Five-year survival does not equal cure in non-small cell lung cancer: a Surveillance, Epidemiology, and End Results-based analysis of variables affecting 10- to 18-year survival

Abstract

Objective: Five-year survival after the diagnosis of non-small cell lung cancer is the most common benchmark used to evaluate long-term survival. Data on survival beyond 5 years are sparse. We sought to elucidate variables affecting 10- to 18-year survival.

Methods: A total of 31,206 patients alive at least 5 years after diagnosis of non-small cell lung cancer who were registered in the Surveillance, Epidemiology, and End Results database from 1988 to 2001 were examined. Primary end points were disease-specific survival and overall survival. Survival analysis was performed with Kaplan-Meier estimates, multivariable Cox proportional hazards regression, and competing risk models.

Results: Overall survival at 10, 15, and 18 years was 55.4%, 33.1%, and 24.3%, respectively. Disease-specific survival at 10, 15, and 18 years was 76.6%, 65.4%, and 59.4%, respectively. In multivariable regression analysis, squamous cell cancers had a disease-specific survival advantage (hazard ratio, 0.88; P < .0001) but an overall survival disadvantage (hazard ratio, 1.082; P = .0002) compared with adenocarcinoma. Pneumonectomy (hazard ratio, 0.44) and lobectomy (hazard ratio, 0.474) had improved disease-specific survival compared with no surgery (P < .0001). Left-sided tumors (hazard ratio, 0.723; P = .036) and node-negative cancers (hazard ratio, 0.562; P < .001) also had a better disease-specific survival and, to a lesser extent, overall survival advantage.

Conclusions: Five-year survivors of non-small cell lung cancer have a persistent risk of death from lung cancer up to 18 years from diagnosis. More than one half of all deaths in 5-year survivors are related to lung cancer. In multivariable regression analysis, age, node-negative disease, and lobar or greater resection were strong predictors of long-term survival (ie, 10-18 years).