Outcomes following surgical treatment compared to radiation for stage I NSCLC: A SEER database analysis

Introduction: Outcomes following surgery are better than following radiation therapy (RT), for stage I NSCLC. Whether this is due to selection of healthier patients for surgery is unclear. This study was undertaken to compare outcomes between surgical patients and patients who were surgical candidates but did not receive surgery. Methods: Data of patients with stage I NSCLC between 1988 and 2007, included in the SEER database were analyzed. Overall survival (OS) was examined by treatment type (surgery only, radiation only, surgery and radiation, and no treatment). OS was compared between RT patients who refused surgery and those not fit for surgery. Cox proportional hazards model was used to compare outcomes by treatment type. Results: Data from 8579 patients with stage I NSCLC during 1988-2007 were analyzed. Use of RT alone increased during the study period. An increasing proportion of patients with stage I lung cancer chose to have no treatment. On multivariate analysis, OS was better among patients who had surgery. There was a 56% improvement in survival among patients who had surgery compared to fit patients who refused surgery (HR 0.437, 95% CI 0.301-0.632). Patients who refused surgery had a better OS than those who were not fit for surgery (log-rank p= 0.01). Patients who received RT alone or no treatment had a significant improvement in five-year OS during the latter part of the study period (1998-2002 vs. 1988-1992). Conclusions: In medically fit patients, outcomes following surgery are better than those following conventional radiation. Hence surgery should be chosen over conventional radiation, whenever possible. Outcomes following RT show an improvement over time reflecting improvement in radiation techniques.