- 2026-06
- Lung cancer (Amsterdam, Netherlands) 216
- Lizheng Zhao
- Feng Li
- Zheqing Zhu
- Jinwei Zhao
- Pengjuan Zhang
- Yong Ma
- Fangfang Shen
Study Design
- Type
- Meta-Analysis
- Sample size
- n = 259
- Population
- an overall population of 4,259 patients with non-small cell lung cancer (NSCLC)
- Methods
- We searched major databases through July 2025. IPD were reconstructed from published Kaplan-Meier curves using the Guyot algorithm. To account for immortal time bias, a 6-week landmark analysis was performed. Pooled hazard ratios for OS and PFS were estimated using random-effects models, stratified by treatment regimen (ICI monotherapy vs. chemo-immunotherapy/mixed).
Background
Previous meta-analyses associate cutaneous immune-related adverse events (cirAEs) with improved survival in non-small cell lung cancer (NSCLC). However, quantifying their actual prognostic benefit is complicated by immortal time bias inherent in aggregate data. Furthermore, the prognostic utility of cirAEs across distinct treatment regimens-specifically anti-PD-1/PD-L1 monotherapy versus chemo-immunotherapy-has not been rigorously evaluated using time-adjusted methods. We applied individual patient data (IPD) reconstruction and landmark analyses to assess the regimen-specific prognostic impact of early-onset cirAEs.Methods
We searched major databases through July 2025 (PROSPERO: CRD420251107392). IPD were reconstructed from published Kaplan-Meier curves using the Guyot algorithm. To account for immortal time bias, a 6-week landmark analysis was performed. Pooled hazard ratios (HRs) for overall survival (OS) and progression-free survival (PFS) were estimated using random-effects models, stratified by treatment regimen (ICI monotherapy vs. chemo-immunotherapy/mixed).Results
Fourteen studies comprising an overall population of 4,259 patients were included. The development of cirAEs was associated with prolonged PFS (HR, 0.40; 95% CI, 0.33-0.50) and OS (HR, 0.43; 95% CI, 0.35-0.53). In reconstructed IPD analyses, cirAEs were associated with prolonged survival versus no irAEs, and with longer PFS compared with any irAEs (HR, 0.70; 95% CI, 0.55-0.89). However, in the 6-week landmark IPD analysis, the prognostic association of early-onset cirAEs differed significantly by treatment regimen. For patients receiving ICI monotherapy, cirAEs were associated with improved OS (HR, 0.45; 95% CI, 0.35-0.58) and PFS (HR, 0.54; 95% CI, 0.44-0.65). In contrast, this survival advantage was not observed in chemo-immunotherapy or mixed cohorts (OS: HR, 1.08; 95% CI, 0.69-1.70; PFS: HR, 0.99; 95% CI, 0.62-1.56).Conclusions
This meta-analysis demonstrates that cirAEs, including early-onset events, are associated with robust survival benefits in patients with NSCLC receiving PD-1/PD-L1 inhibitors. CirAEs provide refined prognostic discrimination compared with any irAEs, supporting their utility as a distinct biomarker to guide treatment decisions.