Neoadjuvant treatment response in negative nodes is an important prognosticator after esophagectomy.

Background: The current American Joint Committee on Cancer Seventh Edition (AJCC7) pathologic staging for esophageal adenocarcinoma (EAC) is derived from data assessing the outcomes of patients having undergone esophagectomy without neoadjuvant treatment and has unclear significance in patients who have received multimodality therapy. Lymph nodes with evidence of neoadjuvant treatment effect without residual cancer cells may be observed and are not traditionally considered in pathologic reports, but may have prognostic significance.

Methods: All patients who underwent esophagectomy after completing neoadjuvant therapy for EAC at our institution between 2006 and 2012 were reviewed. Slides of pathologic specimens were reexamined for locoregional treatment-response nodes lacking viable cancer cells but with evidence of acellular mucin pools, central fibrosis, necrosis, or calcifications suggesting prior tumor involvement. Kaplan-Meier survival functions were estimated, and Cox proportional hazards regression models were used to compare staging models.

Results: Ninety patients (82 men) underwent esophagectomy after neoadjuvant therapy for EAC (mean age, 61.8 ± 8.9 years). All patients received preoperative chemotherapy, and 50 patients also underwent preoperative radiotherapy. Median Kaplan-Meier survival was 55.6 months, and 5-year survival was 35% (95% confidence interval, 19% to 62%). A total of 100 treatment-response nodes were found in 38 patients. For patients with limited nodal disease (62 ypN0-N1), the presence of treatment-response nodes was associated with significantly worse survival (p = 0.03) compared with patients lacking such nodes. Adjusting for patient age and AJCC7 pathologic stage showed the presence of treatment-response nodes significantly increased the risk of death (hazard ratio, 2.7; 95% confidence interval, 1.1 to 6.9; p = 0.04). When stage-adjusted survival was modeled, counting treatment-response nodes as positive nodes offered a better model fit than ignoring them.

Conclusions: Treatment-response lymph nodes detected from esophagectomy specimens in patients having undergone neoadjuvant chemotherapy or combined chemoradiation for EAC provide valuable prognostic information, particularly in patients with limited nodal disease. The current practice of considering lymph nodes lacking viable cancer cells, but with evidence of tumor necrosis, as pathologically negative likely results in understaging. Future efforts at revising the staging system for EAC should consider incorporating treatment-response lymph nodes in the analysis.