelapse mortality, and cumulative incidences of relapse were 68%, 5%, and 46%, respectively. In comparison with a propensity score-matched cohort group of patients with AML receiving allogeneic transplant, the 3-year OS was similar (73% vs 76%). Patients in the study had inferior LFS (51% vs 75%) but superior postrelapse survival (45% vs 14%). In conclusion, programmed cell death protein-1 blockade after autologous transplant is a safe and effective alternative postremission strategy in patients with nonfavorable risk AML who are ineligible for allogeneic transplant, a context in which there is significant unmet need. This trial was registered at www.clinicaltrials.gov as #NCT02771197.