Abstract
OBJECTIVES: This study aimed to investigate whether the body composition metrics extracted from quantitative CT (QCT) are associated with the survival prognosis of acute myeloid leukemia (AML) patients and to evaluate the impact of a nomogram based on QCT and clinical-physical factors in predicting the prognosis of AML. METHODS: The clinical factors and QCT metrics of 127 AML patients undergoing initial chest CT were analyzed retrospectively. The AML patients were divided into favorable and poor prognosis groups based on the threshold of median overall survival (OS). A QCT metrics- and clinical factors-derived nomogram was constructed using multivariate Cox regression. The performance of the nomogram was assessed with a receiver operating characteristic curve (ROC), calibration curve, and decision curve analysis (DCA). RESULTS: Compared to patients in the favorable survival prognosis group, patients with poor prognosis were older (p = 0.027), had higher risk stratification (p = 0.006), more positive minimal residual disease (MRD) (p = 0.014), lower skeletal muscle index (SMI) (p = 0.045), and a higher incidence of volumetric bone mineral density (vBMD) ≤ 120 (p = 0.035). Older age, higher risk stratification, positive MRD, and SMI < 15.74cm(2)/m(2) were independent risk factors for poor prognosis in AML patients. The areas under the ROC curve (AUCs) of the nomogram, which included SMI and independent clinical factors, for predicting 1- and 2-year OS were 0.792 and 0.794, respectively. The calibration curve and DCA demonstrated the good performance of the nomogram prediction model. CONCLUSIONS: Sarcopenia revealed by QCT, integrated into a nomogram with age, risk stratification, and MRD, can facilitate individualized prediction of survival prognosis in AML patients.