Novel metabolic prognostic score for predicting survival in patients with cancer

Cancer is a fatal disease with a high global prevalence and is associated with an increased incidence of metabolic disorders. This study aimed to develop a novel metabolic prognostic system to evaluate the overall metabolic disorder burden in cancer patients and its relationship with their prognosis. The patients in this study were enrolled from the Investigation on Nutrition Status and Clinical Outcome of Common Cancers (INSCOC) project. The least absolute shrinkage and selection operator (LASSO) analysis was used to screen for indicators of metabolic disorders. Cox regression analysis was used to evaluate the independent association between indicators of metabolic disorders and mortality in patients. The Kaplan-Meier method was used to evaluate the survival of patients with varying burdens of metabolic disorders. Finally, nomogram prognostic models and corresponding calculators were constructed and evaluated using the areas under the receiver operating characteristic curves (AUC), decision curve analysis (DCA), and calibration curves. Five of the 19 hematological indexes, including hemoglobin, neutrophils, direct bilirubin, albumin, and globulin, were selected as the evaluation indicators of metabolic disorder burden and independent risk factors for prognosis in cancer patients. Patients with a higher metabolic disorder burden had poorer survival rates. The AUC of the 1-year, 3-year, and 5-year overall survival of the prognostic nomogram was 0.678, 0.664, and 0.650, respectively. DCA and calibration curves indicated that the clinical benefit rate of metabolic disorder burden prognostic markers was high. Patients with a higher metabolic disorder burden had poorer survival rates. The nomogram and corresponding calculator can accurately evaluate the metabolic disorder burden and predict the prognosis of patients with cancer.