Management of pathologic node-positive disease following initial surgery for clinical T1-2 N0 esophageal cancer: patterns of care and outcomes from the national cancer data base

Purpose: Although clinical T1-2N0 esophageal cancer (EC) is often initially surgically resected (without neoadjuvant therapy), several studies have illustrated substantial rates of discovering pathologically node-positive disease. This study evaluated national practice patterns of adjuvant therapy for this population. Methods: The National Cancer Database (NCDB) was queried (2004–2013) for patients with cT1-2N0M0 EC that received up-front surgery (esophagectomy/local techniques) with subsequent discovery of nodal metastasis. Patients receiving any neoadjuvant therapy were excluded. Multivariable logistic regression determined factors predictive of receiving adjuvant therapy. Kaplan–Meier analysis evaluated overall survival (OS), and Cox proportional hazards modeling determined variables associated with OS. Propensity score matching assessed groups in a balanced manner while reducing indication biases. Results: Altogether, 715 patients met inclusion criteria; 114 (16%) underwent adjuvant chemotherapy, 183 (26%) chemoradiation, 16 (2%) radiotherapy alone, and 402 (56%) observation. Observation was more likely performed with advanced age (p =.002) and at nonacademic centers (p =.001). Median OS in the respective cohorts were 42.6, 35.1, 22.2, and 27.0 months. Both chemotherapy and chemoradiation were statistically similar (p =.462) but superior to observation (p <.05 for both). There was a survival benefit to any adjuvant treatment (median OS 38.5 vs. 27.0 months, p <.001), which persisted after propensity matching (median OS 35.1 vs. 24.3 months, p <.001). On multivariable analysis, any adjuvant treatment was independently associated with improved OS, along with treatment at an academic center (p <.05 for all). Conclusions: In the largest study to date evaluating patterns of care for pN + disease following resection of cT1-2N0 EC, a strikingly high proportion of patients were observed. Adjuvant treatment, ideally chemotherapy or chemoradiation, independently correlated with higher survival, and should be considered in able patients. Treatment at academic facilities also associated with higher survival, which has implications for patient counseling.