Outcomes of patients with relapsed/refractory FLT3mut+ Acute Myeloid Leukemia who resumed gilteritinib therapy after HSCT: Post hoc analysis from the ADMIRAL and COMMODORE trials

Abstract Allogeneic hematopoietic stem cell transplantation (HSCT) remains the only curative option for patients with relapsed/refractory (R/R) acute myeloid leukemia (AML). Compared with patients in first complete remission (CR1), patients with R/R AML are often more heavily pretreated and may harbor persistent measurable residual disease (MRD) or multiclonal mutations, increasing the risk of post-HSCT relapse. Data suggest patients with FMS-like tyrosine kinase 3-mutated (FLT3mut+)AML in CR1 benefit from post-HSCT maintenance therapy with FLT3 inhibitors, including gilteritinib (GILT), particularly those with persistent MRD in the peri-HSCT period. However, limited data exist on the efficacy of post-HSCT FLT3 inhibition in patients with R/R FLT3mut+ AML. This pooled analysis of the ADMIRAL (NCT02421939) and COMMODORE (NCT03182244) randomized phase 3 trials comparing GILT vs salvage chemotherapy (SC) evaluated overall survival (OS) in 2 subgroups of patients with R/R FLT3mut+AML: (1) patients who received HSCT vs those who did not receive HSCT during the study period independent of treatment received before HSCT; and (2) who received HSCT during the study period, were relapse-free at day 60 and resumed GILT vs those who did not resume GILT post-HSCT. All randomized patients who received GILT or SC in both trials were pooled, and summarized by HSCT (on-study or subsequent) status and post-HSCT GILT resumption. Only patients treated with GILT were allowed to resume GILT post-HSCT if they met criteria, including being alive 30–90 days post-HSCT, having achieved successful engraftment without transfusions (absolute neutrophil count ≥500/mm3; platelets ≥20,000/mm3), not having grade ≥2 acute graft-versus-host disease (GvHD), and being in composite CR (CRc). A landmark analysis was conducted on day 60 post-HSCT to mitigate survivorship bias by including only patients who remained relapse-free up to day 60. A total of 647 patients were included; 125 (19.3%) patients who received HSCT and 522 (80.7%) patients who did not receive HSCT during the study period. Of 125 HSCT recipients, 95 (76.0%) and 30 (24.0%) were from GILT and SC arms, respectively. Among GILT-treated HSCT recipients, 58 (61.1%) resumed GILT and 37 (38.9%) did not resume GILT after HSCT. At baseline, patients who resumed GILT post-HSCT were younger (median [range] age 43.5 [18–71] vs. 54.0 [18–70] years), had a lower FLT3-internal tandem duplication and tyrosine kinase domain co-mutation rate (1.7% vs. 8.1%) and a nucleophosmin-1 co-mutation (57.7% vs. 81.3%) rate than those who did not resume GILT, although there were no significant differences between groups. Peri-HSCT MRD was not measured. Median (range) post-HSCT GILT treatment duration was 11.1 (0.03–66.8) months with a median (range) daily dose of 119.4 (40–200) mg. Median (95% CI) follow-up for OS was 42.0 (37.1, 46.1) months in patients who received HSCT and 37.1 (35.4, 38.4) months in patients who did not receive HSCT during the study period. Median (95% CI) OS was 24.1 (18.8, not estimable [NE]) months in patients who received HSCT and 7.4 (6.6, 8.3) months in patients who did not receive HSCT (Mantel-Byar p<0.0001). In patients who received HSCT, OS rates at 6, 12, 24 and 36 months were 92.5%, 67.7%, 50.3%, and 44.9%. In patients who did not receive HSCT, OS rates were 58.8%, 26.0%, 12.7%, and 8.1%, respectively. Among GILT-treated HSCT recipients who were relapse-free at day 60 post-HSCT, 54 resumed GILT and 28 did not resume GILT after HSCT. Possible reasons for not resuming GILT included not achieving any type of CR (39.3%), not being alive 30–90 days post-HSCT (14.3%), grade ≥2 GvHD (3.6%) and other (50.0%). Median OS (95% CI) was NE (16.8, NE) in patients who resumed GILT vs 10.1 (3.8, 17.4) months in patients who did not (HR [95% CI] 0.370 [0.198, 0.692]). In patients who resumed GILT, OS rates at 6, 12, 24 and 36 months were 83.3%, 68.5%, 62.8% and 60.8%. In patients who did not resume GILT, OS rates were 59.6%, 47.7%, 26.0%, and 26.0%, respectively. In patients with R/R FLT3mut+ AML, long-term survival was strongly associated with receipt of HSCT, with the most favorable outcomes among the groups tested seen in patients randomized to GILT who underwent HSCT and resumed GILT maintenance post-transplant. Given the non-randomized nature of HSCT status, pre-HSCT factors, and FLT3mut+ characteristics subgroup comparisons should be interpreted with caution.