The optimal surgical technique for lobectomy in lung cancer is not well defined. Proponents of video-assisted thoracic surgery (VATS) hypothesize that less trauma leads to quicker recovery, whereas those who advocate thoracotomy claim it as an oncologically superior procedure. However, a well-balanced comparison of the two procedures is lacking in the literature.
All patients who underwent lobectomy for clinical stage 1A lung cancer by computed tomographic and positron emission tomographic scan were identified from a prospective database. Patient characteristics were compared by the Student t test, Pearson chi(2), and Fisher exact test. A propensity score-matched analysis was performed. Survival was assessed by Kaplan-Meier and Cox proportional hazards analysis. Complications were assessed by a multivariate logistic regression model evaluating age, sex, comorbidities, pulmonary function, tumor size, nodal status, surgeon, and histologic characteristics.
From May 2002 to August 2007, 398 patients underwent an attempt at VATS lobectomy and 343 underwent thoracotomy. An "intent-to-treat" analysis was performed. There was 1 postoperative death in each group. Survival by Cox model was no different for VATS versus thoracotomy (hazard ratio 0.72; P = .12), whereas age (hazard ratio 1.03; P < .001), larger tumor size (hazard ratio 1.34; P < .001), and higher nodal stage (hazard ratio 1.92; P < .001) were associated with worse survival. Logistic regression demonstrated fewer complications for VATS lobectomy (odds ratio 0.73; P = .06), whereas age (odds ratio 1.04; P < .001) and tumor size (odds ratio 1.2; P < .020) correlated with a greater number of complications. Patients undergoing VATS lobectomy demonstrated a 2-day shorter length of stay than patients undergoing thoracotomy (P < .001). Propensity score-matched analysis supported these findings.
VATS lobectomy and thoracotomy demonstrated similar 5-year survivals. However, VATS lobectomy was associated with fewer complications and shorter length of hospital stay.