Predicting central lymph node metastasis in papillary thyroid carcinoma combined with Hashimoto's thyroiditis: a preoperative study

BackgroundCurrent studies have reported a high association between the Hashimoto's thyroiditis (HT) and papillary thyroid cancer (PTC). However, studies on the characteristics of central lymph node metastasis (CLNM) in PTC patients with HT are scarce. Therefore, this study aims to evaluate the risk factors of CLNM in PTC and HT patients. MethodsWe retrospectively collected clinical data from 933 PTC patients with HT who underwent thyroid surgery. Of these, 653 patients were categorized into a training cohort for constructing the nomogram, and 280 patients formed a validation cohort to verify the performance of the model. Least absolute shrinkage and selection operator (LASSO) regression and multivariate logistic regression analyses were used to select risk factors. A nomogram model for predicting CLNM was developed and internally validated. We subsequently evaluated thyroid function within 3 years after surgery and estimated the prevalence and incidence of postoperative complications between the CLNM ( +) and CLNM (-) groups. ResultsLASSO regression revealed that 19 nonzero variables were associated with CLNM. Multivariate logistic regression analysis revealed that younger, patients of low body mass index (BMI), drinking, intranodular hyperechoic (IH) status, diameter >= 1 cm, multifocality, extrathyroidal extension (ETE), enlarged central lymph nodes (ECLNs) and lateral lymph node metastasis (LLNM) were at higher risk of CLNM (P < 0.05). A nomogram to predict CLNM in PTC patients with HT was constructed and internally validated on the basis of risk factors. The areas under the ROC curve (AUCs) of nomogram were 0.768 (95% CI, 0.723-0.812) and 0.773 (95% CI, 0.705-0.841) in training and validation groups, respectively. Moreover, the nomogram data showed a good discrimination and calibration ability to training and validation groups. Postoperative follow-up revealed that TGAb levels and the incidence of hypothyroidism were significantly greater in CLNM ( +) group than CLNM (-) group, respectively. ConclusionsOur nomogram schedule developed and validated with a comprehensive set of preoperative risk factors showed a high benefit in predicting CLNM in PTC patients with HT. Postoperative follow-up of thyroid function allow to clarify the trend, as well as prevalence and incidence of common thyroid complications.