Uveal melanoma (UM) is the most common primary cancer of the adult eye (1). Owing to its rarity and complexity, it is one of the most challenging and hardest cancers to study. Its aggressiveness, invasion potential, and high metastasis susceptibility in almost half of the patients impact its prognostic value and eventually decrease patient survival (1, 2). Various clinical and histological aspects are related to worse prognosis in patients with UM, including tumor location, tumor thickness, large tumor basal diameter, involvement of the ciliary body, epithelioid subtype, and cytogenetic features (3). Furthermore, when small, UM cells can be killed by the immune system. However, in later stages, immune checkpoints help the tumor grow and spread by weakening the immune system (4). To identify and eliminate these cancer cells, some immune checkpoints are turned on or overexpressed to stop the immune response against the tumor (5). UM may use these pathways to avoid being attacked by the immune cells. This may lead to tumor escape from the immune system, resulting in tumor growth and spread (6).

While Programmed cell Death protein-1 (PD1), Programmed Death Ligand-1 (PD-L1), and Cytotoxic T-Lymphocyte-Associated protein-4 (CTLA-4) blockade showed successful responses in patients with cutaneous melanoma (7, 8), patients with UMs were excluded from these studies because their molecular characteristics tend to be widely different from cutaneous melanoma (9, 10). Nevertheless, even with the discovery of different therapies, no definitive cure has been established, especially in patients with metastatic UM (11). Although surgery and radiotherapy are conservative treatment options for a subset of patients, up to one-third of UM may ultimately metastasize (12). This lack of effectiveness may be due to the exceptional microenvironment of the eye and the special mechanisms by which UM escapes the immune response. In fact, little is known about the implication of these mechanisms in this type of cancer, and the cause of the limited response of UM patients to immunotherapy is still unclear and blurred. Hence, our current challenge remains in the identification of additional suppressive pathways.

VISTA (V-domain Ig Suppressor T-cell Activation) encoding C10orf54, also known as V-set immunoregulatory receptor (Vsir), is a type I transmembrane protein located on chromosome 10q22.1. Demonstrated to dull T cell activation, VISTA is highly expressed in myeloid cells (such as macrophages, monocytes, and dendritic cells), on T cells (such as CD4+ and CD8+), and negatively expressed on B cells in several in vitro and in vivo studies (13). VISTA is a novel immune checkpoint that regulates T-cell function. Its role has been studied in several types of cancer, including gastric and ovarian cancers (14, 15). In murine melanoma tumor models, VISTA mAb treatment induced the activation of T cells, suppression of tumor growth, and stimulation of the immune response (16); proving its role in regulating the tumor immune response.

Little is known about the expression of distinct immune checkpoints and their function in UM, and no data exists regarding VISTA expression or its impact on the UM microenvironment. In the present study, we aimed to analyze VISTA expression within the tumor microenvironment of UM patients, to examine its association with clinicopathological features, to evaluate its prognostic factor, and to correlate it with PD1 and CTLA-4, two immune checkpoints already studied in UM pathology. Here, we propose that VISTA may be a novel engaging immune checkpoint and a new target for cancer immunotherapy in UM patients.

Over the past decades, therapeutic strategies have revolutionized cancer treatment (17). Still, no treatment has been totally effective for UM patients to date (11). Therefore, a better understanding of the molecular signatures of UM is necessary. To the best of our knowledge, this is the first study to assess the prognostic value of VISTA in UM patients.

The Cancer Genome Atlas (TCGA) is a database with a large amount of genomic data that enables us to study the immune microenvironment and molecular pathways involved in UM progression and invasion. It is also an interesting tool for confirming research studies worldwide because of its open accessibility. Here, we performed transcriptomic analysis as it is an advantageous tool filling the gap between genomics and proteomics and guiding translational and clinical studies (18, 19).

In this dataset, we analyzed the recently discovered immune checkpoints, VISTA, LAG-3, PD-L1, PD1, TIGIT, CTLA-4, and BTLA, to compare their expression within the UM immune microenvironment. Except for VISTA, these available genes in our TCGA database have been previously studied in UM (20–23). Remarkably, VISTA expression was the highest ( Figure 2 ). RT-PCR analysis and flow cytometry of several mouse tissues determined that VISTA is not expressed in the normal eye. Besides, higher expression has been observed in the thymus, spleen, and bone marrow (13). This supports the fact that VISTA may be highly expressed in UM patients but not in healthy individuals, possibly playing a major role in cancer progression.

Clinicopathological features of our patients were assessed according to the AJCC classification which considers ciliary body involvement and episcleral extension. Subsequently, both cohorts showed no significant differences in sex or age ( Figures 3A, B ).

Indeed, no gender preference is stated for this disease, while others claim that UM is most common in males (24). It has also been found that UM is common in elderly patients and is linked to lower overall survival. In addition, fair skin, light eye color, and Caucasian populations are among the risk factors for UM (25). Consistent with our data reporting that VISTA is associated with clinicopathological features linked to poor prognosis in the UM microenvironment, for instance, advanced stages and epithelioid cells ( Figures 3D, E ); it was reported that the presence of epithelioid cells is associated with UM progression and metastasis development. However, better survival was assigned in patients with the spindle subtype (1).

Subsequently, immunohistochemistry was used to quantify VISTA protein expression and different scoring approaches have been assessed in this study in order to confirm transcriptomic results. Eventually, both the percentage of positive immunostaining and the intensity were considered. Previous studies have only limited the expression of VISTA to immune cells (16, 26). Conversely, recent studies have provided clear evidence that both immune and tumor cells express it (14, 27, 28). Remarkably, immune cells, tumor cells, and endothelial cells all were shown to express VISTA ( Table 2 ), as found in patients with gastric cancer (14). Likewise, UM harbors higher expression of CD4+ CD8+ and CD11b+ cells, which are constitutively expressed by VISTA (29). Consistent with these results, other studies have reported that VISTA protein is expressed on T cells, and its high expression is associated with high levels of CD3+, CD4+, and CD8+ T cells (30).

In the same cohort, the results indicate that the VISTA protein has elevated expression in higher stages. This suggests that it would have a worse prognosis in UM patients with advanced stages. Indeed, it has been proven that VISTA has a worse prognosis in multiple types of cancers, for instance, ovarian cancer, human non-small cell lung cancer, cutaneous melanoma, glioma, and colon cancer (15, 27, 29, 31, 32). In contrast, VISTA has a good prognosis in malignant pleural mesothelioma and breast cancer, specifically in triple-negative patients, cervical cancer, and endometrial cancer, with better overall survival in patients with higher VISTA expression. Single-cell analysis and proteomic studies using both immunohistochemistry and quantitative immunofluorescence results reported lower VISTA expression levels in adjacent tissues than in breast cancer cells. Interestingly, VISTA expression was significant in terms of overall survival rate (33–36). In our sections, VISTA staining was detected mainly in T cells, monocytes, macrophages, and neutrophils. Consistent with our findings, VISTA expression in T cells and in the myeloid lineage leads to their regulation and was found to be associated with poor prognosis. Consequently, when VISTA is upregulated, the levels of IL10, IFN gamma, and FOXP3 decrease (27–29). The latter suggests the role of VISTA in regulating the tumoral immune response; its upregulation is therefore an independent marker of poor survival.

Since anti-PD1 (nivolumab, pembrolizumab) and anti-CTLA-4 (ipilimumab) have been approved by the U.S. Food and Drug Administration (FDA), and since VISTA was found to be correlated with these two immune checkpoint inhibitors ( Figure 5A ). Consistent with the findings of Böger et al., we decided to study their expression in patients with UM to further support our findings, since mRNA expression analysis solely cannot predict protein levels (14).

PD1, also known as CD279, is a transmembrane glycoprotein expressed in activated T and B lymphocytes, NK cells, and monocytes. It contains two tyrosine kinase domains in its cytoplasmic tail (37). Once activated by binding to either PD-L1 or PDL-2 ligands, the TCR and BCR signaling pathways are blocked (38).

CTLA-4, also known as CD152, is a membrane glycoprotein that binds to ligands of the B7 family (CD80 and CD86) on the surface of APCs, and is highly similar to CD28, which is a stimulatory checkpoint molecule. By binding to its ligand, CTLA-4 suppresses signaling pathways in T cells, leading to T cell anergy, fatigue, and a diminished T cell immunological response (39). CTLA-4 might suppress T-cell-mediated antitumor immune responses by attenuating tumor-specific T-cell activation before these T-cells eradicate the tumor. Its blockade was thought to increase T cell-mediated antitumor immunity by eliminating this inhibitory signal (40).

Numerous clinical trials have demonstrated the function of VISTA in the inhibition of T cell responses and its overexpression after anti-PD1 therapy. Otherwise, the response of melanoma patients to immunotherapy induces resistance to anti-PD1 and anti-CTLA-4 treatment (41, 42). Increased expression of VISTA was observed not only after treatment with anti-PD1 alone but also in combination with CTLA-4, proving the activation of VISTA pathway after these treatments and explaining their non-efficacy in various aggressive cancers (41).

PD1 protein expression was astonishingly not noticed in any of our samples. To emphasize, our UM sections did not show any expression of PD1 protein ( Figures 5D, E ). This suggests that VISTA and PD1 pathways are non-redundant in the cancer immune response, consistent with the results reported by Liu et al. (43). Similarly, it has been shown that UM undergoes decreased PD1 and PD-L1 protein expression levels in patients compared with other tumors such as cutaneous melanoma (44). In contrast, Jiang et al. observed PD1 expression in 50% of primary UM cases, and higher expression of PD1 in tumor cells was linked to progression of UM cells and lower patient survival (45). However, CTLA-4 protein was detected only in inflammatory cells, specifically in monocytes and T cells, which are mainly TILs, and was noticed in higher stages of UM ( Figure 5I ). Similar to anti-PD1 agents, treatment with anti-CTLA-4 antibodies showed discouraging results in patients with UM, with a minimal response rate that did not exceed 10% (46).

As the presence of PD1 is required for PD1/PD-L1 interaction, its expression was found to vary among studies. The prognostic value of the PD1/PD-L1 pathway as a regulator of T cell activation in UM was assessed. An in vitro study using RT-PCR and flow cytometry analysis revealed that PD-L1 was constitutively expressed in five of nine primary UM cell lines. However, immunohistochemical analysis showed that PD-L1 protein was not expressed in patients with primary UM and was negatively expressed in all cases. However, its expression has been observed in patients with metastatic UM (47). This may explain that PD-L1 expression was influenced by the eye microenvironment. Another study demonstrated that almost half of patients with primary UM express PD-L1 protein (48). The mRNA expression in the same study analyzing two different cohorts established that the first cohort had a favorable prognosis and decreased infiltration of TILs, whereas the second cohort showed no significance in terms of overall survival. Our data suggest that the expression of PD1 and PD-L1 may be specific to each tumor type; in particular, it may depend on the tumor’s molecular phenotype. Although PD1 and PD-L1 were found to be expressed in various studies, their expression was very low in UM compared to other types of cancers. This low expression may explain the non-effectiveness of anti-PD1 therapies tested to date in the treatment of patients with UM, suggesting that it may be less effective in treating this type of tumor.

The key question of this study is why the inflammatory cells present in our UM sections do not present any PD1 staining?

Positive significant correlation was found between the average mRNA and the average protein expression (49); however, we here found that PD1 protein is different from that of mRNA. Coupled with our results, different assumptions can be made. Since PD1 is expressed in TILS, and our samples had lower expression of TILs infiltration, this may explain the non-expression of PD1. Based on PD-L1 and TILs expression, Teng et al. classified the TME into four forms. PD-L1+/TILs+; PD-L1+/TILs-; PD-L1^-/TILs+ and PD-L1^-/TILs-. The last study indicated the implication of other regulators of the anti-tumor immune response (50).

The prognosis of TILs in UM is a matter of reflection from various studies and reviews. Indeed, their presence may predict the response to treatment, and is linked to a better outcome in several types of cancers. The tumor mutational burden was assumed to be reflected by the infiltration levels of TILs. Singh et al. established that TILs infiltration is related to worse prognosis in UM, suggesting that patients with decreased infiltration of TILs may benefit from the effectiveness of immunotherapy (21). It is not to forget that PD1 is not only expressed on T cells, but also on myeloid cells, thymocytes, and on B cells (38). Therefore, eliminating the fact that the absence of T cells may justify the absence of PD1 staining, the mRNA expression of PD1 was already high in patients with UM in TCGA cohort. This may be due to multiple biological processes and genetic mechanisms, such as post-transcriptional modifications, that may explain the lack of protein expression. The limited sample size included in our study might also have influenced this expression.

Furthermore, while seeking to elucidate PD1 and CTLA-4 survival analysis, we noticed that both high PD1 and CTLA-4 expression decreased patient survival. Although statistically significant, they did not show beneficial outcomes in clinical trials for primary or metastatic UM patients. In addition to the low response rates in retrospective data, as previously stated, treatment with anti-CTLA-4 antibodies, especially Ipilimumab and Tremelimumab, may have many immune-related adverse events and unwanted side effects. A meta-analysis of 81 fully reviewed articles reported that most immune-related adverse events of anti-CTLA-4 treatment occur as skin lesions (dermatitis, epidermal spongiosis, and Sweet’s syndrome), hormonal deficiencies, hepatitis, colitis, pancreatic abnormalities, neurologic complications, ocular diseases, visual disturbances, and severe immune complications (51).

To go more in depth in our study, T cells gene expression profile, their infiltration, and their association with VISTA expression were studied; since many studies have reported that UM microenvironment is infiltrated by CD3, CD4 and CD8 lymphocytes (52–55). According to TCGA database, strong infiltration of CD8+ T cells, CD4+ T cells and regulatory T cells (FoxP3) is noticed in the tumor microenvironment of UM patients expressing high levels of VISTA. Besides, regulatory cytokines like IL-10 and TGFB, play an important role in modulating the immune system, as IL-10 regulates T cell proliferation (56, 57). In line with our findings, they were found to be overly expressed in UM microenvironment, creating an immunosuppressive microenvironment. Suggesting that VISTA may play a compelling role in creating an immunosuppressive microenvironment as well as in the induction of regulatory T cells overexpression.

In order to study the functional role of CD8+ T cells infiltrating our microenvironment, we attempted to characterize their phenotype among UM patients. As reported by Sun et al., persistent carcinogenesis may lead to a T cells phenotype termed “exhausted”; the presence of strong infiltration of these exhausted CD8 + T cells induce the promotion of immune evasion in UM (58–61). Additionally, Chen and Mallmen identified EOMES ^+/-, TBET ^+/- and PD1 ^med as a phenotype for recoverable and exhausted effector CD8+ T cells markers (62). These data are consistent with our findings that report high expression of EOMES, TBET, and PD1 in the presence of high VISTA. EOMES deletion was found to impact the exhaustion process, confirming that it is associated with the terminal differentiation of exhausted CD8+ T cells (63). Here we suggest the potential role of VISTA in T cell exhaustion in UM microenvironment; it may therefore be considered as a marker of exhaustion in UM. Surprisingly, patients with higher VISTA gene-level RNA-seq expression showed the worst survival, which was dramatically decreasing. However, lower VISTA mRNA expression was associated with better outcomes (P= 0,0003) ( Figure 7 ). Comparatively, promising results were obtained when blocking VISTA in murine models of various tumors (16, 19). Taken together, we hypothesized that VISTA expression may play a pivotal role in the survival of patients with UM. This might explain the failure of anti-PD1 therapy and anti-CTLA-4 inefficiency in clinical trials of patients with UM. Our study suggests an alternative pathway by which malignant cells may escape the immune response.

Our findings show that VISTA may be a possible immune checkpoint in patients with malignant UM. Its blockade might be a better potential treatment among different immune checkpoint inhibitors that have previously failed in patients with UM. Evidently, other studies might be necessary to determine whether VISTA may be combined with other immune checkpoint inhibitors to achieve an optimal strategy leading to long-term survival of patients with UM.

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 humans were approved by Institutional Ethics Committee of Biomedical Research of Casablanca (N°04/21), Faculty of Medicine and Pharmacy, Casablanca. The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study. Ethical approval was not required for the study involving animals in accordance with the local legislation and institutional requirements because animals were not used in this study.

Conceptualization: NIIS and AB; Data curation: NIIS, FM, MK, and MEB; Formal analysis: NIIS; Funding acquisition: AL and AB; Investigation: FM, MEB, and AB; Methodology: NIIS, and AB; Project administration: MEB and AB; Resources: NIIS, FM, MK, and MEB; Software: NIIS and AB; Supervision: FM, MEB, and AB; Validation: FM, MEB, and AB; Writing original draft: NIIS; Writing review & editing: FM, MEB, and AB. All authors contributed to the article and approved the submitted version.