Comparative analysis of neoadjuvant chemoimmunotherapy (chemoIO) versus PD-(L)1 monotherapy (mono) across PD-L1 expression strata in resectable non–small cell lung cancer (NSCLC): A systematic review and meta-analysis of prospective trials.

Abstract

e20069 Background: Neoadjuvant chemoIO is a standard treatment for resectable NSCLC, and PD-L1 expression correlates with outcomes. However, cross-trial comparative activity versus PD-(L)1-mono across PD-L1 subgroups remains unclear and may help identify patients for chemotherapy-free approaches, as in the metastatic setting. Methods: MEDLINE and SCOPUS (1/2018 to 8/2025) identified prospective neoadjuvant PD-(L)1 mono or chemoIO trials reporting pCR or MPR (excluding observational reports, CTLA-4 or dual ICB, and RT). Random-effects meta-analyses of arm-level proportions used inverse-variance logit models to estimate pooled pCR/MPR by regimen and PD-L1 status, with meta-regression including regimen, PD-L1, and their interaction. Individual patient data (IPD) were reconstructed from Kaplan-Meier curves for exploratory survival analyses using shared-frailty Cox models. Results: Thirty-four treatment arms comprising 2,640 patients were included. Pooled pCR/MPR increased from 6.4%/16.5% with PD-(L)1-mono to 19.7%/35.2% with chemoIO in PD-L1-negative tumors, and from 13.2%/25.8% to 36.1%/53.7% in PD-L1-positive tumors. In PD-L1 ≥50% disease, pooled MPR was 39.3% with PD-(L)1-mono versus 61.3% with chemoIO. Meta-regression estimated a 16.6% higher pCR with chemoIO (95% CI, 8.9 to 24.3) averaged across PD-L1 strata (Table), with a similar pattern for MPR. Exploratory KM-reconstructed IPD analyses, using a shared-frailty Cox model, suggested an apparent EFS signal favoring PD-(L)1-mono versus chemoIO in PD-L1-positive cohorts (HR 0.43; 95% CI, 0.19 to 0.98), noting cross-trial confounding. Conclusions: ChemoIO was associated with higher pCR/MPR than PD-(L)1-mono across PD-L1 strata. Exploratory EFS findings suggest some patients with PD-L1-positive disease may achieve favorable long-term outcomes with PD-(L)1-mono despite nominally lower pCR/MPR, supporting further prospective biomarker-directed studies to refine perioperative strategies. pCR by treatment regimen and PD-L1 status. Treatment regimen PD-L1 status No. of arms Meta-analysis pooled pCR rate % (95% CI) Meta-regression estimated pCR rate % (95% CI) Absolute difference from reference % (95% CI) P value PD-(L)1-mono PD-L1 negative 7 6.38 (2.66–14.54) 2.63 (-4.60–9.85) Reference — PD-(L)1-mono PD-L1 positive 7 13.20 (7.71–21.70) 16.40 (9.21–23.58) 13.77 (6.48–21.06) 0.0002 ChemoIO PD-L1 negative 13 19.69 (13.12–28.48) 19.24 (13.56–24.91) 16.61 (8.93–24.28) <0.0001 ChemoIO PD-L1 positive 13 36.07(28.83–44.00) 33.01 (27.10–38.91) 30.38 (19.64–41.11) <0.0001