Evaluation of disease-free survival as a predictor of overall survival and assessment of real-world burden of disease recurrence in resected early-stage non–small cell lung cancer

BACKGROUND: Intermediate endpoints, such as disease-free survival (DFS), have shown good correlation with overall survival (OS) in early-stage non–small cell lung cancer (NSCLC) clinical trials. However, real-world data are limited, and no previous real-world study has quantified the clinical and economic burden of disease recurrence. OBJECTIVE: To examine the association between real-world DFS (rwDFS) and OS and quantify the association between NSCLC recurrence and health care resource utilization (HCRU), health care costs, and OS in patients with resected early-stage NSCLC in the United States. METHODS: Data from the Surveillance, Epidemiology, and End Results–Medicare database (2007-2019) for patients with newly diagnosed stage IB (tumor size ≥ 4 cm) to IIIA (American Joint Committee on Cancer 7th edition) NSCLC who underwent surgery for primary NSCLC were analyzed in this retrospective observational study. Baseline patient demographic and clinical characteristics were described. rwDFS and OS were compared between patients with vs without recurrence using Kaplan-Meier curves and the log-rank test; their correlation was assessed using normal scores rank correlation. All-cause and NSCLC-related HCRU and health care costs were summarized, and mean monthly allcause and NSCLC-related health care costs were compared between cohorts using generalized linear models. RESULTS: Of the 1,761 patients who underwent surgery, 1,182 (67.1%) had disease recurrence; these patients had shorter OS from the index date and shorter subsequent OS at each postsurgery landmark (ie, 1, 3, and 5 years) than those without recurrence (all P < 0.001). OS and rwDFS were significantly correlated (0.57; P < 0.001). Patients with recurrence also had significantly higher all-cause and NSCLC-related HCRU and mean monthly all-cause and NSCLC-related health care costs during the study period. CONCLUSIONS: Postsurgery rwDFS was significantly correlated with OS in patients with early-stage NSCLC. Patients with postsurgery recurrence had a higher risk of death and incurred higher HCRU and health care costs than those without recurrence. These findings highlight the importance of preventing or delaying recurrence in patients with resected NSCLC.