Leveraging Landmark Analysis for Tailored Surveillance in Stage I Non-Small-Cell Lung Cancer

Background: Current guidelines for NSCLC follow-up lack specific recommendations on surveillance duration. This study aims to analyze survival and surveillance data in resected stage I NSCLC. Methods: We retrospectively reviewed 759 pathological stage I NSCLC (9thTNM ed.) patients with no history of lung cancer (LC) undergoing surgery from January 2003 to December 2018. Overall survival (OS), incidence of relapse (IR), and incidence of new primary LC (NP) were analyzed. Long-term effect of follow-up beyond 5 years was assessed by landmark analysis of OS, IR, and NP at 10 years, restricted to individuals alive without relapse or NP at 5 years (5-year event-free survivors, 5y-EFSs). Results: The rates of 10-year OS, 10-year IR, NP incidence, and 5y-EFSs were, respectively, 75%, 18%, 1.1%/year, and 59.1% (449 patients). Carcinoid IA/IB (0-10%) and adenocarcinoma IA/IB without lung nodules (LNs) (8-12%) had a similarly lower risk of relapse (p = 0.5088) compared to adenocarcinoma with LNs (p = 0.0191). Similarly, carcinoid (0-0.2%/year) and adenocarcinoma without LNs (0-0.3%/year) had the same lower incidence of NP (p = 0.8062) compared to patients with LNs (p < 0.0001). The group of 5y-EFSs had a conditional 10-year OS, IR, and NP incidence of 92%, 5%, and 0.8%/year. In 5y-EFSs, 10-year OS was better in carcinoid (100%) and adenocarcinoma (94%, p = 0.0009) patients; 10-year IR was lower in stage IA (4%) vs. IB (10%, p = 0.0444), and NP was lower in patients with no pre-surgery (0.5 vs. 1.5%/year, p = 0.0147) and no post-surgery LNs (0.6 vs. 1.1%/year, p = 0.0202). Conclusions: Based on our results, we propose a tailored surveillance strategy by de-escalating follow-up for low-risk patients while maintaining intensive monitoring for high-risk individuals.