Long-Term Survival in Patients with Gastroesophageal Junction Cancer Treated with Preoperative Therapy: Do Thoracic and Abdominal Approaches Differ?

Background: The optimal surgical approach for gastroesophageal junction (GEJ) cancer treated with preoperative therapy remains controversial. We compared the outcomes of patients who underwent either esophagectomy or gastrectomy and identified variables associated with overall survival (OS). Methods: We reviewed records of patients with Siewert types II and III GEJ adenocarcinoma who were treated with preoperative therapy followed by resection from 1995 to 2013. OS was assessed using Kaplan–Meier curves and associated variables were analyzed using Cox proportional hazards models. Results: Of 143 patients, 110 (76.9 %) had type II and 33 (23.1 %) had type III tumors. Most (86 %) patients had stage T3 or T4 disease, and more than half had N+ (62 %) disease. The majority (93 %) received neoadjuvant chemoradiation; 7 % received chemotherapy alone. Patients with type II tumors underwent either esophagectomy (75 %) or gastrectomy (25 %). Patients with type III tumors primarily underwent gastrectomy (88 %). Eighty-six (60 %) patients underwent extended (D1+/D2) abdominal lymphadenectomy. We saw no differences between esophagectomy and gastrectomy patients in R0 resection rate (94 vs. 95 %; p = 0.9), number of nodes removed (mean, 18.3 vs. 19.3; p = 0.6), or 60-day mortality rate (4 vs. 4 %; p = 1.0). The median follow-up period for survivors was 65 months. Esophagectomy and gastrectomy showed similar 5-year OS rates (49 vs. 53 %; p = 0.8). Surgical approach was not associated with OS [hazard ratio (HR) 1.30; 95 % confidence interval (CI) 0.68–2.45; p = 0.43]. The strongest predictor of OS was extended lymphadenectomy (HR 0.55; 95 % CI, 0.32–0.94; p = 0.03). Conclusions: R0 resection and OS rates were similar in patients undergoing esophagectomy or gastrectomy after neoadjuvant therapy; however, extended abdominal lymphadenectomy may improve OS rates.