1 Introduction
Oral squamous cell carcinoma (OSCC) is the most commonly diagnosed oral cancer, which accounts for approximately 90% of all malignant oral neoplasm1. Programmed cell death ligand-1 (PD-L1) is a surface glycoprotein of cancer cells, which is regarded as the inhibitory receptor programmed cell death-protein 1 (PD-1) on T lymphocytes2. PD-L1 involves in the immune response evasion by binding to PD-1 that downregulates T-cell responses.
As one of the most common immunologic checkpoints, the axis PD-1/PD-L1 has showed prognostic significance that mediates immune tolerance. Accompanied by the increasing number of researches finding that activation of immune checkpoint blockade is effective in tumour growth control and survival extension, controversy has also been persistent due to non‐homogeneous conclusions. Several studies approved of either positive3,4[4, 5], or negative association of PD-L1 overexpression with prognosis5,6, while other studies showed no prognostic significance of PD-L1 expression7,8.
Inconsistent results were mainly presumed to be related to the various methods for immunohistochemical staining, stage of diagnosis, tumor location, and the feasibility of radical surgical resection9. The previous meta-analyses reached a consensus that PD-L1expression in OSCC patients was not significantly correlated with overall survival (OS), but major limitation of them did not consider oral subsites or immunostaining patterns as confounding factors.
To date, the role of PD-L1 expression in OSCC of tongue has attracted attention of the increasing number of studies. It was reported that high PD-L1 expression was associated with worse OS that was exclusively found in OSCC of the tongue. However, the prognostic significance of PD-L1 expression was not observed in oropharyngeal carcinomas10. To find the discrepancy complicated by the use of different immunostaining patterns, comparison was conducted stratified by various scoring systems, antibodies clone. One study indicated that no relationship between PD-L1 expression and survival was found by using either combined positive score (CPS) or tumour proportion score (TPS)9. In contrast, other reports have found high PD-L1 expression was linking to better prognosis when cut off of ≥ 5% of TPS and ≥ 1 of CPS11. A recent approach assessed PD-L1 expression measured by two different anti-PD-L1 antibodies (22C3, E1LN3), and revealed that high PD-L1 expression detected by 22C3 presented a worse prognosis in terms of disease-specific survival (DSS)12.
Therefore, it was appropriate for performing a meta-analysis to determine whether the prognostic significances of PD-L1 expression in OSCC differed by oral compartments and methods for PD-L1 immunostaining.