Combined Conventional Ultrasound and Contrast‐Enhanced Computed Tomography for Cervical Lymph Node Metastasis Prediction in Papillary Thyroid Carcinoma

Shanshan Zhao,Wenwen Yue,Hui Wang,Jincao Yao,Chanjuan Peng,Xiatian Liu,Dong Xu
DOI: https://doi.org/10.1002/jum.16024
2022-06-02
Journal of Ultrasound in Medicine
Abstract:Objectives This study aimed to evaluate conventional ultrasound (US) combined with contrast‐enhanced computed tomography (CT) of the neck to predict central lymph node metastasis (CLNM) in clinical lymph‐negative patients with papillary thyroid carcinoma (PTC), establish a simple preoperative risk‐scoring model, and validate its effectiveness in a two‐center dataset. Methods A total of 423 patients with PTC preoperatively evaluated by US and contrast‐enhanced CT were included in the modeling group, and 102 patients from two hospitals were enrolled in the validation group. Independent predictive factors were determined using multivariate logistic regression analysis. Diagnostic performance was evaluated using receiver operating characteristic curve analysis. Results The independent predictive factors for CLNM were age ≤45 years (odds ratio [OR[ = 3.950), nodule presence in the non‐upper pole (OR = 2.385), nodule size >12.5 mm (OR = 2.130), Thyroid Imaging Reporting and Data System score ≥9 (OR = 2.857), normalized enhancement CT value ≥0.75 (OR = 3.132), central enhancement (OR = 0.222), and capsular invasion (OR = 3.478). The area under the curve (AUC) of the model was 0.790 (95% confidence interval [CI]: 0.747–0.834), and the sensitivity and specificity were 70.4% and 73.9%, respectively. The AUC in the validation group was 0.827 (95% CI: 0.747–0.907), and the sensitivity and specificity were 88.9% and 63.2%, respectively. Conclusions We found conventional US combined with contrast‐enhanced CT of the neck to be useful in predicting CLNM preoperatively and established a simple risk‐scoring model that might help surgeons with appropriate surgical plans and prognostic evaluation.
radiology, nuclear medicine & medical imaging,acoustics
What problem does this paper attempt to address?