PD-L1 Expression in Triple-negative Breast Cancer-a Comparative Study of 3 Different Antibodies

BACKGROUND: Assessment of programmed death protein-ligand 1 (PD-L1) in triple-negative breast cancer (TNBC) has entered daily practice to identify patients eligible for treatment with immune checkpoint inhibitors. However, different antibodies and different cut-offs for PD-L1 positivity are used, and the interchangeability of these methods is not clear. The aim of our study was to analyze whether different PD-L1 antibodies can be used interchangeably to identify TNBC patients as PD-L1 positive. METHODS: A tissue microarray encompassing 147 TNBC cases was immunohistochemically analyzed using 3 different antibodies against PD-L1: SP142, SP263, and E1L3N. PD-L1 positivity was determined as ≥1% of positive tumor-associated immune cells. The staining patterns of the 3 antibodies were compared and correlated with clinicopathological data. RESULTS: A total of 84 cases were evaluable for PD-L1 analysis with all 3 antibodies. PD-L1 was positive in 50/84 patients (59.5%) with SP263, in 44/84 (52.4%) with E1L3N, and in 29/84 (34.5%) with SP142. There was no statistical difference between the performance of SP263 and E1L3N, but both antibodies stained significantly more cases than the SP142 antibody. CONCLUSIONS: Our results show that the 3 PD-L1 antibodies identify different TNBC patient subgroups as PD-L1 positive and, therefore cannot be used interchangeably. Additional studies are needed to further investigate the use and impact of different PD-L1 antibody clones for predictive selection of TNBC patients for treatment with immune checkpoint inhibitors.