Abstract
BACKGROUND: Medullary thyroid carcinoma (MTC) frequently presents with lateral lymph node metastasis (LLNM), a critical determinant of postoperative recurrence. While surgery remains the cornerstone of MTC treatment, the indications for lateral lymph node dissection (LLND) remain contentious. This study aimed to develop and validate a predictive nomogram for assessing LLNM risk in patients with MTC. METHODS: We retrospectively analyzed 87 treatment-naïve MTC patients who underwent primary surgical resection at our institution. Univariate and multivariate logistic regression analyses were performed to identify independent risk factors for LLNM. A nomogram was constructed and internally validated, with its clinical utility evaluated through discrimination, calibration, and decision curve analyses. RESULTS: Univariate analysis identified multifocality, intrathyroidal lymphovascular invasion (IT-LVI), extrathyroidal extension (ETE), central lymph node metastasis (CLNM), maximum tumor diameter (MTD), serum calcitonin (Ctn), and carcinoembryonic antigen (CEA) as significantly associated with LLNM (P < 0.05). Multivariate logistic regression analysis revealed ETE (OR = 14.37; 95% CI: 2.11-100.24; P = 0.007), CLNM (OR = 4.97; 95% CI: 1.06-23.26; P = 0.042), and natural log-transformed Ctn (Ln_Ctn) (OR = 2.72; 95% CI: 1.49-4.99; P<0.001) as independent predictors. The resulting nomogram demonstrated excellent discriminative ability (AUC = 0.941), good calibration, and strong clinical utility. CONCLUSION: We developed a novel nomogram incorporating ETE, CLNM, and Ln_Ctn to accurately estimate LLNM probability in MTC patients. This predictive model significantly improves risk stratification, provides valuable guidance for surgical decision-making regarding LLND, and supports personalized surgical planning.