Abstract
Immune checkpoint inhibitors (ICI) have demonstrated efficacy in the treatment of solid cancers. However, there is no unified predictive biomarker available for ICIs. We aimed to compare the prognostic impact of using three PD-L1 antibodies (SP142, SP263, and 22C3) for immunohistochemical (IHC) analysis. We retrospectively investigated tumor tissues derived from 316 breast cancer cases, by constructing tissue microarrays and by performing IHC staining. The immune-cell expression rate (for SP142 and SP263) and combined proportional score (for 22C3) were evaluated, and survival outcomes were analyzed. Prediction models were developed, and values of Harrel's c-index and areas under curves were calculated to compare the discriminatory power. Negative PD-L1 expression based on the 22C3-IHC assay was determined to be an independent prognostic marker for recurrence-free survival (RFS, P = 0.0337) and distant metastasis-free survival (DMFS, P = 0.0131). However, PD-L1 expression based on SP142- and SP263-IHC assays did not reveal a prognostic impact. Among the three antibodies, adding PD-L1 expression data obtained via 22C3-IHC assay to the null model led to a significant improvement in the discriminatory power of RFS and DMFS. We suggest that PD-L1 expression based on the 22C3-IHC assay is a superior prognostic marker than that based on SP142- and SP263-IHC assays.