Cholesterol-modified prognostic nutritional index (CPNI) as an effective tool for assessing the nutrition status and predicting survival in patients with breast cancer

Abstract

Abstract Background Malnutrition is associated with poor overall survival (OS) in breast cancer patients; however, the most predictive nutritional indicators for the prognosis of patients with breast cancer are not well-established. This study aimed to compare the predictive effects of common nutritional indicators on OS and to refine existing nutritional indicators, thereby identifying a more effective nutritional evaluation indicator for predicting the prognosis in breast cancer patients. Methods This prospective study analyzed data from 776 breast cancer patients enrolled in the “Investigation on Nutritional Status and its Clinical Outcome of Common Cancers” (INSCOC) project, which was conducted in 40 hospitals in China. We used the time-dependent receiver operating characteristic curve (ROC), Kaplan–Meier survival curve, and Cox regression analysis to evaluate the predictive effects of several nutritional assessments. These assessments included the patient-generated subjective nutrition assessment (PGSGA), the global leadership initiative on malnutrition (GLIM), the controlling nutritional status (CONUT), the nutritional risk index (NRI), and the prognostic nutritional index (PNI). Utilizing machine learning, these nutritional indicators were screened through single-factor analysis, and relatively important variables were selected to modify the PNI. The modified PNI, termed the cholesterol-modified prognostic nutritional index (CPNI), was evaluated for its predictive effect on the prognosis of patients. Results Among the nutritional assessments (including PGSGA, GLIM, CONUT, NRI, and PNI), PNI showed the highest predictive ability for patient prognosis (time-dependent ROC = 0.58). CPNI, which evolved from PNI, emerged as the superior nutritional index for OS in breast cancer patients, with the time-dependent ROC of 0.65. It also acted as an independent risk factor for mortality (p < 0.05). Moreover, the risk of malnutrition and mortality was observed to increase gradually among both premenopausal and postmenopausal age women, as well as among women categorized as non-overweight, overweight, and obese. Conclusions The CPNI proves to be an effective nutritional assessment tool for predicting the prognosis of patients with breast cancer.