Abstract
BACKGROUND: Immune checkpoint inhibitors (ICIs)-based neoadjuvant therapy has been regulatory approved in clinical practice since 2021. However, it is still difficult to determine which patients can benefit from it. Here, we conducted a meta-analysis to evaluate the predictive values of programmed cell death ligand 1 (PD-L1) in pan-cancer neoadjuvant immunotherapy. METHODS: We searched MEDLINE and EMBASE for randomized controlled trials (RCTs) to collect information regarding pathological complete response (pCR) and event-free survival (EFS) in patients with PD-L1-positive and PD-L1-negative tumors. Odd ratio (OR), hazard ratio (HR), and their 95% confidence intervals (CIs) were calculated. RESULTS: Totally, 10353 patients with 6 tumor types in 23 RCTs were included in this study. Neoadjuvant immunotherapy was associated with increased pCRs in both patients with PD-L1-positive (OR, 3.22; 95% CI, 2.25-4.61; P < 0.001) and PD-L1-negative tumors (OR, 2.07; 95% CI, 1.42-3.00; P < 0.001). However, compared with PD-L1 negative tumors, PD-L1 positive tumors benefited more from ICB-based neoadjuvant therapy (interaction effect, 0.65; 95% CI, 0.45-0.94; P(Interaction) = 0.01). Similarly, neoadjuvant immunotherapy resulted in favorable EFS in patients with PD-L1 positive (HR, 0.55; 95% CI, 0.46-0.66; P < 0.001) and PD-L1 negative tumors (HR, 0.70; 95% CI, 0.62-0.80; P < 0.001), the efficacy differences were also significant (interaction effect, 1.24; 95% CI, 1.03-1.50; P(Interaction) = 0.04). CONCLUSION: Both patients with PD-L1-positive and PD-L1-negative tumors can benefit from neoadjuvant immunotherapy. However, the magnitude of efficacy is greater in patients with PD-L1-positive tumors. Accordingly, rather than serving as an independent marker for patient selection, PD-L1 expression is more effectively applied as a prognostic biomarker.