Matched-pair and propensity score comparisons of outcomes of patients with clinical stage I non-small cell lung cancer treated with resection or stereotactic radiosurgery

BACKGROUND Stereotactic body radiotherapy (SBRT) is an alternative to surgery for clinical stage I non-small cell lung cancer (NSCLC), but comparing its effectiveness is difficult because of differences in patient selection and staging. METHODS Two databases were combined which contained patients treated from 1999 to 2008 by lobectomy (LR, n=132), sublobar resection (SLR, n=48), and SBRT (n=137) after negative staging. Univariate and multivariate analysis were performed for survival (OS), total recurrence control (TRC comprises local-regional and distant control), and locoregional control (LRC) in our entire population. A matched-pair analysis was also performed that compared surgery and SBRT results. Median follow-up for the entire study population was 25.8 months. RESULTS On univariate analysis, OS was significantly worse with SBRT and also correlated with histology, the Charlson comorbidity index, tumor size, and aspirin use; TRC correlated only with histology; and no variable significantly correlated with LRC. OS was significantly poorer for SBRT in the matched-pair analysis than for patients treated with surgery, but TRC and LRC were not significantly different between these groups. Multivariate analyses including propensity score as a covariate (controlling for all factors affecting treatment selection) found that OS correlated only with Charlson comorbidity index, and TRC correlated only with tumor grade. LRC correlated only with tumor size with or without propensity score correction. CONCLUSIONS This retrospective study has demonstrated similar OS, LRC, and TRC with SBRT or surgery after controlling for prognostic and patient selection factors. Randomized clinical trials are needed to better compare the effectiveness of these treatments. Cancer 2013;119:2683-2691. (c) 2013 American Cancer Society.