Thoracoscopic lobectomy is associated with improved short-term and equivalent oncological outcomes compared with open lobectomy for clinical Stage I non-small-cell lung cancer: A propensity-matched analysis of 963 cases

Abstract

OBJECTIVES: Previous literature has reported lower morbidity for video-assisted thoracoscopic surgery lobectomy (VL) compared with open lobectomy (OL); however, most comparative studies have been retrospective and have failed to compare well-matched patient groups, therefore allowing selection bias to influence results. Furthermore, oncological adequacy of VL has recently been questioned, particularly with respect to lymphadenectomy. This study aimed to evaluate short- and long-term outcomes of a large cohort of consecutive patients with c-stage I non-small-cell lung cancer (NSCLC) that underwent either VL or OL. METHODS: Consecutive patients with c-stage I NSCLC who underwent lobectomy without preoperative therapy were reviewed. Univariable, multivariable and propensity-matched analyses were performed. VL patients who underwent conversion to OL were analysed within the VL group. RESULTS: VL was performed in 307 (32%) patients and OL in 656 (68%). Twenty-two (7%) patients converted from VL to OL. Although there were no differences in overall p-stage grouping, there were fewer patients with pT2 tumours in the VL group (39 vs 48%, P = 0.012) and fewer patients with squamous cell histology (26 vs 18%, P = 0.006). These differences resolved after propensity matching. In unmatched and matched analyses, VL was associated with less overall morbidity, less pulmonary morbidity, fewer atrial arrhythmias, shorter chest tube duration and shorter hospital stay than patients who had OL. Thirty-day in-hospital mortality was 0.3 and 1.4%, for VL and OL groups, respectively (P = NS). In unmatched analysis (log rank), 5-year survival favoured VL (78 vs 68%, P = 0.007); however, after propensity matching there was only a trend towards improved survival with VL (78 vs 73%, P = 0.071). Multivariable Cox regression analysis revealed VL (hazard ratio (HR) 0.64, 95% confidence interval (CI) 0.46-0.92), male sex (HR 1.43, 95% CI 1.10-1.86), Zubrod performance status (HR 3.42, 95% CI 1.26-9.29) and increasing age (HR 1.04, 95% CI 1.03-1.06) to be independent predictors of survival. CONCLUSIONS: Patients with clinical Stage I NSCLC undergoing VL have less perioperative morbidity compared with matched OL controls. Regional lymphadenectomy, nodal upstaging, overall and disease-free survival were similar between VL and OL groups. In experienced centres, VL is an acceptable operation for patients with c-stage I NSCLC.