Is sleeve lobectomy comparable in terms of short- and long-term results with pneumonectomy after induction therapy? A multicenter analysis.

Background: Sleeve lobectomy (SL) is considered a valid therapeutic option in untreated, centrally located non-small cell lung cancer (NSCLC) even in patients "fit" for pneumonectomy (PN). Nevertheless, SL feasibility and long-term results after induction therapy (IT) have been only rarely investigated. We herein report the results of a multicenter retrospective study on NSCLC patients who underwent PN or SL after IT for locally advanced NSCLC.

Methods: From January 1992 to January 2012, 119 consecutive patients (94 males, 25 females) underwent in three tertiary referral centers either SL (bronchial, arterial, or both) or PN for locally advanced NSCLC after IT (chemotherapy alone or combined chemoradiotherapy). The indication for SL was based on technical feasibility. Clinical and pathologic variables were retrospectively reviewed, and treatment results were assessed and compared in both groups. Survival was calculated by Kaplan-Meier method and compared by the log-rank test as well the Cox regression model.

Results: Sleeve lobectomy was performed in 51 patients and PN, in 68 patients. Thirty-day mortality and morbidity rates were 3.9% and 9.8% for SL and 2.9% and 22.1% for PN, respectively. Five-year survival rates were 53.8% after SL and 43.1% after PN, respectively (p = 0.28). Overall recurrence rate was 42.8% after SL and 47.0% after PN (p = 0.34); relapse was locoregional in 22.4% of SL cases and 12.1% after PN, respectively (p = 0.011). The Cox analysis suggested pN status and right side as independent risk factors for death in the SL group (hazard ratio, 1.96; 95% confidence interval, 1.12 to 3.44; p = 0.018; and hazard ratio, 2.96; 95% confidence interval, 1.13 to 8.66; p = 0.047, respectively). As well, pN status and right side were a strong predictor of relapse (hazard ratio, 2.33; 95% confidence interval, 1.17 to 4.64; p = 0.016; and hazard ratio, 2.96; 95% confidence interval, 1.13 to 8.66; p = 0.046, respectively) in SL patients.

Conclusions: For locally advanced NSCLC, SL represents a safe and effective surgical option when compared with PN even after IT, with substantially comparable early and long-term results. Nevertheless, further investigations on a large cohort of patients are needed.