Induction-refractory acute myeloid leukemia (AML) occurs in approximately 30% of AML patients and is mostly associated with inferior outcome. A sequential approach offers shorter time to allogeneic hematopoietic cell transplantation (HCT), reduces salvage-associated toxicities while providing significant anti-leukemic activity. We hypothesized that early intervention with a sequential approach based on the original FLAMSA regimen salvage these patients more effectively as compared with a standard salvage therapy. Methods:
We retrospectively compared results between 2 tertiary centers in Israel that apply different therapeutic modalities for first-induction failure: The first (TASMC -"cases"
group), employs a transplant-based sequential approach for patients with first-induction-refractory AML, while the second center (RMC – "controls"
group), applies mostly conventional salvage regimens. All consecutive patients ≥18 years who were treated in both centers between 2013 and 2021 and given either high-intensity induction chemotherapy or low-intensity venetoclax (VEN)-based regimens were included. Refractory to first induction was defined as either blasts >10% on day 30 marrow after 7+3 regimen, or after 2-3 cycles of a VEN-based regimen. We compared baseline characteristics between the 2 groups and used time-to-event analyses to calculate non-relapse mortality (NRM) and overall survival (OS) with day of allocation (either sequential therapy or non-sequential approach) defined as day 0. Controls who underwent HCT were not censored at transplant. Cox regression analysis was used to identify predictors for survival. This trial was approved by both Institutional Ethic Committees. Results:
Between 01/2013 and 12/2021 we identified 91 patients that fulfilled the eligibility criteria (cases, n=55; controls, n=36). Characteristics of cases and controls were similar, except for higher % of patients with metabolic syndrome in the control group, Table
. Mean days from diagnosis to HCT in the cases group was 69±25 with longer mean time in patients who were given VEN-based compared to high-intensity induction chemotherapy (94±23 vs. 62±21 days, respectively, p<.001). All patients advanced to transplantation. Median days to neutrophils and platelets engraftment were 10 (range, 7-16) and 16 (range, 9-35), respectively. Incidences of grade 2-4 and 3-4 acute GVHD were 55% (95%CI 47%-63%) and 15% (95%CI 47%-63%), respectively, and incidences of overall and moderate-severe chronic GVHD at 3 years were 64% (95%CI 44%-84%) and 37% (95%CI 25%-49%), respectively. In the controls group, patients were given high-dose cytarabine-based salvage regimens (n=21, 58%), novel targeted therapies (n=5, 14%), HCT (n=4, 11%), VEN-based regimen (n=3, 8%), and palliation (n=2, 6%). Of those patients who were given chemotherapy/other targeted therapies (n=30), 11 (37%) responded and of them, 9 (82%) proceeded to HCT. Five additional patients underwent HCT in active disease. Thus, overall, 18 patients (50%) in the controls group underwent HCT. Percentage of complete remission was higher in the cases group (measured at 1 month after HCT) vs. controls group (measured at any time after allocation) (87% vs. 39%, p<.001). 1-year NRM was higher in the cases group compared to controls group, however this did not reach statistical significance (37% (95%CI 32%-42%) vs. 20% (95%CI 14%-26%), p=0.126). Median OS was similar in patients given intensive induction chemotherapy between the cases and controls groups (8.3 (95%CI 2.74-14.4) months vs. 7.1 (95%CI 2.15-11.4) months, p=0.66), although in patients with adverse ELN risk, OS within the intensive group was better (8.2 (95%5.1-11/7) months vs. 2.4 (95%CI 0-4.8) months, p=0.027). Importantly, OS was better in cases refractory to VEN-based induction regimens compared to controls (22 (95%CI n/a) months vs. 7.8 (95%CI 0-15.9) months, p<.001). Cox regression analysis identified adverse ELN risk group and lower performance status as predictors for mortality (HR=2.5, p=0.02 and p=0.008, respectively). Conclusions:
sequential therapy in refractory to induction therapy AML may benefit patients with poor salvage options such as those with adverse risk AML and those with refractory disease to VEN-based induction regimens. Better salvage regimens and improved transplantation techniques may improve both strategies to enhance patients' outcomes.