A Model Predicting Occult Metastases in Lateral Lymph Nodes in pN1a Stage Papillary Thyroid Cancer.

OBJECTIVE

This study aimed to develop and validate a nomogram for predicting occult lateral neck lymph node metastasis (LLNM) in patients with pN1a papillary thyroid carcinoma (PTC), addressing the clinical controversy surrounding prophylactic lateral neck dissection (PLND).

METHODS

A retrospective analysis was conducted on 128 pN1a PTC patients who underwent total thyroidectomy with bilateral central lymph node dissection and ipsilateral PLND between 2020 and 2023. Clinical and pathological data, including tumor location, size, capsular invasion, and nodal status, were collected. Least Absolute Shrinkage and Selection Operator (LASSO) regression and multivariate logistic regression were employed to identify independent risk factors for lymph node metastasis (LNM). A nomogram was constructed based on these factors and internally validated using bootstrap resampling (B=1000). External validation was performed on an additional 37 patients treated between 2023 and 2024.

RESULTS

Tumor location in the upper pole (odds ratio [OR]: 2.45), size >10 mm (OR: 2.12), and capsular invasion (OR: 1.89) were identified as independent predictors of occult LLNM. The nomogram demonstrated robust discriminative ability, with an area under the curve (AUC) of 0.826 (95% confidence interval [CI]: 0.736-0.916) in internal validation and 0.858 (95% CI: 0.740-0.975) in external validation. Calibration curves indicated excellent agreement between predicted and observed outcomes. Decision curve analysis confirmed the model's clinical utility for threshold probabilities exceeding 25%.

CONCLUSION

The proposed nomogram effectively stratifies the risk of occult LLNM in pN1a PTC patients, providing a valuable tool for individualized surgical planning. By integrating tumor-specific features, this model aids in selecting patients who may benefit from PLND while minimizing overtreatment and associated complications. Further multicenter studies are warranted to enhance its generalizability.