Establishing Benchmarks for Relapse-Free and Overall Survival Among Patients (Pts) with Newly Diagnosed AML Not Undergoing Allogeneic Stem Cell Transplantation (SCT) in First Remission

Background

Despite ongoing progress in understanding leukemogenic mechanisms and the development of novel therapeutic interventions, relapse continues to be a primary cause of treatment failure and mortality in pts with acute myeloid leukemia (AML). Allogeneic stem cell transplantation (alloSCT) remains the most effective post-remission treatment strategy for long-term survival and reduction of relapse risk. For pts who are ineligible for alloSCT, maintenance therapy has been proposed as a beneficial approach to reduce the risk of relapse, particularly for high-risk AML pts. The survival outcomes of pts who do not undergo alloSCT remain largely undefined. However, understanding these expected rates is crucial for guiding future research aimed at developing and integrating innovative post-remission maintenance and consolidation strategies. To this end, we investigated the characteristics and survival outcomes of pts with de-novo AML at first remission who did not undergo alloSCT.

Methods

We conducted a retrospective analysis of 1,686 highly annotated pts treated with first-line therapy for de novo AML at our institution from 2016 onwards. We excluded pts diagnosed with acute promyelocytic leukemia (APL, n=97) and core binding factor (CBF, n=83) AML, as these subtypes are rarely consolidated with alloSCT. To focus the survival analysis on the post-remission period, we included only pts who achieved complete remission (CR) or CR with incomplete hematologic recovery (CRi) after first-line therapy and did not proceed to alloSCT (n=355). Minimal residual disease (MRD) was measured using multiparameter flow cytometry. Baseline patient characteristics, genomic classifications, and treatment type were collected. Overall survival (OS) and relapse-free survival (RFS) were compared using the log-rank test. Univariate and multivariable Cox regression models were applied to assess the effect of covariates on OS and RFS.

Results

The median age of the cohort was 69 years (range 19-94), with 248 pts (70%) receiving lower-intensity therapy (LIT). MRD was negative in 55% of pts at the time of best response. Twenty-seven percent and 51% of the pts were classified as intermediate (INT) or adverse risk (ADV) by ELN 2022 classification, respectively. NPM1 and FLT3 mutations were detected in 29% and 14% of pts, respectively. TP53 mutations were detected in 18% of pts, and myelodysplastic (MDS) related gene mutations were observed in 33% of pts. Additionally, a complex karyotype was identified in 21% of the pts. Overall, the median OS for the entire cohort was 18.6 months [95% CI 15.1-22.8], and the median RFS was 11 months [95% CI 9-14]. Interestingly, there was no statistically significant difference in OS or RFS between patients aged < 60 yrs and those aged ≥ 60 years. The median RFS and OS of pts treated with IC were 26 and 28.2 months, respectively compared to 8.5 and 15.2 months, respectively among those treated with LIT. We next assessed the impact of venetoclax (VEN) on outcomes across both LIT and IT treatment groups. Pts receiving VEN had better RFS regardless of treatment intensity, with median RFS of 11 months (with VEN) [95% CI 9-15] vs 6 months [95% CI 5-10] in the LIT group, and not reached vs 16 months [95% CI 7-not reached] in the IC group. The effect on OS was not statistically significant. Median OS and RFS were significantly different by: MRD status (median OS and RFS of 24.6 and 17 months for MRD-negative vs 14 and 8 months for MRD-positive, respectively, p<0.001); TP53 mutation status (median OS and RFS of 23.3 and 14 months for wild-type vs 11.1 and 5 months for mutated TP53, respectively, p<0.001); and ELN2022 risk classification (median OS and RFS of 20.2 and 13 months for INT vs 14 and 8 months for ADV, respectively, p=0.03). After adjusting for potential confounders, TP53, MRD status, and ELN2022 risk classification remained significant in a Cox regression multivariable analysis (HR 2.34, 95% CI 1.6-3.2, HR 0.72, 95% CI 0.54-0.96 and HR 0.5, 95% CI 0.32-0.76, respectively, p<0.05 for all).

Conclusion

This comprehensive survival analysis of newly diagnosed AML pts who achieved CR/CRi and did not undergo alloSCT can serve as a benchmark for future trials on novel maintenance therapies for these pts in the post-remission period. A relapse-risk score based on disease and treatment characteristics can be useful to assess novel post-remission strategies.