Respiratory complications after oesophagectomy for cancer do not affect disease-free survival

Recent studies have suggested that postoperative complications could have a potential negative effect on long-term outcome after oesophagectomy for cancer. Because respiratory failures represent the most frequent postoperative complication, we have investigated the prognostic impact of these complications on disease-free survival (DFS). From a prospective single-institution database of 405 consecutive patients who underwent transthoracic oesophagectomy for cancer, we retrospectively analysed medical charts of all patients with microscopically complete resection (R0, n = 384 patients). Complications were graded according to the modified Clavien classification. Respiratory complications were defined as atelectasis, pneumonia or acute respiratory distress syndrome in the absence of early surgical complications. Patients with grade 5 (postoperative mortality, n = 43, 11%) were excluded from the analysis. The remaining 341 patients were analysed for estimation of DFS according to the Kaplan-Meier method. Logistic regression analysis was conducted to discriminate predictive factors affecting DFS. According to the modified Clavien classification, postoperative complications rates were grade 0: 147 (44%), grade 1: 7 (2%), grade 2: 56 (16%), grade 3: 69 (20%) and grade 4: 62 (18%). Five-year DFS rates were not significantly different between grade 0 (no complication, 38%, n = 147) and other grades (grade 1, 2, 3 and 4 (64, 45, 56 and 48%, respectively)). Respiratory complications occurred in 107 patients (31%) and the 5-year DFS in this subgroup was 47% compared with 38% observed in grade 0 patients (P = 0.75). Clavien classification and respiratory complications did not come out in the univariate analysis of factors affecting DFS. On logistic regression, only two variables affected DFS: c-N stage and extracapular lymph node involvement. When postoperative mortality is excluded, postoperative complications do not affect DFS in patients with complete resection. This deserves substantial information regarding the prognosis of subgroup of patients in critical situations where incrementing intensive care is debated.