Nasopharyngeal carcinoma (NPC), can be distinguished from other head and neck cancers based on its clinical presentation, pathological features, and epidemiology, as well as its close association with the Epstein-Barr virus (EBV) (1). Cases of NPC have been observed worldwide, and the reported number of new cases and the number of deaths in both males and females of NPC in 2020 were 133,354 and 80,008 worldwide. The age standardized incidence and mortality rates were both remarkedly higher in Southeastern Asia than in any other region (2–4). The main treatment modality for early-stage NPC is radiotherapy (5, 6). In patients with locally advanced NPC, treatment using a combination of radiotherapy and chemotherapy has increased the 5-year local control rate to 90% (7, 8). However, about 10-20% patients can experience recurrence or distant metastasis (9–12). The treatment options for recurrence and metastatic patients are limited, and the efficacy also needs further improvement (13). Thus, identifying effective markers for stratifying NPC patients by risk of metastasis may aid in their timely treatment and management.

NPC is a virus-driven malignancy (14, 15) that is characterized by EBV infection and the presence of immune cell infiltration around the cancer nests (16). Pretreatment and postradiotherapy plasma EBV DNA levels correlate with disease outcomes and long-term survival of patients with advanced NPC (17). Although pretreatment plasma EBV DNA has been suggested as a reliable biomarker for the diagnosis, risk stratification, and prognosis of NPC patients (17, 18), studies differ in the cut-off values that they apply (17, 19) and methods to quantify EBV DNA in plasma have not been standardized (19, 20). In addition, a randomized controlled trial analyzed plasma EBV DNA in NPC patients after chemoradiation to identify high-risk patients for adjuvant chemotherapy, results showed that adjuvant chemotherapy with cisplatin and gemcitabine did not improve relapse-free survival (21). This study indicated that the EBV DNA levels alone were unable to identify high-risk patients for adjuvant chemotherapy (21). There is an urgent need to identify efficient markers to help with risk stratification and development of treatment options for NPC patients.

Programmed death-ligand 1 (PD-L1), also known as B7-H1 or CD274, is a surface glycoprotein that induces T cell anergy or apoptosis by binding to its receptor, programmed cell death protein-1 (PD-1), on T lymphocytes (22, 23). PD-L1 is widely expressed on antigen-presenting and other immune cells, and is upregulated in tumor cells in a broad range of tumors, including NPC (24–28). The expression of PD-L1 is correlated with tumor grade or prognosis in several types of carcinomas (22, 23, 29), and patients with EBV-associated malignancies have high levels of PD-L1 (26, 30). PD-L1 has recently been reported as a prognostic indicator in NPC patients, but these findings remain controversial (31–34). Victor H. F. Lee and colleagues reported that patients with higher PD-L1 expression on tumors had longer locoregional failure-free survival and marginally longer PFS (31). Another study also demonstrated that overexpression of PD-L1 on tumor cells was associated with superior OS (34). However, several studies have shown that PD-L1 expression predicts poor prognosis for NPC patients. Yajuan Zhou et al. found that patients with a higher PD-L1 H-score could have a greater risk of death (33). A meta-analysis study including 1,315 patients revealed that PD-L1 overexpression was associated with a poor OS (32). These studies with inconsistent results indicated that PD-L1 alone may not be a perfect prognostic indicator of NPC.

EBV associated malignancies had distinct regulation mechanisms on the expression of PD-L1. In NPC, PD-L1 expression was higher in EBV-positive NPC cell lines than in EBV-negative cell lines, and that PD-L1 expression can be enhanced by the exogenous and endogenous induction of the latent membrane protein 1 (LMP1) and IFN-gamma pathways (35). EBV-encoded circular RNA also upregulated the PD-L1 expression in NPC by retinoic acid-inducible gene I (36). Therefore, a combination of PD-L1 expression and EBV infection status may be effective at identifying NPC patients at higher risk of metastasis or poor survival.

In this study, we examined the tumoral PD-L1 expression and pretreatment plasma EBV DNA status in a cohort of patients with non-keratinizing NPC to assess the combined effect of these two factors on risk stratification and prognosis prediction.

NPC is an EBV-associated malignancy with high metastatic potential (39), and metastasis is the most frequent reason for treatment failure (40). Therefore, it is important to identify a reliable and efficient marker to predict metastasis risk. In this study, we examined the joint role played by tumoral PD-L1 expression levels and EBV status in predicting metastasis risk and prognosis of NPC patients. Our findings reveal that patients exhibiting high PD-L1 expression and EBV-positive status at NPC diagnosis are more likely to experience metastasis and shorter PFS.

Several studies have reported correlations between plasma EBV DNA levels and the prognosis and treatment outcomes of NPC patients (17, 18). However, the stratifying power of EBV DNA levels on their own may be limited (21). Better results may be obtained by using a combination of two or more biomarkers, as shown for the combinations of tumor volume and EBV status, pretreatment EBV DNA levels and clinicopathological variables, or pretreatment EBV DNA levels and cervical node necrosis (41–43). However, none of these studies stratified patients based on treatment modality. With the current advances in immunotherapy, it is worth considering the addition of immunotherapy to chemotherapy, especially for patients at high risk of metastasis (44).

PD-L1, which is expressed on immune, antigen-presenting, and tumor cells, plays an important role in T cell tolerance and immune escape. This might be the reason why patients with higher PD-L1 expression had shorter PFS and correlates with metastasis in this cohort. The mechanisms of PD-L1 upregulation induced by EBV infection are complex. In NPC, PD-L1 expression was reported to be regulated by EBV-induced LMP1 and IFN-gamma pathways (35). Recently, Ge and colleagues found that EBV-encoded circular RNA CircBART2.2 upregulated the expression of PD-L1 in NPC and further inhibited the T-cell function (36). In other malignancies with positive EBV infection, IFN-gamma mediated signaling pathways, EBV microRNA, somatic gene mutations, and epigenetic modifications were also reported to regulate PD-L1 expression (45, 46). Therefore, PD-L1 may be an effective target for immunotherapy as well as a useful biomarker for predicting metastasis and survival. Many studies have investigated the role of PD-L1 expression and EBV in the prognosis and risk stratification of NPC, as well as other EBV-associated malignancies, but most of them did not combine tumor PD-L1 expression with EBV-DNA to further stratify NPC patients (30, 33, 47–51) Hu and colleagues stratified NPC patients into three groups based on the combination of EBV DNA load and PD-L1 expression on tumor infiltrating lymphocytes (TILs). The results indicated that a high log EBV-DNA load with low TIL PD-L1 was a poor prognostic factor for PFS (52).

The results of present study indicate that tumoral PD-L1 expression and pretreatment plasma EBV status act as independent prognostic indicators, and both of them are associated with metastasis in NPC patients. Our results show that patients with high tumoral PD-L1 expression levels and detectable EBV DNA have the highest incidence of metastasis and worst PFS, indicating that the combination of these two biomarkers can help identify high-risk patients. For those patients with higher risk for metastasis and worse survival stratified by the combination of PD-L1 and EBV status, more intensive anti-tumor treatment may necessary to reduce the incidence of metastasis and recurrence, further to improve patients’ survival. Adding anti-PD-L1/PD-1 immunotherapy to conventional treatment regimens for these high-risk patients may further prolong their PFS. Several PD-1 blocking antibodies have shown promising results in the treatment of recurrent and metastatic NPC failed on previous treatment (53–56). Recently, results of two-phase 3 clinical trials showed that anti-PD-1 antibodies plus gemcitabine-cisplatin chemotherapy as first-line treatment in advanced NPC patients had superior PFS compared to chemotherapy alone, and this regimen could be a new standard of care for metastatic or recurrent NPC (57, 58). This progress might help to improve the treatment efficacy of patients with high risk stratified by PD-L1 expression and positive EBV status.

To our knowledge, this is the first study to examine the combination of tumoral PD-L1 expression and pretreatment plasma EBV DNA status as biomarkers for risk stratification and prognosis in NPC. The EBV DNA status is routinely tested during the management of NPC, and PD-L1 expression can be evaluated using low-cost immunohistochemistry. Using these two indices together may provide good risk stratification without the need for complex nomograms. Although our study was able to identify biomarkers that can be used for risk stratification in NPC patients, our findings were limited by the small sample size of the cohort and by heterogeneity in PD-L1 expression within the same patient. Future studies should investigate these associations further and validate the relationship between metastasis and PD-L1 expression and EBV status using larger samples.

This study revealed that pretreatment plasma EBV status and tumoral PD-L1 expression levels are closely related to prognosis of NPC patients. Using the two biomarkers together may facilitate risk stratification and timely, effective treatment.

The work was presented at the 60th American Society for Radiation Oncology Annual Meeting, 2018.

The original contributions presented in the study are included in the article/supplementary material. Further inquiries can be directed to the corresponding author.

The studies involving human participants were reviewed and approved by Institutional Review Board of West China Hospital of Sichuan University. Written informed consent for participation was not required for this study in accordance with the institutional requirements.

The corresponding author PA concepted and designed the study, edited and reviewed the manuscript. XL performed the experiment study, data analysis, statistical analysis, and manuscript preparation. XP worked on the clinical data acquisition, data analysis, statistical analysis and manuscript editing. SZ worked on the sample collection, data acquisition, data analysis, and manuscript editing. HZ worked on data acquisition. YJ worked on sample collection, experimental studies. FL searched the literature. All authors reviewed the manuscript and agreed with the submission.

This study was supported by grants from the Department of Science and Technology of Sichuan Province (Grant No. 2020YFS0276), the Health Department of Sichuan Province (130087) and the Wu Jieping Medical Foundation (320.6750.13317).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.