Abstract
Background: Immune checkpoint inhibitors (ICIs) have been an emerging treatment strategy for advanced triple-negative breast cancer (TNBC). Some studies have shown that high expression of programmed death-ligand 1 (PD-L1) can achieve a better response of clinical efficacy. However, the efficacy of ICIs in advanced TNBC remains controversial. In this meta-analysis, we evaluated the correlation of PD-L1 expression with the efficacy of ICIs in patients with advanced TNBC. Methods: We conducted a systematic search using four databases until March 2022 to obtain eligible randomized controlled trials (RCTs). The quality of the studies was assessed by the Cochrane risk of bias tool. Hazard ratio (HR) was extracted to evaluate the relationship between PD-L1 expression and progression-free survival (PFS) or overall survival (OS) in patients with advanced TNBC. Results: Five randomized controlled clinical trials (RCTs) with 3104 patients were included in this meta-analysis. The results demonstrated that ICIs could significantly improve the OS (HR 0.77, 95% CI 0.60-0.98, p = 0.03) in PD-L1 positive TNBC group. In the subgroup analysis, longer OS was observed (HR: 0.70, 95% CI: 0.60-0.82, p = 0.00001) in PD-L1 positive TNBC patients receiving ICIs alone or ICIs combined with nab-paclitaxel. In terms of PFS, PFS was significantly improved (HR: 0.68, 95% CI: 0.58-0.79, p < 0.00001) in PD-L1 positive patients receiving first-line ICIs and chemotherapy compared to those with ICIs alone. No significant improvement was observed for OS or PFS in PD-L1 negative group. Conclusion: Our study indicated significant improvement for OS in advanced TNBC with ICIs therapy in the PD-L1 positive status, and ICIs alone or ICIs combined with nab-paclitaxel might be a excellent choice in terms of OS. Although PFS has no significant benefit in PD-L1 positive patients, the subgroup analysis showed that ICIs combined with chemotherapy could achieve the PFS benefit in the first-line treatment. However, further clinical studies are needed to validate our conclusions due to limited relevant research.