# Multifactorial risk stratification for central lymph node metastasis in papillary thyroid carcinoma: a predictive model integrating clinical and tumor characteristics

## TL;DR

This study creates a model to predict central lymph node metastasis in thyroid cancer patients using clinical and tumor features.

## Contribution

A novel predictive model for central lymph node metastasis in papillary thyroid carcinoma integrating clinical and tumor characteristics.

## Key findings

- Key risk factors include young age, high TRAb levels, calcification, and tumor size.
- The model achieved an AUC of 0.76, showing robust predictive accuracy.
- Clinical decision analysis suggests the model can improve treatment decisions for most patients.

## Abstract

With the increasing prevalence of papillary thyroid carcinoma PTC) and advancements in auxiliary examination technology, the holistic detection rate of malignant thyroid nodules, particularly small ones, continues to rise. However, there remains controversy surrounding the optimal treatment for PTC, and a crucial factor influencing treatment decisions is the status of central lymph node metastasis (CLNM). There is a lack of research on the relationship between clinical laboratory results and tumor characteristics observed during surgery and CLNM status. Therefore, our research aims to systematically explore the risk factor of CLNM in patients with PTC.

We systematically gathered and analyzed clinical features and pathological data of 2,435 PTC patients who underwent surgery. After variable screening, the selected variables were included in logistic regression analysis, and a Nomogram prediction model was constructed according to the analysis results. To investigate the risk factors for CLNM in patients with PTC.

This study included a total of 2,435 patients diagnosed with PTC, among whom 933 patients were confirmed as CLNM by postoperative pathology. Univariate and multivariate regression analysis identified age, serum TRAb levels, calcification, multifocality, extrathyroidal invasion, tumor size, and tumor location as risk factors associated with CLNM. The prediction model based on these risk factors demonstrated robust accuracy with an AUC of 0.76. Clinical decision curve analysis indicated that aside from a small range of low threshold probabilities, intervening based on the model’s predictions can yield greater clinical benefit.

Key risk factors for CLNM in PTC patients include young age, high serum thyrotropin receptor antibody (TR-Ab) levels, calcification, multifocality, extrathyroidal extension, larger tumor size, and tumor location in the middle or lower poles of the thyroid. The clinical prediction model established based on these critical risk factors can provide a more accurate reference standard for clinical decision-making in practice.