Search | Search by Center | Search by Source | Keywords in Title
Backhus LM, Farjah F, Liang CK, He H, Varghese TK, Au DH, Flum DR, Zeliadt SB. Imaging surveillance and survival for surgically resected non-small-cell lung cancer. The Journal of surgical research. 2016 Jan 1; 200(1):171-6.
INTRODUCTION: The importance of imaging surveillance after treatment for lung cancer is not well characterized. We examined the association between initial guideline recommended imaging surveillance and survival among early-stage resected non-small-cell lung cancer (NSCLC) patients. METHODS: A retrospective study was conducted using Surveillance, Epidemiology, and End Results-Medicare data (1995-2010). Surgically resected patients, with stage I and II NSCLC, were categorized by imaging received during the initial surveillance period (4-8 mo) after surgery. Primary outcome was overall survival. Secondary treatment interventions were examined as intermediary outcomes. RESULTS: Most (88%) patients had at least one outpatient clinic visit, and 24% received an initial computerized tomography (CT) during the first surveillance period. Five-year survival by initial surveillance imaging was 61% for CT, 58% for chest radiography, and 60% for no imaging. After adjustment, initial CT was not associated with improved overall survival (hazard ratio [HR], 1.04; 95% confidence interval [CI] 0.96-1.14). On subgroup analysis, restricted to patients with demonstrated initial postoperative follow-up, CT was associated with a lower overall risk of death for stage I patients (HR, 0.85; 95% CI, 0.74-0.98), but not for stage II (HR, 1.01; 95% CI, 0.71-1.42). There was no significant difference in rates of secondary interventions predicted by type of initial imaging surveillance. CONCLUSIONS: Initial surveillance CT is not associated with improved overall or lung cancer-specific survival among early-stage NSCLC patients undergoing surgical resection. Stage I patients with early follow-up may represent a subpopulation that benefits from initial surveillance although this may be influenced by healthy patient selection bias.