Ultrasound and clinicopathological characteristics of papillary thyroid carcinoma predict the coexistence of TERT promoter and BRAFV600E mutations

Abstract

Abstract Purpose: The coexistence of TERT promoter and BRAFV600E mutations is strongly associated with high aggressiveness and poor prognosis in papillary thyroid carcinoma. The aim of this study was to construct a preoperative and postoperative predictive coexisting mutation model based on ultrasound and clinicopathological characteristics for the prognostic risk stratification of papillary thyroid cancer that can guide the choice of clinical treatment modalities. Methods: Retrospective analysis of the ultrasound and clinicopathological characteristics of 113 patients with a surgical pathology of papillary thyroid carcinoma with TERT promoter and BRAFV600E gene testing results in the Affiliated Hospital of Jining Medical University from December 2020 to August 2022. Correlations between ultrasound and clinicopathological characteristics and combined mutations were analyzed by univariate and multivariate binary logistic regression, independent predictors were screened, and nomograms were constructed. The performance of the risk prediction model was assessed by plotting receiver operating characteristic curves (ROC curves), calibration curves, and decision curves. Results: The multivariate logistic regression analyses determined that tumor size (OR: 6.572; 95% CI 2.101-20.555, P=0.001), lateral lymph node metastasis (OR: 9.099; 95% CI 1.408-58.819, P=0.020) and microlobulated or ill-defined margins (OR: 14.092; 95% CI 1.598-124.250, P=0.017) were all independent predictors for the coexistence of BRAFV600E and TERT promoter mutations. Two models were established with the above three independent predictors to predict coexisting mutations in the preoperative and postoperative periods. The AUCs of the preoperative and postoperative prediction models were 0.781 (95% confidence interval, 0.781-0.951) and 0.875 (95% confidence interval, 0.830-0.970), respectively. The calibration curve and decision curves of the two prediction models had good calibration ability and good clinical practicability. Conclusion: The established prediction model using ultrasound and clinicopathological characteristics can predict coexisting mutations before or after surgery, stratify prognostic risks and guide the choice of treatment.