Stereotactic body radiation therapy (SBRT) has become the treatment of choice for early-stage non–small-cell lung cancers (NSCLCs) in nonoperative candidates. The purpose of this study was to examine the efficacy and toxicity of SBRT for stage T2 NSCLC.
Eileen Harder, BS, Henry S. Park, MD, MPH, Brandon R. Mancini, MD, Roy Decker, MD, PhD; Yale University
Introduction and Purpose: Stereotactic body radiation therapy (SBRT) has become the treatment of choice for early-stage non–small-cell lung cancers (NSCLCs) in nonoperative candidates. Large prospective and retrospective series have established excellent local control (LC) and minimal toxicity in peripheral tumors smaller than 3 cm in size (stage T1). The outcomes for T2 tumors are less well established. The purpose of this study was to examine the efficacy and toxicity of SBRT for stage T2 NSCLC.
Materials and Methods: Patients treated with SBRT for T1 or T2 NSCLC were identified in a prospectively maintained institutional database that includes outcomes, demographic, clinicopathologic, and radiation treatment information. Toxicity was scored using the National Institutes of Health Common Terminology Criteria for Adverse Events version 4.0 (CTCAE v4.0). An acute toxicity was defined as a treatment-related side effect within 90 days of the first fraction; a late toxicity occurred after this time.
The Kaplan-Meier method was used to estimate local control (LC), distal control (DC), disease-free survival (DFS), and overall survival (OS). Subgroups were compared with the log-rank test. Univariate and multivariate analyses were performed with SPSS.
Results: A total of 244 consecutive patients who received SBRT between October 2007 and October 2013 were included-192 had T1N0M0 lesions and 52 had T2N0M0 disease; median tumor diameter was 2.0 cm (range: 0.54–6.4 cm). The median follow-up was 25.2 months (range: 0.9–68.5 mo). Median dose was 54 Gy with a range of 25–60 Gy (in 2–5 fractions); median biologically effective dose (BED) was 151.2 Gy (range: 56–180 Gy) using the linear-quadratic method with α/β = 10. The 2-year OS for T1 vs T2 disease was 59% vs 40% (median OS 31.2 mo vs 17.6 mo; P = .016). Both univariate and multivariate Cox regression analyses demonstrated a significantly increased risk of death for patients with T2 vs T1 disease (hazard ratio [HR] = 1.65; P = .018). Replacing T-stage with size thresholds of 2 cm, 3 cm, and 4 cm yielded similar results. T-stage was not significantly associated with 2-year LC (90% vs 77%; P = .42), DC (84% vs 77%; P = .47), or DFS (62% vs 52%; P = .50). T-stage was not significantly associated with 3-month freedom from acute toxicity (73% vs 79%; P = .47) or 2-year freedom from late toxicity (77% vs 78%; P = .52) on univariate or multivariate analyses. Of note, one acute grade 4 toxicity (pneumonitis) occurred in a patient after synchronous treatment of T1 lesions, and two late grade 4 toxicities (aspergilloma, pneumonitis) occurred in another two T1 patients. One late grade 5 toxicity (hemorrhage) occurred after synchronous treatment of T2 lesions.
Conclusions: SBRT for T2-stage NSCLC is associated with worse OS compared with T1 disease. There was a nonsignificant trend toward worse LC, DC, and DFS in T2 lesions, although these findings were not significant. Overall, SBRT for T2 NSCLC was accomplished with acceptable toxicity, which was not significantly worse than that of T1 tumors.