Background: The aim of this study was to analyze the time-varying pattern of recurrence risk of early-stage (T1a-T2bN0M0) non-small cell lung cancer (NSCLC) after surgery using the hazard function and identify patients who might benefit from adjuvant chemotherapy.
Patients and methods: This retrospective study enrolled 994 patients with early-stage NSCLC who underwent radical surgical resection between January 1999 and October 2009. Survival curves were generated using the Kaplan-Meier method, and the annual recurrence hazard was estimated using the hazard function.
Results: The median recurrence-free survival (RFS) was 8.8 years. The life table survival analysis showed that the 1-year, 3-year, 5-year and 10-year recurrence rates were 82.0%, 67.0%, 59.0% and 48.0%, respectively. Approximately 256 (25.7%) patients experienced relapse [locoregional: 32 (3.2%) and distant: 224 (22.5%)], and 162 patients died from cancer. The annual recurrence hazard curve for the entire population showed that the first major recurrence surge reached a maximum 1.6 years after surgery. The curve subsequently declined until reaching a nadir at 7.2 years. A second peak occurred at 8.8 years. An analysis of clinical-pathological factors demonstrated that this double-peaked pattern was present in several subgroups.
Conclusions: The presence of a double-peaked pattern indicates that there is a predictable temporal distribution of the recurrence hazard of early-stage NSCLC. The annual recurrence hazard may be an effective method of selecting patients at high risk of recurrence, who may benefit from adjuvant therapy.