Venetoclax and Hypomethylating Agents Confirmed to be Standard Therapy in De Novo and ELN17 Not Adverse Acute Myeloid Leukemia Patients, Unfit for Chemotherapy According to SIE/Sies/GITMO Criteria. a Multicentric Real-Life Study on 379 Patients, By the Hematological Network Rete Ematologica Lombarda (REL)

Treatment of elderly acute myeloid leukemia (AML) is still challenging. In patients (pts) unfit for chemotherapy (CT) the best results are obtained by the association venetoclax (VEN) and hypomethylating agents (HMA) but, given its substantial toxicity compared to HMA alone, it is unclear whether in clinical practice it should be reserved to a proportion of relatively less “unfit” pts. Age, level of comorbidities and functional impairment, therapy benefits and risks, pts preferences, and AML features may influence the treatment choice.

Within the Rete Ematologica Lombarda (REL), we conducted a multicenter retrospective study, to assess the real-life experience in managing of elderly pts with newly diagnosed AML. In this report, we focused on pts defined as unfit for CT according the SIE/SIES/GITMO (“fitness”) criteria, a validated and comprehensive tool to define pts fitness for specific treatments. We evaluated the effect on survival (OS) of the presence of a specific criterion of unfitness, PS (ECOG), Charlsons Comorbidity Index (CCI), molecular data, and treatment received.

From Jan 20 to Dec 22, 726 pts (median age 75 years (y), range: 65-93) were diagnosed consecutively at 13 Hematology Units, 379 (52%) of whom were defined unfit for CT (median age 78 y, 50.6% de novo) and represent the study cohort. The majority of pts were classified as “unfit” according to the age criterion (>75 y) (65%), followed by the comorbidity criterion (22%) or by the presence of comorbidity and age criteria in 11% of cases. The most common comorbidities were severe pulmonary disease (30%), severe cardiopathy (10%), mental illness (9%), active infection (11%) or any other comorbidity according to physician’s judgment (40%). CCI was >3 in 43% and PS was >3 not-leukemia related in 2% of pts.

According to ELN17, evaluable in 273 pts, 18% were favorable (fav), 22% intermediate (int) and 60% adverse (adv). Mutations involved IDH1/2 in 34%, ASXL1 in 30%, TET2 in 30%, RUNX1 in 26%, TP53 in 20%, DNMT3A in 21%, NPM1 in 20%, K/NRAS in 21%, EZH2 in 19%, FLT3-ITD in 12% and splicing mutations in 33% of informative cases.

According to physicians’ decision, 75% of pts received non-intensive therapy (niT), HMA (azacytidine or decitabine) alone (27%, n=76) or in combination with VEN (73%, n=208). Only 1% of pts received CT and 23% best supportive care (BSC). The overall response rate in treated pts was 57%, including 45% of complete remission (CR), significantly higher in VEN-HMA than HMA [54% vs 20%, p:0.0001], including pts with adv ELN risk (40% vs 5.7%, p<0.0001) or secondary AML (s-AML) (40% vs 8.6%; p:0.0003).

After a median follow-up of 21.4 months (m) for living pts, the median OS of whole cohort was 13.6 m, 9.3 m in pts treated with niT and only 2 m in pts receiving BSC (p<0.0001). OS was significantly better in pts treated with VEN-HMA compared to HMA alone both overall (10 vs 7.2 m, p:0.002), and in the subgroups of pts categorized as unfit for severe comorbidities (12 vs 4 m, p<0.0001), in pts with CCI>2 (9.1 vs 7.3 m, p:0.019) and in pts with PS >2 (9.5 vs 6.8 m, p:0.027). In addition, VEN-HMA conferred better OS compared to HMA in de novo AML (13 vs 7m; p:0.0005), in fav and int ELN risk (not-reached vs 6.8 m in fav, p:0.04; 19.7 vs 3.7 m in int, p<0.0001, respectively). Conversely OS was similar between VEN-HMA and HMA alone in s-AML (8 vs 7 m, p: 0.3) and ELN adv risk (7.9 vs 5.8 m, p: 0.074).

Within VEN-HMA group, inferior survival was predicted by a diagnosis of s-AML (8 m vs 13 m in de novo, p:0.019), adv ELN (7.9 vs 10.4 m in int and 19.7 m in fav, p: 0.001), TP53 mutation (6.2 vs 16 m, p<0.001), absence of IDH mutations (8.9 vs 24.1 m; p 0.0022), absence of NPM1mutation (9.6 m vs not-reached, p:0.0007) or absence of DNMT3A (12.7 vs 33.7 m, p: 0.01). Conversely, no differences in OS were observed according to PS or CCI, to ELN risk or type of AML (de novo vs s-AML) in pts receiving HMA alone.

The main cause of death was disease progression (63%). The death rate in CR due to infection or induction-aplastic death was similar in VEN-HMA and in HMA (18% vs 17%).

This real-world study confirmed that VEN-HMA can be considered the standard of care for elderly AML pts, unfit for CT. Comorbidities or age should not limit the choice of VEN-HMA. However, in subgroups of unfit pts with s-AML or ELN adv-risk as well as with specific molecular characteristics, such as TP53 mutation, the combination obtains higher rates of CR, but it does not offer a survival advantage when compared to HMA alone.