A Review on the Application of PD-1 Blockade in EBV-Associated Nasopharyngeal Carcinoma Immunotherapy.

Epstein-Barr virus (EBV) linked with nasopharyngeal carcinoma (NPC) is considered to be one of the most prevalent head and neck malignancies in East and Southeast Asia. Although radiotherapy and chemotherapy are effective treatments for NPC, they have immunosuppressive effects. Immunotherapy has got considerable attention of clinicians for cancer treatment in recent years due to proven success of PD-1/PD-L 1 inhibition in solid tumors trials. The distinct immunological environment of EBV-associated NPC presents a reasonable therapeutic target for PD-1/PD-L 1 inhibition. Immune checkpoint blockade therapy targeting the programmed cell death-1 (PD-1) and programmed cell death ligand-1 (PD-L 1) receptors have shown efficacy in early phase I clinical trials, with ongoing phase III clinical trials. Herein, we have extensively addressed the role of the PD-1/PD-L1 axis in the immunotherapy of EBV-associated NPC. Immunotherapeutic strategies are anticipated to enter mainstream clinical practise and provide long-term remissions in patients with severe NPC.

1. Introduction

Epstein-Barr virus (EBV) belongs to the gamma-herpesviruses subfamily that infects over 90% of the population across the globe [1]. It is the first identified oncogenic virus in human cancers and is considerably linked with nasopharyngeal carcinoma (NPC) [2]. EBV-associated NPC is a kind of head and neck cancer that is characterized by its epidemiological and pathological properties, as well as its clinical presentation, such as poor tumor cells differentiation, high susceptibility to chemotherapy and radiotherapy, and geographical global distribution [3, 4]. Based on the International Agency for Research on Cancer, around 129,000 new NPC cases were diagnosed in 2018 across the globe [5]. Although it accounts for only 0.7% and 0.8% of the total cases and deaths, accordingly, its geographical global distribution is heavily unbalanced where more than 70% of new cases come from East and Southeast Asia [5].

Over the past decades, radiotherapy and chemotherapy have been considered basic treatments for EBV-associated NPC. Immunotherapy has recently opened up a new avenue for treating EBV-associated NPC. Unlike traditional radiotherapy and chemotherapy, which have immunosuppressive side effects, immunotherapy turns the cold immunosuppressive tumor microenvironment into a hot one to better recognize and attack tumor cells. So far, there are several types of immunotherapy that have been applied for cancer treatment, such as adoptive immunotherapy, therapeutic vaccines, T cell transfer therapy, and immune checkpoint inhibitors [6]. Immunological checkpoint inhibitors have become the focus of intense research in recent years as one of the critical therapeutic strategies for NPC. Immune checkpoint inhibitor is an effective immunotherapy, releasing the immune system's brake, and reactivates endogenous antitumor immunity. Programmed death-1 (PD-1) has a key contribution to fine-tuning T lymphocyte functioning and regulation of the immune-system, which makes it one of the most extensively researched regulators. There are two ligands of PD-1, namely, B7-H1 (PDL 1) and B7-DC (PD-L2), which are upregulated during an inflammatory response to infectious agents in peripheral tissues [7]. In this review, we discussed in detail about the use of the PD-1/PD-L1 axis in the immunotherapy of EBV-associated NPC. This review will provide a consolidated piece of information for researchers working on immunotherapeutic strategies against EBV-associated nasopharyngeal carcinoma.

2. Tumor Immunity Cycle and Its Mechanisms

Malignant cell changes often occur in the normal body. The body's immunity can identify and kill the cancerous cells via cellular immune mechanisms, thus generating the “immune surveillance theory.” Hence, it has been suggested that there is a considerable link between the tumor and the immune system. The immune system not only has the protective function of resisting cancer but also has the function of the tumor cell immune selection, making reshape tumor cells' immune, weakening immunogenicity cells that can further enhance immune escape, and the theory of “immune to edit”.

It is now believed that the generation of immunity to cancer is a cyclic process which is completed in 7 basic steps. Cancer antigens are secreted by tumor cells postapoptotic process, in Step 1. DCs, which form and migrate to lymph nodes at the same time, collect the antigens produced in the second stage. In Step 3, at the lymph nodes, the acquired tumor antigen is presented to the major histocompatibility complex-I (MHC-I) molecule, resulting in T cell priming. The activated T cells then begin to migrate to the tumor (Step 4), followed by infiltration into the tumor tissue (Step 5), tumor cell identification (Step 6), and damage (Step 7). Steps 3 through 6 function as immunological checkpoints that regulate T cell activation. Cancer antigens are released from cancerous cells damaged through T lymphocytes in Step 7, and the cancer-immunity cycle returns to Step 1 and repeats. One or more of these stages can be disrupted in many cancer patients, resulting in inefficient immune responses to tumors [8]. As can be seen, there are many immune factors involved in many aspects of the body's immune reaction to tumor, such as stimulating factors and inhibiting factors; thus, they can provide many potential therapeutic targets for immunotherapy of cancer.

3. The Discovery of PD-1/PD-L1 Axis

PD-1 was first reported in 1992 by Ishida et al. [9]. There are three domains of PD-1 including Ig variable-type extracellular domain, a transmembrane domain, and a cytosolic domain. CD28, CTLA-4, and ICOS all share a 21–33% sequence identity with the extracellular domain [10]. It has also been revealed that the membrane-proximal Cys residue required for homodimerization is missing in PD-1. PD-1 is monomeric in solution and on the cell surface, according to biochemical and biophysical analysis [11]. The physiological relevance of PD-1 remaining elusive for several decades post its discovery until numerous studies demonstrated that mice with low levels of PD-1 acquired autoimmune disorders including glomerulonephritis, lupus-like arthritis, diabetes, and dilated cardiomyopathy [12-15]. The underlined data reveals that PD-1 considerably contributes in maintaining peripheral self-tolerance as well as negatively regulating immunological reactions.

Freeman et al. reported PD-L1 as a PD-1 ligand in 2000, and it is expressed by APCs, particularly in the lungs, heart, placenta, and kidney [16]. In 2001, PD-L2, the second ligand of PD-1, was identified, and the functional study revealed the prominence of PD-1/PD-L1/PD-L2 in controlling T cell reactions [17]. PD-L1 and PD-L2 have IgC and IgV extracellular domains that share approximately 40% sequence identity [16, 17]. PD-L1 and PD-L2 contain short cytoplasmic tails with unknown motif, and their functions in signal transduction reveal that both PD-L1 and PD-L2 do not transduce signals through their binding with PD-1 [10].

4. Contribution of PD-1 in Tumor Immunological Tolerance

Different from other members in a CD28 family that are expressed on the surface of T lymphocytes, PD-1 is present on both T cell surfaces and activated B cell surfaces [18]. The wider expression of PD-1 indicates that PD-1 regulates several immune reactions [10]. The ligands of PD-1 have distinct expression patterns. PD-L1 expression has been observed in many cells, such as hematopoietic and nonhematopoietic cells such as vascular endothelial cells, as well as immune privilege sites including the testes and eye [19]. PD-L1 has been expressed in various tumor and virus-infected cells. In contrast, PD-L2 expression has more restrictions, primarily on DCs and macrophages. In normal tissues, but PD-L1 contributes in peripheral immune tolerance to control autoimmune reactions postpersistent inflammatory reactions to tissue injury. Tumor cells express PD-L1, which they use as a molecular shield to reduce cytotoxicity (activated by T cells) and avoid immune monitoring [20]. Furthermore, PD-L1 decreases the proliferative potential of stimulated T cells in peripheral tissues by interacting with PD-1, resulting in T cell exhaustion that can be reversed by PD-1 attenuation [16].

5. PD-1/PD-L1 Expression and the Clinical Stage and Survival of Patients with EBV-Associated NPC

PD-1 is normally expressed on the membrane of triggered T lymphocytes. As a ligand of PD-1, PD-L1 is predominantly expressed on the tumor cell membrane and is also expressed on tumor-related macrophages and DCs. The link between PD-1 and PD-L1 can prevent T cells from activation and proliferation, leading to tumor immunosuppression and immunological escape, as well as tumor recurrence and metastasis [21].

The PD-1/PD-L1 expression in various tumor tissues is linked with the stage and survival rate of tumor patients. Shi et al. [22] revealed that an elevated PD-L1 expression in cancerous cells of patients with colorectal cancer was linked with TNM stage and survival rate. Karim et al. revealed that PD-L1 was expressed in a few samples of cervical cancer tissues, affecting patients' prognosis [23]. According to Nomi et al., PD-L1-positive pancreatic cancer patients had a bad survival than PD-L1-negative patients [24]. Muenst et al. revealed that PD-L1 expression in breast cancer (BC) tissues was linked to tumor size, pathological grade, and lymph node metastasis (LNMets), with positive patients having a considerably shorter overall survival (OS) [25]. According to Frigola and Thompson et al., the expression of PD-L1 in patients with renal cancer was linked to tumor stage and prognosis [26, 27]. Azuma et al. [28] indicated that elevated PD-L1 expression in NSCLC tumor cells was linked to EGFR mutation and that all of them had a bad prognosis. According to the reported studies, PD-L1 was substantially expressed in the cancer tissues of patients with ovarian cancer [29], malignant melanoma [30], and esophageal cancer [31], and all patients had a bad prognosis.

Similarly, Zhang et al. [32] investigated the relationship between PD-1/PD-L1 expression and posttreatment outcomes in NPC patients. They discovered increased levels of PD-L1 expression in the malignant tissues of nasopharyngeal carcinoma patients and revealed that increased expression of PD-L1 also existed in cancer tissues of patients with NPC. In 52 out of total 139 tumor patients, 37.4% of immune cells were found to have positive PD-1 expression. PD-L1 expression was found to be 95.0% in 132 of the tumors. An elevated level of PD-L1 expression in cancerous cells was linked to a bad prognosis for DSL (P value = 0.009). It has been revealed that the coexpression of PD-1 and PD-L1 in NPC was associated with a poor disease-free survival (DFS) prognosis (P value = 0.038). PD-1/PD-L1 coexpression reflected the tumor microenvironment's selective cytotoxic T cell inhibition and predicted NPC recurrence and progression postconventional treatments. PD-1/PD-L1 cascade attenuation could be a candidate target for NPC. The summary of PD-L1 in different carcinomas and its functions is shown in Table 1.

Table 1

PD-L1 correlation in different carcinomas and its functions.

Type of cancer	Function	References
PD-L1 in colorectal carcinoma (CRC)	The presence of PD-L1 is an independent predictor of CC prognosis. Proliferating, migrating, and invading cells can be prevented by B7-H1 knockdown.	[22]
PD-L1 in cervical cancer (CC)	PD-L1 has no considerable impact on the survival of patients with CC.	[23]
PD-L1 in pancreatic cancer (PC)	PD-L1 is a new prognostic factor for patients suffering from PC.	[24]
PD-L1 in BC	The PD-L1 expression is an independent negative prognostic factor in human BC.	[25]
PD-L1 in renal cell carcinoma (RCC)	Soluble PD-L1 has been determined in the sera of patients suffering from RCC and may systemically impair host immunity. Furthermore, the survival rate of RCC patients (with B7-H1 tumors) is very low.	[26, 27]
PD-L1 in non-small cell lung carcinoma (NSCLC)	In the case of surgically removed NSCLC, PD-L1 elevated expression was linked with the mutations in EGFR that independently predicted the survival outcomes for this carcinoma.	[28]
PD-L1 in ovarian cancer (OC)	PD-L1 expression on cancerous cells is an independent predictor of OC.	[29]
PD-L1 in malignant melanoma (MM)	Melanoma patients with an elevated expression of PD-L1 have a poorer survival rate.	[30]
PD-L1 in esophageal cancer (EC)	PD-L1 and PD-L2 could predict the survival outcomes of patients suffering from EC, paving the way for new immunotherapies that target the PD-1/PD-L cascade.	[31]
PD-L1 in nasopharyngeal carcinoma (NPC)	The co-expression of PD-1/PD-L1 revealed the tumor microenvironment's selective suppression of cytotoxic lymphocytes and predicted NPC recurrence and development post conventional treatments.	[32]

The continuous PD-1 expression (on T lymphocytes) is a high indication of T cell depletion and decreased function. Muenst et al. [33] revealed that BC PD-1 positive TILs (tumor-infiltrating lymphocytes) were found to be correlated with tumor size, grade, LNMets, and poor OS. Maxime et al. [34] revealed that high PD-1 expression (on T lymphocytes) in malignant melanoma had a higher incidence of distant metastasis. According to Zhang et al. [35], elevated PD-1 expression by CD8+ T ILs in patients with NSCLC resulted in reduced cytokine production and decreased proliferation. Furthermore, Thompson et al. [36] reported that positive tumor immune cell PD-1 was associated with late-stage and tumor-related death in renal cancer patients.

Hsu et al. [37] used immunofluorescence staining methods for analysis of PD-1 in normal nasopharyngeal tissue and expression in nasopharyngeal carcinoma tissue. According to his findings, the PD-1 on CD8+ T lymphocytes in nasopharyngeal carcinoma tissue was considerably elevated than normal nasopharyngeal tissue (P < 0.0001) but had no correlation with clinical stage of the disease. NPC patients with an elevated PD-1 expression had a poorer survival, shortened DFS and OS, and increased local recurrence (P values = 0.05, 0.007, and 0.004, accordingly). Thompson et al. [36] revealed that the positive rates of PD-1 and PD-L1 were 37.4% and 95%, accordingly, in NPC tissues. Age, sex, LNMets, and the disease stage had no effect on the expression of PD-1/PD-L1. In addition, PD-1 expression had no obvious impact on prognosis (P = 0.57). Elevated expression of PD-L1 considerably shortened disease-free survival and had poor prognosis (P = 0.009), while the simultaneous expression of PD-1/PD-L1 had the worst prognosis.

It has been revealed that PD-1 signaling cascade is important for regulating tumor recurrence. The PD-1 is identified as a key marker for T cell dysfunction. Evaluation of PD-L1 expression showed that 16/18 (89%) EBV-associated NPC patients indicated positive PD-L1 staining in malignant cells [38]. According to the reported study by Hsu et al., CD8 T cells of tumor tissue have a significantly elevated PD-1 expression level compared to the cells of normal tissue. This suggests a close relationship between PD-1 and NPC recurrence, metastasis, and patient clinical progression [37]. In a cord blood-humanized mouse model, combining PD-1 and CTLA-4 blockade reduces the proliferation of lymphomas (induced by EVB), suggesting that PD-1/CTLA-4 blockade could serve as a candidate therapeutic target for EBV related complications [39].

6. Difficulty in Targeting PD-1/PD-L1 Cascade in EBV-Associated NPC

Emerging immunotherapy offers more treatment options for patients with nasopharyngeal cancer. The clinical efficacy of immunotherapy is also worthy of recognition. Recently, it has been reported that several specific antibodies against PD-1 and PD-L1 are active in head and neck squamous cell carcinoma [40]. Evaluation of several anti-PD-1 monoclonal antibodies in recurrent or metastatic NPC showed promising clinical activity. For example, a phase II clinical trial evaluated nivolumab, which is a human IgG4 anti-PD-1 monoclonal antibody. In 44 patients with pretreated metastatic or recurrent NPC, the objective response rate (ORR) was 20.5%. The 1-year OS rate was 59%, and the 1-year progression-free survival (PFS) rate was 19.3% [41]. Pembrolizumab is another PD-1 monoclonal antibody and was received by 27 patients with metastatic or unresectable NPC in a phase I clinical evaluation. The treatment resulted in a 25.9% ORR, 63% 1-year OS, and 33.4% 1-year PFS [42]. In another phase I clinical trial, Fang et al. evaluated camrelizumab (anti-PD-1 antibody) alone as the second-line therapy and in combination with gemcitabine and cisplatin as the first-line therapy in patients with recurrent or metastatic NPC. In the camrelizumab monotherapy trial for 93 patients, the ORR was 34% and the 1-year PFS was 27.1%. The combination treatment for 23 patients led to the ORR of 91% and a 1-year PFS rate of 61.4% [43]. The summary of PD-1/PD-L1 antibodies in different clinical evaluations is shown in Table 2. Since these clinical trials with anti-PD-1 antibodies exhibit promising anticancer potentials and manageable toxicity profiles, a couple of phase III clinical trials for NPC treatment are ongoing [4]. Hopefully, these ongoing investigations will provide valuable information for the usage of PD-1/PD-L1 axis in NPC patients.

Table 2

Summary of PD-1/PD-L1 antibodies in several clinical evaluations.

PD-1/PD-L1 antibody applied in NPC patients	Clinical trial phase	Objective response rate	The 1-year OS rate	The 1-year DFS rate	References
Nivolumab	II	20.5%.	59%	19.3%	[41]
Pembrolizumab	I	25.9%	63%	33.4%	[42]
Camrelizumab	I	34%	—	27.1%	[43]

However, PD-1/PD-L1 inhibition also has some disadvantages in clinical application, which cannot be ignored. Drug resistance is a major problem faced by immunocheckpoint inhibitors. Studies have shown that the observed drug resistance is as high as 60% in patients receiving anti-PD-1 treatment [44]. Therefore, how to overcome the durability of immunocheckpoint inhibitors will be a next new challenge. In addition, studies [45] have shown that immune tolerance is a major obstacle to immunotherapy. Recent studies [46] have shown that the interaction between ANXA 2 (Annexin A2) and DC-SIGN (dendritic cell-specific intracellular adhesion meggle-3 grabbing nonintegrin) molecules can cause immunosuppression and lead to tumor immune escape. Thus, the development of ANXA 2 targeted therapy may overcome the drug resistance of patients with nasopharyngeal carcinoma with high expression of ANXA 2.

A number of recent research studies support the combining immunotherapy with traditional therapies, including combining immunonode inhibitors with chemotherapy or radiotherapy, which not only provides more options for patients with nasopharyngeal cancer but also offers the possibility of comprehensive and individualized treatment of NPC. The combination of nasopharyngeal immunotherapy and traditional therapy is not simply a random combination of existing approaches but to find the treatment combination that can achieve the highest synergistic effect. Although most of the immunotherapy strategies for EBV-related nasopharyngeal carcinoma are still in clinical trials, based on the previous research results and the emerging latest research results, the immunotherapy of nasopharyngeal carcinoma will be promoted to a new level.

7. Conclusions

In conclusion, with the deepening of people's understanding of the relationship among tumorigenesis, development, and immune system, immunotherapy has gradually become a development direction of comprehensive treatment model for nasopharyngeal carcinoma. Among them, immunocheckpoint inhibitor therapy has shown the most promising prospect in the basic and clinical research of nasopharyngeal carcinoma.

In spite of the fact that PD-1/PD-L1 immunocheckpoint inhibitor therapy has enhanced the prognosis of patients with NPC, it is currently only helpful for a few patients and has some disadvantages that cannot be ignored. In a number of domestic and foreign studies on the linkage between the PD-1/PD-L1 expression in nasopharyngeal carcinoma and clinical stage and prognosis, the results are not completely consistent [36, 40]. However, based on the above studies, we can draw the following conclusions: patients with nasopharyngeal carcinoma simultaneous PD-1/PD-L1 expression have the worst survival, and the disease-free survival period is significantly shortened. Therefore, in tumor immunotherapy, attenuating the PD-1/PD-L1 cascade might be relatively more significant for these patients.

At present, many studies are only retrospective analysis, and the study sample size is small; also the follow-up time of patients is short. Therefore, in order to explore the immune system's role in the occurrence and development process of nasopharyngeal carcinoma (NPC), it is extremely important to offer a foundation for the comprehensive treatment model of radiotherapy and chemotherapy combined with immunotherapy for nasopharyngeal carcinoma. Therefore, in order to understand a better theoretical basis for the prediction of tumor immunotherapy efficacy for the PD-1/PD-L1 cascade in nasopharyngeal cancer, there is still a need to conduct extensive and in-depth translational studies on immunotherapy for nasopharyngeal carcinoma. Meanwhile, we should reasonably choose the treatment plan according to the tumor staging and physical condition of the patients, so as to bring greater survival benefits to the vast number of patients with nasopharyngeal carcinoma.