Abstract
Background. NPM1 and FLT3-ITD are the most frequently observed mutations in acute myeloid leukemia (AML). However, the intensity of treatment in elderly patients (pts) with these mutations is notwelldefined.
Methods. Within the Rete Ematologica Lombarda, we conducted a retrospective analysis of 718 AML pts aged ≥65 y, diagnosed between 2020 and 2022. In pts harboring NPM1 and/or FLT3-ITD mutations (n=151), we evaluated clinical and molecular features, treatments received —either chemotherapy (CT) or non-intensive therapy (niT: hypomethylating agents [HMA] ± venetoclax [VEN])—and their impact on overall survival (OS).
Results. NPM1 and/or FLT3-ITD mutations were found in 110/555 (19%) and in 68/527 (12.9%) molecularly evaluable pts, respectively; additionally, 30 pts harbored both mutations. The median age was 73 y for each group (range 65-90). Most pts had a good Charlson Comorbidity Index (CCI) and a good ECOG performance status (PS) [CCI<2: 115/151 (73.5%); PS <2: 98/151 (85%)], in both mutational groups. According to SIE/SIES/GITMO fitness criteria, 66 pts (44%) were classified as fit, 68 (45%) as unfit, and 17 (11%) as frail. All frail pts received best supportive care (BSC), regardless of mutational status.
Among NPM1-mutated pts, 95 (86%) had a de novo AML and 5 of 103 evaluable (4.9%) had an adverse karyotype. According to fitness criteria, 52 pts (47%) were classified as fit, 47 (43%) unfit, and 11 (10%) as frail. In fit pts, CT was given in 46/52 (88%), niT (VEN+HMA) in 6. All unfit pts received niT: 39 (83%) VEN+HMA and 8 (17%) HMA alone. Ten pts (10%) underwent allogeneic stem cell transplantation (alloSCT).
In the FLT3-ITD-mutated pts, 49 (72%) had a de novo AML and 7 of 63 evaluable pts (11%) had an adverse karyotype. Thirty-four pts (50%) were fit, 25 (36.7%) unfit, and 9 (13%) frail. Among fit pts, CT was administered in 28/34 (82%), niT (VEN+HMA) in 6 pts. Among unfit pts, 22 received niT — 14 (56%) VEN+HMA and 8 (32%) HMA alone — and 3 (12%) BSC. AlloSCT was performed in 12 pts (18%).
Complete remission (CR) was achieved in 71% (n=78) of NPM1 pts and in 53% (n=36) of FLT3-ITD pts, without statistically significant differences according to treatment received (NPM1: 89% CT vs 73.5% niT, p=0.07; FLT3: 68% CT vs 59% niT, p=0.5). Relapse occurred in 29 (37%) NPM1 pts and 21 (58%) FLT3-ITD pts.
At a median follow-up of 25 months (mo), median OS was 18.8 mo in NPM1 pts and 10.8 in FLT3-ITD pts.
In NPM1 mutated pts, longer OS was observed in pts without adverse karyotype (26 vs 6 mo; p=0.002), with de novo AML (26 vs 7.6 mo, p= 0.012), and in pts achieving CR (45.9 vs 6.3 mo; p=0.000). No significant OS differences were observed by PS, CCI, fitness (fit vs unfit: 30 vs 19 mo; p=0.3), FLT3-ITD co-mutation (26.3 vs 18.6 mo, p=0.5), treatment type (CT vs VEN+HMA: not reached vs 24.3 mo; p=0.2), or alloSCT (17.2 vs 18.8 mo; p=0.7). Among fit pts, OS was not significantly different by treatment (CT: 30 mo vs VEN+HMA:18.8 mo; p=0.48). Among unfit pts, OS was not reached with VEN+HMA group vs 6.9 mo with HMA-alone (p=0.005).
In FLT3-ITD cohort, significantly better OS was observed in pts without adverse karyotype (13.6 vs 4.4 mo; p=0.002), with co-mutation NPM1 (18.6 vs 7.8 mo; p=0.003), in pts undergoing alloSCT (20.2 vs 8.1 mo; p=0.014), and in those achieving CR (18.8 vs 4.4 mo; p=0.0001). No significant OS differences were observed by AML type, PS, CCI, fitness (fit vs unfit: 15 vs 13.4 mo; p=0.29). Among fit pts, OS did not differ by treatment (15 mo with CT vs 18.8 mo with VEN+HMA; p=0.9). Among unfit pts, OS was superior in pts treated with VEN+HMA to HMA alone (13.6 vs 4.5 mo; p=0.045).
Achieving CR was the only independent predictor factor of OS in the multivariable analysis.
Conclusions. In this real-world cohort of elderly AML pts with NPM1 and/or FLT3-ITD mutations, treatment intensity modulated by their fitness status, clinical outcome was mainly influenced by disease biology (adverse karyotype, de novo AML) and particularly by the achievement of CR. The combination VEN+HMA improves OS compared to HMA alone in unfit pts across both mutational groups. In fit pts OS was similar between CT and VEN+HMA, although in NPM1 pts treated with CT a longer follow-up could highlight a better OS. Comutational status had a similar prognosis to NPM1 mutation alone in both fit and unfit pts. AlloSCT improved outcome in FLT3-ITD-mutated pts; however, the prognosis remains poor highlighting the need for tailored strategies.
