INTRODUCTION: Whether programmed cell death ligand 1 (PD-L1) expression differs between different specimen types remains unclear. This meta-analysis aimed to compare PD-L1 expression between biopsy and surgical specimens in non-small cell lung cancer (NSCLC). METHODS: We searched the PubMed and Web of Science databases from January 2015 to December 2025 to evaluate PD-L1 expression detection across different types of specimens. The QUADAS-2 tool was used to evaluate the quality of the included studies. Review Manager 5.4 and Stata 17.0 were used to calculate the relative risk (RR) and the corresponding 95% confidence interval (95% CI). RESULTS: This meta-analysis of 13 studies involving 966 patients demonstrated that no significant difference in PD-L1 positivity rates between biopsy and surgical specimens was observed at the 1% cutoff (RR = 0.85, 95% CI [0.71 - 1.01], P = 0.07), but a significant difference was detected at the 50% cutoff (RR = 0.76, 95% CI [0.65 - 0.90], P < 0.01). Similar findings were observed in patients without neoadjuvant therapy. Additionally, PD-L1 expression in tumor cells was significantly lower with the SP142 assay than with other antibodies at both the 1 and 50% cutoffs. CONCLUSIONS: PD-L1 expression may differ between biopsy specimens and surgically resected specimens in patients with advanced NSCLC, even in patients who have not received neoadjuvant therapy. Consequently, PD‑L1 status evaluated from biopsy specimens should be interpreted with caution in clinical practice. In addition, the selection of PD-L1 immunohistochemistry (IHC) detection methods may also affect the evaluation results.
Keywords
