Stereotactic Ablative Radiation Therapy Versus Surgery in Early Lung Cancer: A Meta-analysis of Propensity Score Studies

PURPOSE: As no completed randomized trials of surgery versus stereotactic ablative radiation therapy (SABR) in patients with early-stage non-small cell lung cancer are available, numerous propensity score studies have attempted to mimic the setting of clinical trials using nonrandomized data. We performed a meta-analysis of propensity score studies comparing SABR and surgery.

METHODS AND MATERIALS: The MEDLINE and Embase databases were queried up to December 2016. Two authors independently reviewed the records for inclusion and extracted outcome measures. The study was conducted according to PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) and MOOSE (Meta-analysis of Observational Studies in Epidemiology) guidelines. The primary meta-analysis and secondary analyses were carried out using R (version 3.3.2) at a significance level of .05.

RESULTS: Sixteen studies were included in the meta-analysis. Overall survival favored surgery (hazard ratio for SABR vs surgery, 1.48 [95% confidence interval, 1.26-1.72]; I2 = 80.5%). Lung cancer-specific survival was not significantly different between SABR and surgery (hazard ratio, 1.17 [95% confidence interval, 0.92-1.50]; I2 = 18.6%). On stratification, overall survival favored both lobectomy and sublobar resection over SABR, although lung cancer-specific survival was again not significantly different. On secondary analysis, the lymph node upstaging rate was 15.6% following surgery, with 11.4% of patients receiving chemotherapy. The propensity score-matching caliper distance and first-author specialty were found to be associated with survival endpoints on regression.

CONCLUSIONS: For patients with early-stage non-small cell lung cancer who are eligible for either treatment, better overall survival was seen after surgery compared with SABR. However, lung cancer-specific survival was similar for both treatments. Prospective clinical trials are preferred to propensity analyses in evaluating the nature of non-cancer-related death after SABR.