Background: Conditional survival (CS) is defined as living some additional time predicated on living to a certain time point. This study aimed to evaluate the usefulness of CS analyses for postoperative follow-up of surgically treated patients with lung cancer.
Methods: We retrospectively analyzed survival and clinicopathologic data from 859 patients with non-small cell lung cancer who underwent complete resection. CS is the probability of surviving additional time (y), after an individual has already survived for some time (x), and can be calculated from the following formula: CS(y|x) = S(x + y)/S(x), where S is the overall survival at a specific time.
Results: The 5-year cumulative overall survival rate was 75%. The additional 5-year survival rates at 1, 2, 3, and 4 years after operation (5Y-CS) were 75%, 77%, 77%, and 81%, respectively. The likelihood of surviving for a total of 5 years, 3 years after surgery was 89%. Age less than or equal to 70 years, female sex, no or light smoking, adenocarcinoma histologic type, pathologic stage I, and normal serum carcinoembryonic antigen levels were the clinicopathologic features associated with a high cancer-specific CS. For patients with favorable factors, the 2Y recurrence-free CS (3) reached 100%, whereas 5Y-CS did not change.
Conclusions: Estimates of CS probabilities may provide more meaningful information than traditional cumulative survival. Even in patients with favorable factors, a postoperative follow-up visit after 3 postsurgical years may be required.
Copyright © 2016 The Society of Thoracic Surgeons. Published by Elsevier Inc. All rights reserved.