Real-world outcomes of anti-PD1 antibodies in platinum-refractory, PD-L1-positive recurrent and/or metastatic non-small cell lung cancer, and its potential practical predictors: first report from Korean Cancer Study Group LU19-05.

2021 
Purpose Although immune-checkpoint inhibitors have become a new therapeutic option for recurrent/metastatic non-small cell lung cancers (R/M-NSCLC), its clinical benefit in the real-world is still unclear. Methods We investigated 1181 Korean patients with programmed death-1 ligand 1 (PD-L1)-positive [tumor proportion score (TPS) ≥ 10% by the SP263 assay or ≥ 50% by the 22C3 assay] R/M-NSCLC treated with pembrolizumab or nivolumab after failure of platinum-based chemotherapy. Results The median age was 67 years, 13% of patients had ECOG-PS ≥ 2, and 27% were never-smokers. Adenocarcinoma was predominant (61%) and 18.1% harbored an EGFR activating mutation or ALK rearrangement. Pembrolizumab and nivolumab were administered to 51.3% and 48.7, respectively, and 42% received them beyond the third-line chemotherapy. Objective response rate (ORR) was 28.6%. Pembrolizumab group showed numerically higher ORR (30.7%) than the nivolumab group (26.4%), but it was comparable with that of the nivolumab group having PD-L1 TPS ≥ 50% (32.4%). Median progression-free survival (PFS) and overall survival (OS) were 2.9 (95% CI 0-27.9) and 10.7 months (95% CI 0-28.2), respectively. In multivariable analysis, concordance of TPS ≥ 50% in both PD-L1 assays and the development of immune-related adverse events (irAEs) were two significant predictors of better ORR, PFS, and OS. EGFR mutation could also predict significantly worse OS outcomes. Conclusion The real-world benefit of later-line anti-PD1 antibodies was comparable to clinical trials in patients with R/M-NSCLC, although patients generally were more heavily pretreated and had poorer ECOG-PS. Concordantly high PD-L1 TPS ≥ 50% and development of irAE could independently predict better treatment outcomes, while EGFR mutation negatively affected OS.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    33
    References
    0
    Citations
    NaN
    KQI
    []