UCSF

Cervical lymph node metastases are common in papillary thyroid cancer (PTC). Clinically negative lymph nodes confer uncertainty about true lymph node status, potentially prompting empiric postoperative radioactive iodine (RAI) administration even in low-risk patients. We examined the association of clinically (cN0) versus pathologically negative (pN0) lymph nodes with utilization of RAI for low-risk PTC. Using the National Cancer Database 1998-2011, adults with PTC who underwent total thyroidectomy for Stage I/II tumors 1-4 cm were evaluated for receipt of RAI based on cN0 versus pN0 status. Cut-point analysis was conducted to determine the number of pN0 nodes associated with the greatest decrease in the odds of receipt of RAI. Survival models and multivariate analyses predicting RAI use were conducted separately for all patients and patients <45 years. 64,980 patients met study criteria; 39,778 (61.2 %) were cN0 versus 25,202 (38.8 %) pN0. Patients with pN0 nodes were more likely to have negative surgical margins and multifocal disease (all p < 0.001). The mean negative nodes reported in surgical pathology specimens was 4; ≥5 pathologically negative lymph nodes provided the best cut-point associated with reduced RAI administration (OR 0.91, CI 0.85-0.97). After multivariable adjustment, pN0 patients with ≥5 nodes examined were less likely to receive RAI compared to cN0 patients across all ages (OR 0.89, p < 0.001) and for patients aged <45 years (0R 0.86, p = 0.001). Patients with <5 pN0 nodes did not differ in RAI use compared to cN0 controls. Unadjusted survival was improved for pN0 versus cN0 patients across all ages (p < 0.001), but not for patients <45 years (p = 0.11); adjusted survival for all ages did not differ (p = 0.13). Pathological confirmation of negative lymph nodes in patients with PTC appears to influence the decision to administer postoperative RAI if ≥5 negative lymph nodes are removed. It is possible that fewer excised lymph nodes may be viewed by clinicians as incidentally resected and thus may suboptimally represent the true nodal status of the central neck. Further research is warranted to determine if there is an optimal number of lymph nodes that should be resected to standardize pathological diagnosis.