Purpose: The aim of this cohort study was to assess the benefit that patients with lower esophageal or gastroesophageal junction (E/GEJ) adenocarcinoma receive by continuing perioperative chemotherapy post-surgery.
Methods: Three hundred twelve patients underwent radical tumor surgical resection after preoperative chemotherapy. Chemotherapy was mainly ECX (epirubicin, cisplatin, capecitabine). Propensity score matching (PSM) was used to compare continuation of chemotherapy post-surgery vs. no postoperative treatment.
Results: Two hundred ten patients (67.3%) had GEJ and 102 (32.7%) lower esophageal adenocarcinoma. Microscopically clear surgical margins (R0), according to the Royal College of Pathologists, were achieved in 208 patients (66.7%). In total, 225 patients (72.1%) continued perioperative chemotherapy post-surgery. PSM was used to create two patient groups, well-balanced for basic epidemiological, clinical, and histopathological characteristics. The first included 148 patients who continued perioperative chemotherapy after surgery and the second 86, who did not receive postoperative treatment. The first group had non-significantly different median time-to-relapse (TTR 22.2 vs. 25.7 months, p = 0.627), overall survival (OS 46.1 vs. 36.7 months, p = 0.199), and post-relapse survival (15.3 vs. 8.7 months, p = 0.122). Subgroup analysis showed that only patients with microscopically residual disease after surgery (R1 resection) benefited from continuation of chemotherapy post-surgery for both TTR (hazard ratio [HR] 0.556, 95% CI 0.330-0.936, p = 0.027) and OS (HR 0.530, 95% CI 0.313-0.898, p = 0.018).
Conclusions: Continuation of perioperative chemotherapy post-surgery was not associated with improved outcome in patients with E/GEJ adenocarcinoma. Patients with microscopically residual disease post-surgery might receive a potential benefit from adjuvant chemotherapy.
Keywords: Adjuvant chemotherapy; Esophageal–gastroesophageal junction adenocarcinoma; Perioperative chemotherapy; Prognosis; Propensity score matching analysis.