Sleeve lobectomy may provide better outcomes than pneumonectomy for non-small cell lung cancer. A decade in a nationwide study

International audienceIntroduction: Whenever feasible, sleeve lobectomy is recommended to avoid pneumonectomy for lung cancer, but these guidelines are based on limited retrospective series. The aim of our study was to compare outcomes following sleeve lobectomy and pneumonectomy using data from a national database. Methods: From 2005 to 2014, 941 sleeve lobectomy and 5318 pneumonectomy patients were recorded in the French database Epithor. Propensity score was generated with 15 pretreatment variables and used to create balanced groups with matching (794 matches) and inverse probability of treatment weighting (standardized difference was 0 for matching, and 0.0025 after weighting). Odds ratio (OR) of postoperative complications and mortality and hazard ratio (HR) for overall survival and disease-free survival were calculated using propensity adjustment techniques and a sensitivity analysis. Results: Postoperative mortality after sleeve resection was similar to that after pneumonectomy (matching OR, 1.24; P = .4; weighting OR, 0.77; P = .4) despite significantly lower odds of pulmonary complications with pneumonectomy (matching OR, 0.4; P < .0001; weighting OR, 0.12; P < .001). The adjusted HR for death after pneumonectomy was significantly higher when analyzed using matched analysis but not with weighting (matching HR, 1.63; P = .002; weighting HR, 0.97; P = .92). The same was true for disease-free survival (matching HR, 1.49; P = .01; weighting HR, 1.03; P = .84). Conclusions: Despite early differences in perioperative pulmonary outcomes favoring pneumonectomy, early overall and disease-free survival was in favor of sleeve lobectomy in the matched analysis but not the weighted analysis. In our opinion, when it is technically feasible, sleeve lobectomy should be the preferred technique