Heliyon (Jan 2025)
Development and validation of a prognostic nomogram incorporating neutrophil-to-albumin ratio for predicting overall survival in patients with nasopharyngeal carcinoma undergoing concurrent chemoradiotherapy
Abstract
Background: Recent research suggests that the emerging neutrophil-albumin ratio (NAR) has a significant correlation with the survival outcomes across a range of tumors, yet its predictive significance for nasopharyngeal carcinoma (NPC) remains insufficiently investigated. This study aimed to evaluate the relationship between the neutrophil-to-albumin ratio (NAR) and overall survival (OS) in patients with NPC, as well as to develop a corresponding prognostic model. Methods: This retrospective analysis included 861 NPC patients treated with concurrent chemoradiotherapy (CCRT), who were randomly divided into a training group (n = 605) and a validation group (n = 256). To identify factors associated with OS and construct a prognostic nomogram, both univariate and multivariate Cox regression analyses were performed. The nomogram's prognostic accuracy was evaluated and independently validated. Results: The NAR score successfully segregated NPC patients into two categories with significantly different OS (HR = 0.536; 95 % CI: 0.296–0.972, P = 0.040). Through multivariate analysis, factors such as age, T stage, N stage, and NAR score were identified as independent predictors of OS, leading to the creation of a prognostic nomogram. This nomogram demonstrated superior predictive capability for OS [C-index = 0.702 (95 % CI: 0.636–0.768)], surpassing that of the conventional staging system [C-index = 0.651 (95 % CI: 0.549–0.752)]. The findings underwent internal validation within an independent cohort. Conclusions: The NAR, an emergent biomarker combining nutritional and inflammatory status, offers a practical, low-cost, and non-invasive prognostic measure for NPC patients treated with CCRT. Additionally, the prognostic nomogram derived from NAR surpasses traditional staging systems in predictive accuracy.