The prognostic value of neutrophil-to-albumin ratio for dual outcomes in locally advanced esophageal cancer: an integrated analysis based on propensity score matching and Lasso-Cox regression.

OBJECTIVE(S): Locally advanced esophageal cancer is characterized by high malignancy and poor prognosis. Studies have indicated that preoperative inflammation and nutritional status significantly impact malignancies. This study aims to investigate the effect of the preoperative neutrophil-to-albumin ratio (NAR) on postoperative survival in esophageal squamous cell carcinoma (ESCC) patients treated with neoadjuvant chemotherapy combined with immunotherapy, as well as the construction and validation of a predictive model. MATERIALS AND

methodsClinical data from 222 patients with ESCC who underwent neoadjuvant chemotherapy combined with immunotherapy and surgery between 2020 and 2024 were retrospectively analyzed. The optimal cut-off value for NAR was determined using X-tile, and propensity score matching (PSM) was employed to balance baseline data between groups. Perioperative safety and survival (overall survival [OS], disease-free survival [DFS]) were compared and analyzed. A stepwise statistical strategy was adopted: univariate Cox regression and LASSO regression (to eliminate collinearity) were used to screen independent prognostic factors, followed by multivariate Cox regression. All core variables identified by LASSO (non-zero coefficients) were directly used to construct a prognostic nomogram. Model performance was internally validated using the Bootstrap method (1000 repetitions), and the predictive efficacy and clinical benefit of the model were evaluated using the concordance index (C-index), time-dependent receiver operating characteristic (ROC) curves, calibration curves, and decision curve analysis (DCA).

resultsThe optimal cut-off value for NAR was determined to be 0.07 by X-tile, dividing patients into a high-NAR group (127 patients) and a low-NAR group (95 patients). After PSM matching, each group comprised 95 patients. No statistically significant difference was observed in postoperative complications between the two groups (P > 0.05). Kaplan-Meier survival analysis showed that the low-NAR group had significantly better DFS (P = 0.018) and OS (P < 0.001) than the high-NAR group. For the OS model: univariate Cox regression and LASSO regression screened six core variables [NAR, ypTNM (pathological TNM staging after neoadjuvant therapy), TRG (tumor regression grade), post_cTNM_stage (clinical TNM staging after neoadjuvant therapy), post_SIRI (systemic immune-inflammation index after neoadjuvant therapy), postoperative_T_stage (T staging after surgery)]. Multivariate Cox regression confirmed that NAR (hazard ratio [HR] = 8.819), ypTNM staging, and TRG were independent prognostic factors. For the DFS model: LASSO regression screened five core variables (NAR, ypTNM, TRG, post_cTNM_stage, postoperative_T_stage, post_SIRI). Multivariate Cox regression similarly confirmed that NAR (HR = 1.994), ypTNM, and TRG were independent prognostic factors. Using the core variables selected by LASSO, internal validation demonstrated good model accuracy, with C-index values of 0.826 for OS and 0.723 for DFS. Time-dependent C-index analysis indicated that the model’s predictive stability within 1–5 years postoperatively was significantly superior to that of any single clinical indicator. Decision curve analysis (DCA) showed that the constructed nomogram model provided significantly higher net benefit across a wide range of threshold probabilities compared to single clinical indicators and the “treat-all” or “treat-none” strategies.

conclusionThe preoperative neutrophil-to-albumin ratio is an independent influencing factor for postoperative survival in ESCC patients treated with neoadjuvant chemotherapy combined with immunotherapy. The nomogram model, based on LASSO-selected variables and internally validated, can accurately quantify patients’ recurrence and survival risks, demonstrating significant clinical application value.