BACKGROUND: Recent clinical trials have demonstrated the benefit and feasibility of perioperative chemotherapy for treatment of gastroesophageal adenocarcinoma (GEA). Despite convincing results, patients entering such trials usually represent only a fraction of those who are candidates for treatment. Confirmation of trial-reported effects and tolerability in unselected cohorts is therefore required. The aims of this study were to confirm the safety and efficacy of perioperative chemotherapy for resectable GEA and to delineate risks of treatment failure. METHODS: We conducted a national retrospective cohort analysis of patients admitted for perioperative chemotherapy for resectable GEA. Regimens were epirubicin and capecitabine combined with oxaliplatin or cisplatin. RESULTS: The intention-to-treat analysis included 271 patients. Eighty-seven percent of patients completed preoperative chemotherapy, and 63 % received radical resection. Age >70 years (odds ratio 2.58) and hypoalbuminemia (odds ratio 4.10) were independent predictors of not undergoing scheduled surgery (P = 0.033). Grade 3 or higher febrile neutropenia, fatigue, and diarrhea were common in the oxaliplatin group (n = 128), but hypomagnesaemia and tinnitus/hearing loss were more common in the cisplatin group (n = 135). The median overall survival was 26.4 months, and the 1- and 2-year survival rates were 76 and 53 %, respectively. Performance status >0 (hazard ratio 1.64) and elevated serum lactate dehydrogenase (hazard ratio 3.03) were independent predictors of poor prognosis (P ≤ 0.05). CONCLUSIONS: Perioperative chemotherapy is feasible and well tolerated in patients with good performance status and low incidence of comorbidities.
Annals of Surgical Oncology, 2015, Vol 22, Issue 5, p. 1540-1547