Preoperative NLR Combined with Ultrasound to Construct a Predictive Model for Central Lymph Node Metastasis in cN0 Papillary Thyroid Carcinoma

Abstract

Objective This study aims to evaluate the preoperative neutrophil-to-lymphocyte ratio (NLR) and ultrasound tumor characteristics to identify risk factors for central lymph node metastasis (CLNM) in clinically lymph node-negative (cN0) papillary thyroid carcinoma (PTC) patients. Based on these risk factors, a nomogram predictive model is developed to aid in formulating individualized treatment plans for clinical practice. Methods A retrospective analysis was conducted on 471 patients who underwent surgery and were pathologically diagnosed with PTC in our hospital between January 2021 and January 2022. Preoperative clinical data and ultrasound findings were collected, including NLR from routine blood tests, age, gender, maximum tumor diameter, tumor location, laterality (unilateral or bilateral), aspect ratio, tumor margin, presence of calcifications, tumor relationship with the thyroid capsule, preoperative thyroglobulin (TG), and thyroid-stimulating hormone (TSH). Based on the presence of CLNM, patients were categorized into a metastasis group and a non-metastasis group. Univariate and multivariate analyses were conducted using SPSS software to identify independent risk factors for CLNM. The patients were randomly divided into a training group and a validation group in a 7:3 ratio. Receiver operating characteristic (ROC) curves, calibration plots, and decision curve analysis were generated using R software to assess the feasibility of the predictive model. Results Among the 471 PTC patients, 215 cases (45.6%) were confirmed to have CLNM. Univariate analysis showed that factors such as gender, NLR, tumor diameter, tumor margin, calcifications, aspect ratio, tumor location, tumor relationship with the thyroid capsule, laterality, and TG levels were associated with CLNM. Multivariate logistic regression identified male gender, higher NLR, larger tumor diameter, irregular tumor margins, microcalcifications, and tumor invasion of the thyroid capsule as significant risk factors for CLNM in PTC patients. Based on these findings, a personalized nomogram was constructed and validated internally in two cohorts. The areas under the ROC curve (AUC) for the training and validation groups were 0.836 and 0.825, respectively, indicating good discriminatory power. Hosmer-Lemeshow test results demonstrated good model calibration in both groups. Calibration curves further confirmed the model's predictive accuracy. Conclusion High NLR, microcalcifications, irregular tumor margins, tumor diameter > 1 cm, thyroid capsule invasion, and male gender are independent risk factors for CLNM in PTC patients. The predictive model constructed based on these factors can significantly predict the probability of CLNM in PTC patients. This non-invasive tool will aid in preoperative lymph node assessment and provide effective guidance for surgeons in developing individualized treatment plans.