BACKGROUND: The role of lymphadenectomy in adrenocortical carcinoma (ACC) is controversial, and formal lymph node (LN) dissection is not routine. We sought to determine the minimum number of LNs that must be examined to accurately identify a patient as node negative.

METHODS: The National Cancer Database was used to identify patients diagnosed with ACC from 2004 to 2013 who underwent surgical resection. Patients with distant metastases, multivisceral resection, or missing surgical or lymphadenectomy data were excluded. The primary outcome was overall survival (OS).

RESULTS: LNs were identified on histopathology in 156 patients. Of these, 35 (22%) had at least one positive LN. Positive LNs were associated with positive surgical margins (odds ratio [OR] 5.80, p = 0.002) and increasing LN yield (OR 1.06, p = 0.02). Overall, on Cox regression analysis, LN positivity (hazard ratio [HR] 3.02, p < 0.001) and positive surgical margins (HR 2.06, p = 0.048) independently predicted poor OS after controlling for other factors that may influence survival. LN(-) disease in patients with one to three LNs examined had poorer overall survival compared with when at least four LNs were examined (p = 0.02). None of the other patient, tumor, and treatment variables had any impact on OS of the LN(-) cohort. The likelihood of identifying nodal involvement was higher on examination of at least four LNs compared with examination of one to three LNs (30 vs. 16%, p = 0.03).

CONCLUSIONS: LN positivity in ACC tumors independently predicts worse 5-year OS and a minimum of four LNs may be required to accurately determine LN negativity.