Construction and validation of nomograms to reduce completion thyroidectomy by predicting lymph node metastasis in
Rong Wang1, Zimei Tang1, Zhenghao Wu1
1Department of Breast and Thyroid Surgery, Union Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, 430022, China.
Context:
More than 5 central lymph nodes metastases (CLNM) or lateral lymph node metastasis (LLNM) indicates a higher risk of recurrence in low-risk papillary thyroid carcinoma (PTC) and may lead to completion thyroidectomy (CTx) in patients initially undergoing lobectomy.
Objective:
To screen potentially high-risk patients from low-risk patients by using preoperative and intraoperative clinicopathological features to predict lymph node status.
Methods:
A retrospective analysis of 8301 PTC patients in Wuhan Union Hospital database (2009-2021) was performed according to the 2015 American Thyroid Association (ATA) and 2021 National Comprehensive Cancer Network (NCCN) guidelines, respectively. Logistic regression and best subsets regression were used to identify risk factors. Nomograms were established and externally validated using the Differentiated Thyroid Cancer in China cohort.
Results:
More than 5 CLNM or LLNM was detected in 1648 (19.9%) patients. Two predictive models containing age, gender, maximum tumor size, free thyroxine (FT4) and palpable node (all p < 0.05) were established. The nomogram based on NCCN criteria showed better discriminative power and consistency with a specificity of 0.706 and a sensitivity of 0.725, and external validation indicated that 76% of potentially high-risk patients could achieve preoperative conversion of surgical strategy.
Conclusions:
Models based on large cohorts with good predictive performance were constructed and validated. Preoperative low-risk (T1-2N0M0) patients with age younger than 40 years, male gender, large tumor size, low FT4 and palpable nodes may be at high risk of LLNM or more than 5 CLNM, and they should receive more aggressive initial therapy to reduce CTx.
