Development and Validation of a Risk Prediction Model for Neutropenia in Acute Myeloid Leukemia Patients Treated with Venetoclax

OBJECTIVE To investigate the independent risk factors for grade 4 neutropenia in acute myeloid leukemia(AML) patients treated with venetoclax and to construct a nomogram prediction model.

METHODS A retrospective analysis was conducted on the clinical data of hospitalized AML patients who received venetoclax treatment and underwent therapeutic drug monitoring(TDM) at the First Affiliated Hospital of Nanchang University between August 2022 and March 2025. Patients were divided into a case group and a control group based on the occurrence of grade 4 neutropenia. Univariate analysis was used to screen potential risk factors, and binary Logistic regression analysis was employed to identify independent risk factors. A nomogram prediction model was constructed based on the independent risk factors and validated. The clinical utility of the nomogram was evaluated using decision curve analysis(DCA).

RESULTS Among 138 patients, grade 4 neutropenia occurred in 53 cases(38.4%). Univariate analysis revealed that dosage, venetoclax C₀, venetoclax C₆, underlying disease, concomitant use of CYP3A inhibitors, white blood cell count(WBC), red blood cell count(RBC), hemoglobin, platelet count, direct bilirubin, albumin(ALB), globulin, and uric acid were significantly associated with the occurrence of grade 4 neutropenia(P<0.05). Binary Logistic regression analysis indicated that elevated C₀ (OR=2.794, 95% CI: 1.02–7.65, P=0.046), presence of underlying disease (hypertension) (OR=30.185, 95% CI: 1.85–491.35, P=0.017), decreased WBC(OR=0.490, 95% CI: 0.29–0.84, P=0.009), decreased RBC (OR=0.021, 95% CI: 0.001–0.46, P=0.014), and ALB<35 g·L⁻¹ (OR=11.042, 95% CI: 1.82-66.99, P=0.009) were independent risk factors for grade 4 neutropenia. The nomogram prediction model constructed based on these five indicators had an area under the receiver operating characteristic curve of 0.929 (95% CI: 0.89–0.97). The calibration curve demonstrated high consistency between the model-predicted risk and the actual observed risk. DCA further confirmed that within a reasonable high-risk threshold range of 0.1 to 0.4, the clinical application of this model could provide significant net benefit for patient management.

CONCLUSION Venetoclax C₀, underlying disease (hypertension), WBC, RBC, and ALB levels can serve as key indicators for predicting the occurrence of grade 4 neutropenia in AML patients treated with venetoclax. The constructed model can assist clinicians in early identification of high-risk patients and intervention, thereby optimizing the clinical application of venetoclax.