Development of a clinical prediction model for sensitivity to combination therapy of Bcl-2 inhibitors and hypomethylating agents in elderly/unfit patients with acute myeloid leukemia

Objective This study aims to develop a clinical prediction model for sensitivity to Bcl-2 inhibitors combined with hypomethylating agents (HMAs) in elderly/unfit patients with acute myeloid leukemia (AML). Methods Clinical data, including French-American-British (FAB) classification, chromosomal karyotype, and second-generation sequencing results, were retrospectively collected from consecutive elderly/unfit patients with AML treated with Bcl-2 inhibitors in combination with HMAs between September 2019 and March 2024. Treatment efficacy was assessed in all patients. Logistic regression and Akaike information criterion were used to identify risk variables affecting efficacy. A nomogram was developed based on these variables to assess patient sensitivity to the treatment regimen. The performance of the nomogram was evaluated using a receiver operating characteristic (ROC) curve, calibration plot, and decision curve analysis (DCA). Results This study included 209 patients with AML. The FAB classification, AML type, AML status, prior HMAs exposure, chromosomal karyotype, and mutations in ASXL1 , FLT3 , IDH , NPM1 , and CEBPA were screened to develop the nomogram. The area under the ROC curve indicated a discriminatory power of 0.900 (95% CI, 0.860–0.941). The calibration curve suggested favorable concordance between the predicted and actual occurrence probabilities ( P  = 0.849). DCA revealed a net clinical benefit when the threshold probability ranged from 0 to 0.98. Internal validation, performed 500 times using the bootstrap method, demonstrated a satisfactory model performance in the validation set. Conclusion A prediction model was developed and validated to serve as a decision-making tool for physicians treating elderly/unfit patients with AML prior to initiating therapy with Bcl-2 inhibitors combined with HMAs.