Stereotactic Body Radiotherapy Versus Surgery for Patients with Stage I Non-Small-Cell Lung Cancer: Comparison of Long-Term Outcome with a Propensity Score Matching Analysis

Abstract

Stereotactic body radiotherapy (SBRT) for clinical stage I non-small-cell lung cancer (NSCLC) has shown promising results, but no randomized study compared SBRT and surgery with a large sample size and enough follow-up time. We aimed to perform a propensity score matching (PSM) analysis to avoid selection bias between the two treatment modalities and compare survival outcomes between SBRT and surgery in patients with stage I NSCLC. This retrospective study included 768 patients who underwent SBRT (n = 190) or surgery (n = 578) including lobectomy, sublobar resection, and wedge resection from 2004 to 2014 at our single institution. Progression-free survival (PFS) and overall survival (OS) were calculated using the Kaplan-Meier method. A multivariate analysis was performed for predictors of PFS and OS. PSM was performed for the two groups based on age, gender, performance status (PS), tumor diameter, forced expiratory volume in 1 second (FEV1%), and Charlson comorbidity index (CCI). The median follow-up times after SBRT and surgery were 66 and 64 months, respectively. PSM identified 128 patients from each treatment group with similar characteristics. Although 5-year PFS rates were significantly lower in the SBRT group than in the surgery group before PSM (58% vs 78%, p < 0.001), there was no significant difference between the two groups after PSM (62% vs 74%, p = 0.10). Five-year OS rates were lower in the SBRT group than in the surgery group before PSM (61% vs 85%, p < 0.001) and after PSM (63% vs 78%, p = 0.008). After PSM, the results of multivariate analyses showed that male sex (hazard ratio [HR] 1.9, 95% confidence interval [CI] 1.1-3.2, p = 0.021), histology other than adenocarcinoma (HR 1.7, CI 1.3-2.3, p < 0.001), and SBRT (HR 1.7, CI 1.0-2.5, p = 0.042) were significant predictors for worse PFS. After PSM, multivariate analyses showed that male sex (HR 2.2, CI 1.3-3.8, p = 0.006), PS (HR 1.4, CI 1.0-1.9, p = 0.032), tumor diameter (HR 1.0, CI 1.0-1.1, p = 0.009), histology other than adenocarcinoma (HR 1.6, CI 1.2-2.1, p = 0.004), and SBRT (HR 2.0, CI 1.2-3.1, p = 0.005) were significant predictors for worse OS. Patients with stage I NSCLC treated with SBRT seem to have a slightly higher disease recurrence and a worse prognosis than those treated with surgery. However, patients treated with SBRT might have no true counterpart in the surgery cohort even if patient cohorts were balanced between the two treatment groups. Prospective large studies are needed to confirm the role of SBRT in patients with clinical stage I NSCLC.