Background: The current American Thyroid Association risk-stratification system for papillary thyroid carcinoma (PTC) incorporates the number and size of positive lymph nodes (LNs) but places less weight on extranodal extension (ENE). This study investigated how to incorporate ENE into the current system to predict recurrence better in PTC N1 patients. Methods: A total of 369 N1 PTC patients without distant metastasis were enrolled. The combination of number of positive LNs and LNs with ENE that had the highest C-index were identified with multivariable Cox proportional hazards models. ENE number was incorporated into the current system considering the recurrence rate and unadjusted and adjusted hazard ratios (HRs) of the subgroups. Kaplan-Meier curves for recurrence based on current and alternative systems were compared by log-rank test. Results: The recurrence rate for the subgroup with five or fewer positive LNs and one to three ENEs (7/61; 11.5%) was higher than that of the subgroup with five or fewer positive LNs without ENE (5/129; 3.9%; adjusted HR = 3.42 [confidence interval (CI) 0.99-11.75]; p = 0.050). In contrast, adjusted HRs of the subgroup with more than five positive LNs and one to three ENEs (2.33 [CI 0.52-10.35]) or with four or more ENEs (3.86 [CI 1.05-14.17]) were not higher than those of the subgroup with more than five LNs without ENE (4.47 [1.16-17.19]). Incorporating ENE into the current system as an intermediate-risk group yielded a lower log-rank p-value (0.05 vs. 0.01) than the current system. Conclusions: The presence of ENE in low volume LN metastasis confers an intermediate risk of recurrence. Incorporating ENE into the current system allows more accurate decisions regarding further management of PTC N1 patients.
Keywords: extranodal extension; papillary thyroid carcinoma; recurrence; risk-stratification system.