The predictive value of a combined nomogram model integrating ultrasound and fine-needle aspiration cytology for large-number cervical lymph node metastases in clinically node-negative papillary thyroid carcinoma: a propensity score matching analysis.

Background: Large-number cervical lymph node metastases (CLNMs) in clinically node-negative (cN0) papillary thyroid carcinoma (PTC) patients are challenging to detect preoperatively, but significantly influence prognosis and treatment strategies. This study aimed to evaluate the predictive value of preoperative ultrasound (US) and fine-needle aspiration cytology (FNAC) for large-number CLNMs in cN0 PTC patients to guide early detection and treatment planning. Methods: A retrospective analysis was conducted of cN0 PTC patients who underwent total thyroidectomy and cervical lymphadenectomy across six hospitals from September 2018 to June 2024. Based on the presence or absence of large-number CLNMs, patients were grouped accordingly using 1:1 propensity score matching (PSM) for age and sex. Univariate and multivariate logistic regression analyses were conducted to identify independent predictive US, FNAC, and their combined (US-FNAC) factors. A nomogram was constructed based on the independent predictive factors derived from the US-FNAC combination, and its performance and clinical applicability were subsequently assessed through receiver operating characteristic (ROC) analysis, calibration plots, and decision curve analysis (DCA). Results: The univariate and multivariate logistic regression analyses identified the following independent predictors of large-number CLNMs in PTC patients: escape-like arrangement (P=0.016), nuclear elongation (P<0.001), nuclear outline (P<0.001), nucleolus (P=0.043), size (P<0.001), margin (P=0.003), and calcification (P=0.045). The developed nomogram demonstrated good predictive performance with an area under the curve (AUC) of 0.830 [95% confidence interval (CI): 0.774-0.885] in the training cohort and 0.810 (95% CI: 0.718-0.902) in the test cohort. In the training cohort, the nomogram achieved an accuracy of 76.5%, a sensitivity of 81.4%, and a specificity of 71.8%. While in the test cohort, it achieved an accuracy of 76.7%, a sensitivity of 78.4%, and a specificity of 57.3%. Conclusions: In cN0 PTC patients, large-number CLNMs are associated with several factors, including size, escape-like arrangement, nuclear elongation, nuclear outline, nucleolus, margin, and calcification. Clinicians should pay closer attention to patients with these risk factors. Further, the nomogram developed using these risk factors exhibited promising predictive performance for large-number CLNMs in PTC patients.