Construction of a comprehensive nutritional index and comparison of its prognostic performance with the PNI and NRI for survival in older patients with nasopharyngeal carcinoma: a retrospective study.

ABSTRACT

OBJECTIVES: To explore the relationship between the Comprehensive Nutritional Index (CNI) and survival in older patients with nasopharyngeal carcinoma (NPC) and to compare the prognostic performance of three nutritional indicators (CNI, Prognostic Nutritional Index (PNI), and Nutritional Risk Index (NRI)) for overall survival (OS). METHODS: This retrospective study involved 309 older NPC patients in Guangzhou (China) from November 2006 to November 2017. The CNI comprised five parameters the body mass index (BMI), usual body weight percentage (UBW%), hemoglobin (Hb) level, albumin level, and total lymphocyte count (TLC). All single nutritional indicators were evaluated before and immediately after treatment. The principal component analysis (PCA) was used for calculation of the CNI by single nutritional indicators after treatment. The cutoff point for the CNI was evaluated and logistic regression used to explore the risk factors for the CNI. Univariable, multivariable Cox regression, and Kaplan-Meier methods were applied for OS and disease-free survival (DFS) analyses. Cox proportional hazards models were used to compare the prognostic value of the CNI, PNI, and NRI for OS. RESULTS: All single nutritional indicators decreased significantly after treatment (P < 0.05). The CNI cutoff point for mortality was 0.027, and the logistic regression indicated more complex treatments or higher cancer stage for NPC was associated with a low CNI (HR = 0.179; 95% CI 0.037-0.856; 0.545, 0.367-0.811, respectively). In multivariable Cox regression, the CNI remained an independent prognostic factor of OS and DFS (HR = 0.468, 95% CI 0.263-0.832; 0.527, 0.284-0.977, respectively). Kaplan-Meier curves showed that a low CNI was associated with worse OS and DFS (P = 0.001 and 0.013, respectively). The prognostic predictive performance of the CNI was superior to that of the PNI or NRI. CONCLUSIONS: The CNI can be recommended as an appropriate indicator reflecting the integrated nutritional status of older NPC patients. A low CNI predicted a poor survival outcome and the prognostic performance of CNI was superior to PNI or NRI.