Abstract
Background: Acute myeloid leukemia with MECOM rearrangement (MECOMr AML) represents 1-2% of AML and is associated with poor long-term survival outcomes. Though chemotherapeutic resistance in this population has been previously reported, clinical outcomes in the era of hypomethylating agent+venetoclax (HMA+ven) are not well-defined. To explore this, we performed a retrospective study analyzing outcomes in patients (pts) with MECOMr AML treated with HMA+ven.
Methods: We included clinicopathological and molecular data from 56 pts with MECOMr AML treated with HMA+ven in the newly diagnosed (1L) and relapsed/refractory (RR) setting from 4 academic centers across the United States. MECOMr was detected by karyotype analysis and/or FISH. Composite complete remission (cCR) was defined as CR + CR with incomplete hematologic recovery (CRi) + CR with partial hematologic recovery (CRh). The Kaplan-Meier method was used to estimate overall survival (OS). Log-rank test was used to compare OS between groups.
Results: 56 pts with MECOMr AML treated with HMA+ven were identified, including 27 (48%) 1L and 29 (52%) RR. Median age was 65 years (range 20-84), with 1L being older than RR [69 (34-79) vs. 61 (20-84); p=0.02]. 7% had prior myeloid malignancy and 36% had therapy-related AML. In the RR cohort, prior therapy included intensive chemotherapy (IC) in 15 (52%), HMA in 9 (31%), and IC and HMA in 5 (17%); 5 (17%) had undergone prior allogeneic stem cell transplant (alloSCT). Overall, 21 (38%) pts had inv(3), 10 (18%) had t(3;3), 20 (36%) had t(3;other), and 5 (9%) had MECOM/EVI1-rearrangment detected by FISH. The translocation partner in the t(3;other) group included: chromosome 21 (6 pts); chromosomes 2, 7, and 11 (2 pts each); and chromosomes 5, 8, 10, 12, 14, 19, 21 and 22 (1 pt each). 15 pts (27%) had isolated MECOMr. Complex karyotype was present in 23 (41%), monosomy 7 in 19 (34%), del(5/5q) in 13 (23%), del(7q) in 6 (11%), and del(17p) in 4 (7%). The most common mutations included SF3B1 in 13 (23%) and TP53 in 12 (21%), followed by PTPN11 in 11 (20%), NRAS in 7 (12%), and DNMT3A and NF1 each in 6 (11%) pts. Cytomolecular characteristics were comparable between 1L and RR except for TP53 mutation, which was enriched in the 1L group (75% vs 25%, p=0.02).
In terms of treatment characteristics, HMA+ven contained azacitidine in 38/56 (68%) and decitabine in 18/56 (32%) with a median of 2 cycles received (range 1-12). Overall, cCR rate was 15/56 (27%), comparable in the 1L and RR groups [9/27 (33%) vs 6/29 (20%), p=0.40]. cCR+MLFS rate was 22/56 (39%) in the overall cohort, comparable between 1L and RR [13/27 (48%) vs 9/29 (31%), p=0.27], although numerically higher in 1L. Amongst responders with minimal residual disease (MRD) assessments performed by local flow cytometry, 6/22 (27%) were MRD-negative across groups, including 4 1L and 2 RR. 13/56 (23%) proceeded to alloSCT, with a numerically higher rate in 1L than RR [8/27 (30%) vs. 5/29 (17%), p=0.34]. 29/56 (52%) pts had refractory disease to HMA+ven, including 11/27 (40%) in 1L vs 18/29 (62%) in RR (p=0.27).
Median OS (mOS) for 1L and RR were 6.9 (95% CI 5.3-14.0) vs 6.9 months (mos) (95% CI 5.1-11.5), p=0.72, with a median follow up of 46 and 52 mos, respectively. When assessing mOS from 3-month landmark for responders (cCR+MLFS) vs non-responders, mOS was numerically but not statistically longer in the 1L cohort [9.1 (95% CI 4.2-not reached (NR)) vs 3.9 mos (95% CI 2.3-NR), p=0.82] and statistically longer in the RR cohort [55.0 (95% CI 6.86-NR) vs 3.9 mos (95% CI 2.1-15), p=0.008]. Next, landmark analysis for OS following alloSCT vs no alloSCT from the time of first cCR+MLFS was performed. In the 1L group, mOS was significantly longer amongst those proceeding to alloSCT compared to those who did not [17.4 (95% CI 2.7-NR) vs 3.9 (95% CI 11.0-NR) mos, p=0.026]. Similarly, in the RR group, mOS was improved after alloSCT [55.0 (95% CI 6.9-NR) vs 4.1 (95% CI 2.6-10.4) mos, p=0.01]. Notably, there were 6 long-term survivors. 4 had isolated MECOMr, 1 had concomitant t(6;12)(q13;p13), and 1 had complex karyotype; 5 of them underwent alloSCT.
Conclusions: In this multicenter cohort, survival outcomes for MECOMr AML were overall poor after HMA+ven. Response rates were numerically higher in those treated in 1L compared to RR, but there was no difference in mOS following HMA+ven between the two groups. mOS was improved amongst those who responded to HMA+ven and underwent alloSCT.