[Development and validation of a nomogram for predicting cervical lymph node metastasis based on hematological...

Objective:To explore the predictive value of preoperative peripheral hematological parameters combined with clinicopathological features for cervical lymph node metastasis（CLNM） in patients with laryngeal squamous cell carcinoma（LSCC）, and to construct and validate a nomogram model for CLNM. Methods:A retrospective analysis was conducted on the clinical data of 264 LSCC patients who underwent surgical treatment and were pathologically confirmed, collected from the Second Affiliated Hospital of Shandong First Medical University and Taian 88 Hospital. Specifically, 161 patients from one hospital were allocated to the training cohort, while 103 patients from another hospital constituted the validation cohort. Based on postoperative pathological results, patients were categorized into CLNM-positive and CLNM-negative groups. The general clinical data, clinicopathological features, and hematological parameters of the two groups were analyzed and compared. A preoperative predictive model for CLNM was developed using logistic regression analysis, followed by validation and sensitivity analysis to evaluate the robustness of the model's predictive performance. Results:The results showed that there were significant differences in tumor location, tumor size, tumor differentiation, neutrophil percentage, lymphocyte count, lymphocyte percentage, c-reactive protein（CRP）, fibrinogen, neutrophil-to-lymphocyte ratio（NLR）, platelet-to-lymphocyte ratio（PLR）, systemic immune-inflammation index（SII）, systemic inflammation response index（SIRI）, and prognostic inflammatory index（PIV） between the CLNM-positive and CLNM-negative groups（P<0.05）. Lasso regression identified tumor location, clinical T stage, tumor size, tumor differentiation degree, red blood cell distribution width（RDW） -coefficient of variation（RDW-CV）, CRP, FIB, D-dimer, NLR, and lymphocyte-to-monocyte ratio（LMR） were the most predictive parameters. Multivariate logistic regression revealed that tumor location, tumor size, tumor differentiation degree, CRP, and NLR were independent risk factors for CLNM in LSCC patients（P<0.05）. A nomogram was constructed based on these five factors. The model demonstrated excellent discrimination, with a C-index of 0.837（95%CI 0.766-0.908） in the training cohort and 0.809（95%CI 0.698-0.920） in the validation cohort. Calibration curves and DCA curves in both cohorts confirmed the clinical utility of the model. Sensitivity analysis further supported the robustness of the results, showing good discrimination and calibration across different age and BMI subgroups. Conclusion:Tumor location, tumor size, tumor differentiation degree, CRP, and NLR were independent risk factors for CLNM in LSCC patients. The nomogram based on these variables exhibits strong discrimination, calibration, and clinical applicability, and may serve as a valuable tool for preoperative risk assessment and individualized treatment planning.