Cancer Imaging (Jun 2023)

Nomogram for preoperative estimation risk of lateral cervical lymph node metastasis in papillary thyroid carcinoma: a multicenter study

  • Jialin Zhu,
  • Luchen Chang,
  • Dai Li,
  • Bing Yue,
  • Xueqing Wei,
  • Deyi Li,
  • Xi Wei

DOI
https://doi.org/10.1186/s40644-023-00568-5
Journal volume & issue
Vol. 23, no. 1
pp. 1 – 13

Abstract

Read online

Abstract Background Lateral lymph node metastasis (LLNM) is frequent in papillary thyroid carcinoma (PTC) and is associated with a poor prognosis. This study aimed to developed a clinical-ultrasound (Clin-US) nomogram to predict LLNM in patients with PTC. Methods In total, 2612 PTC patients from two hospitals (H1: 1732 patients in the training cohort and 578 patients in the internal testing cohort; H2: 302 patients in the external testing cohort) were retrospectively enrolled. The associations between LLNM and preoperative clinical and sonographic characteristics were evaluated by the univariable and multivariable logistic regression analysis. The Clin-US nomogram was built basing on multivariate logistic regression analysis. The predicting performance of Clin-US nomogram was evaluated by calibration, discrimination and clinical usefulness. Results The age, gender, maximum diameter of tumor (tumor size), tumor position, internal echo, microcalcification, vascularization, mulifocality, and ratio of abutment/perimeter (A/P) > 0.25 were independently associated with LLNM metastatic status. In the multivariate analysis, gender, tumor size, mulifocality, position, microcacification, and A/P > 0.25 were independent correlative factors. Comparing the Clin-US nomogram and US features, Clin-US nomogram had the highest AUC both in the training cohort and testing cohorts. The Clin‑US model revealed good discrimination between PTC with LLNM and without LLNM in the training cohort (AUC = 0.813), internal testing cohort (AUC = 0.815) and external testing cohort (AUC = 0.870). Conclusion Our findings suggest that the ClinUS nomogram we newly developed can effectively predict LLNM in PTC patients and could help clinicians choose appropriate surgical procedures.

Keywords