Sleeve Lobectomy for Non-Small Cell Lung Cancers: Predictive CT Features for Resectability and Outcome Analysis.

OBJECTIVE. The purpose of this study was to investigate the CT features predictive of resectability in sleeve lobectomy (SL) for patients with non-small cell lung cancers (NSCLCs) and the outcomes of this procedure. MATERIALS AND METHODS. The study included 45 patients with NSCLCs who planned to undergo SL as a first-line surgical option and underwent SL or pneumonectomy between February 2014 and July 2017. The patients were dichotomized according to whether SL was achieved without residual disease. ROC curve and logistic regression analyses were performed to identify CT features predictive of resectability in SL without residual disease. For outcome analysis, postoperative complications and survival were compared between the two patient groups. RESULTS. SL was successfully performed in 25 patients (55.6%), whereas eight patients underwent SL but had a positive resection margin and 12 required pneumonectomy. The AUC value of the tumor-involved length of the salvage bronchus for SL without residual disease was 0.802, with 5 mm considered the optimal cutoff for this tumor-involved length. A tumor-involved length of 5 mm or less for the salvage bronchus representing the distal anastomosis end in SL was the sole significant predictor for resectability in SL without residual disease in logistic regression analysis (odds ratio, 16.0; 95% CI, 3.7-69.6; p < 0.001). The group of patients who underwent SL without residual disease tended to have a more favorable survival outcome than patients who underwent SL but had residual disease or those who underwent pneumonectomy ( p = 0.110), but postoperative complication rates were similar between the two patient groups (20.0% vs 25.0%, respectively). CONCLUSION. A tumor-involved salvage bronchus length of 5 mm or less was a CT feature predictive of resectability in SL without residual disease, which may improve the survival of patients with NSCLCs.