Long-Term Survival Outcomes and Cytogenetic/Molecular Patterns of Relapse in Adults with FLT3-Mutated AML Receiving Frontline Triplet Therapy with a Hypomethylating Agent, Venetoclax and FLT3 Inhibitor

Background: Several small studies have suggested improved outcomes by adding a FLT3 inhibitor (FLT3i) to hypomethylating agent (HMA)/venetoclax in older patients (pts) with newly diagnosed FLT3-mutated acute myeloid leukemia (AML). In this large, pooled analysis of adults with FLT3-mutated AML who received a frontline triplet regimen, we assessed the long-term clinical outcomes with this approach. We also sought to characterize the molecular patterns of relapse in these pts, which could help inform future combination studies.

Methods: We retrospectively analyzed pts with newly diagnosed FLT3-mutated AML treated at our institution with a frontline “triplet” regimen consisting of an HMA, venetoclax, and a FLT3i. Pts with treated secondary AML who had previously received an HMA or chemotherapy for an antecedent hematologic neoplasm were excluded. Molecular analysis was performed at diagnosis and at the time of relapse to characterize patterns of clonal evolution associated with relapse.

Results: Between 9/2017 to 6/2024, 88 pts with newly diagnosed FLT3-mutated AML received a triplet regimen containing a FLT3i. The median age was 70 years (range, 18-88); 26 pts (30%) were ≥75 years. FLT3 mutation was ITD only in 68 pts (77%), D835/D836 only in 15 pts (17%), ITD + D835/D836 in 3 pts (3%), and N676K in 2 pts (2%). The most common co-mutations were DNTMT3A in 40 pts (46%), NPM1 in 39 (44%), RUNX1 in 22 (25%), TET2 in 20 (23%), and WT1 in 18 (21%). ELN 2022 risk was adverse in 48 pts (55%). The FLT3i used was gilteritinib in 61 pts (69%), quizartinib in 18 (21%), sorafenib in 7 (8%), and midostaurin in 2 (2%).

The CR/CRi rate was 92% (CR 83%; CRi 9%); an additional 6% achieved MLFS as best response (CR/CRi/MLFS rate: 98%). 2 pts (2%) died during induction. Among 71 responders with adequate samples for flow MRD, 57 (80%) achieved MRD negativity (sensitivity: 10^-4). 32 pts (36%) underwent allogeneic SCT in first remission.

With a median follow-up of 25.5 months (range, 1-55.5 months), the median RFS for the entire cohort was 16.7 months and the estimated 3-year RFS was 40%; the median OS was 28.1 months, and the estimated 3-year OS was 46%. For pts with FLT3-ITD, the median OS was 24.5 months and the 3-year OS was 40%; for pts with only non-ITD mutations, the median OS was 39.3 months and the 3-year OS was 67%. The 3-year OS was similar for pts with NPM1 mutation vs NPM1 wild type (53% and 41%, respectively; P=0.75), for pts with ELN 2022 adverse vs favorable/intermediate risk disease (44% and 49%, respectively; P=0.60), and for pts <75 years vs ≥75 years (48% and 42%, respectively; P=0.65). Pts with a baseline RAS pathway mutation (including NRAS, KRAS, PTPN11, CBL, NF1 and/or BRAF) had worse survival than those without one of these mutations (3-year OS 20% vs 56%, respectively; P=0.09). In a landmark analysis, there was no benefit for SCT in the entire cohort (3-year OS 53% for SCT vs 52% for no SCT; P=0.92), in FLT3-ITD-mutated pts (3-year OS 46% vs 49%, respectively; P=0.96), nor in pts <75 years of age (3-year OS 52% vs 55%, respectively; P=0.73).

Overall, 24 pts relapsed (28% of responders). 1 pt had extramedullary-only relapse (CSF and skin). 20 pts underwent repeat cytogenetic analysis and molecular sequencing at relapse. FLT3 was undetectable by PCR (assay sensitivity: ~1%) in 12 pts (60%) at relapse. 13 pts (65%) had evidence of clonal evolution at relapse (6 pts with new mutations, 3 pts with new chromosomal abnormalities, and 4 pts with both). No recurring patterns of chromosomal changes were observed. 2 pts (10%) had an emergent FLT3 TKD mutation (1 after gilteritinib and 1 after sorafenib). Other common emergent mutations included RAS pathway mutations in 4 pts (20%, including KRAS/NRAS in 2 pts and PTPN11 and CBL in 1 pt each), GATA2 in 3 (13%), spliceosome mutations in 2 (8%, including SF3B1 and ZRSR2 in 1 pt each), IKZF1 in 2 (8%) and TET2 in 1 (4%). Outcomes were poor after relapse, with a median OS of only 4.5 months.

Conclusion: In this large cohort, triplet therapy with an HMA, venetoclax and FLT3i results in durable remission and encouraging long-term survival in older adults with newly diagnosed FLT3-mutated AML (median OS 28.1 months; 3-year OS 46%). These results are superior to historical expectations with HMA plus venetoclax in FLT3-mutated AML. A majority of relapses are driven entirely by FLT3 wild-type clones. RAS pathway mutations at diagnosis are associated with worse outcomes, and new RAS pathway mutations were observed in 20% of relapses.