Clinical outcome of patients with TP53 AML treated with non-intensive vidaza-venetoclax chemotherapy : Venaura retrospective regional multicenter study.

Abstract

Abstract Acute myeloid leukemias (AML) with TP53 mutations represent a distinct biological entity, recently recognized in the WHO 2022 and ICC classifications, due to their extremely poor prognosis. TP53 mutations are associated with high level of chemoresistance, genomic instability and a rapid disease progression. Patients with TP53mut AML are often ineligible to intensive chemotherapy due to age or co-morbidities, but also due to limited benefit on survival of such approaches. More recently, azacitidine and venetoclax (AZA/VEN) has emerged as a potential therapeutic option in this high-risk population. Based on prospective studies, AZA/VEN seems to improve response rate but without significant survival benefit in patients with poor-risk cytogenetics and TP53mut AML. However, there are very few data regarding TP53mut AML patients treated with AZA/VEN in real life settings. In this retrospective study, we aimed to evaluate the determinants of outcomes in TP53mut AML patients treated with frontline AZA/VEN. Data were retrospectively collected in the VENAURA registry (n = 755) from 10 different French centers (Saint-Etienne, Clermont-Ferrand, Grenoble, Lyon (Hopital Lyon Sud, Centre Léon Bérard), Nîmes, Villefranche sur Saône, Annecy, Valence, Roanne) in Auvergne Rhône Alpes (AURA) region, between January 2019 and February 2024. Overall response rate was defined as in VIALE-A trial with a composite complete remission (CRc) combining complete remission (CR), complete remission with incomplete hematopoietic lineage recovery (CRi) and morphological leukemia-free state (MLFS). We also investigated the relationship between TP53 mutational load (VAF) and treatment response rate, with a cut-off at 23% to distinguish mono-allelic from bi-allelic forms according to the publication of Bahaj W et al (2023). Overall, we identified 72 patients with TP53mut AML. With a median age of 73 years (31-86), our cohort was highly enriched in secondary type AML (21 post MDS, 4 post MPN, 11 AML-pCT) while only 6/72 patients were de novo AML and 28/72 harboured MDS-related gene mutation (MRC). Most frequent co-occuring mutations were DNMT3A (n=12), TET2 (n=10), JAK2 (n=9) and U2AF1 (n=7). Cytogenetics was predominantly complex (56 monosomal complex, 4 non-monosomal complex) while diploid karyotypes were unfrequent (5.5%). Mutation profiling showed 11/72 patients with mono-allelic TP53 alteration (VAF < 23% and no del17p or only del17p) and 61/72 patients with bi-allelic alteration (VAF ≥ 23% and/or del17p in addition to the TP53 mutation). Median TP53 VAF was 52% (range: 7-93). With a median number of 3 cycles (range: 1-28), CRc rate was 54% (39/72) with 21 patients (29.2%) achieving CR, 6 (8.3%) achieving CRi/CRp and 12 patients (16.6%) achieving MLFS. Eight patients (11.1%) had a partial response (PR), while 25 patients (34.7%) experienced treatment failure or could not be evaluated due to early death or missing data. With a median time to achieve best response of 31 days (range: 19-119) and a median time to best response of 1 cycle, median duration of response was 151 days (range: 3-596). Following treatment, 6 patients (8.3%) underwent allogeneic hematopoietic stem cell transplant. Of these, 3 experienced early relapses within 6 months post-transplant. The median overall survival (OS) in our population was 7.4 months (range: 0-39.8). Stratification by allelic status showed a median OS of 11 months for monoallelic patients (with VAF < 23%), versus 5 months for those with biallelic alteration (VAF ≥ 23% or VAF<23% and del17p). Thus, even if it's not statistically significant (p=0.12), AZA/VEN appears more beneficial in patients with lower mutational burden. In univariate and multivariate analysis, only performance status and leucocyte count appear to be predictive of outcome. Additionally, survival was increased in responders (CR/CRi/MFLS) compared to partial responders and non-responders with a median OS of 14 months versus 6.6 months and 2 months respectively (p<0.05). In conclusion, TP53mut AML remains a highly aggressive disease with poor prognosis. However, stratification based on allelic status using a 23% VAF cutoff may have a prognostic value and help to guide therapeutic decisions. Some patients achieved CRc with durable responses, highlighting the potential benefit of AZA/VEN treatment in such patients. Finally, biallelic or non-biallelic involvement and response to treatment appear to be two interesting prognostic factors in our cohort.