Clinicopathological and prognostic significance of programmed death ligand-1 expression in breast cancer: a meta-analysis

Abstract

Background: Programmed cell death-ligand 1 (PD-L1) may be a useful molecule for targeted immunotherapy. Therefore, this meta-analysis aimed to investigate PD-L1 expression in breast cancer and its associations with clinicopathological factors and outcomes, which may help determine whether PD-L1 expression is a useful prognostic marker.

Methods: The Medline Ovid, Cochrane, PubMed, Google Scholar, and Web of Knowledge databases were searched for studies that evaluated the prognostic or clinicopathological significance of PD-L1 expression in patients with breast cancer, and reported at least one survival-related outcome.

Results: Six studies that included 7877 cases were selected for the analysis. Higher PD-L1 expression in all cells was related to higher histological grade and lymph node metastasis. Higher PD-L1 expression in tumor cell was related to larger tumor size, estrogen receptor negativity, progesterone receptor negativity, human epidermal growth factor type-2 positivity, and triple-negative breast cancer. PD-L1 positivity in all cells was associated with poorer disease-free survival, although it was not significantly associated with overall survival.

Conclusion: The present meta-analysis revealed that cases of breast cancer with PD-L1 positivity in all cells exhibited higher histological grades, lymph node metastasis, and poorer disease-free survival. Therefore, positive expression of PD-L1 may be a useful prognostic marker in breast cancer.

Keywords: Breast cancer; Meta-analysis; PD-L1; Prognosis.

Fig. 1

Flow chart of the literature search and study selection

Fig. 2

Forest plots of studies that assessed the association between PD-L1 and clinicopathological factors in all cells. a Histological grade. b Lymph node metastasis

Fig. 3

Forest plots of studies that assessed the association between PD-L1 and clinicopathological factors in tumor cells. a Tumor size. b Estrogen receptor status. c Progesterone receptor status. d Human epidermal growth factor receptor 2 status. e Molecular subtype

Fig. 4

Forest plots of studies that assessed the association between PD-L1 and survival outcome in all breast carcinoma cells. a Disease-free survival. b Overall survival. c Overall survival without one study (Baptista et al. 2016, reference [11])

Fig. 5

Egger’s test and funnel plot results for all included studies. a Overall survival based on all cells (p = 0.17). b Disease free survival based on all cells (p = 0.15)