OX40/OX40 ligand and its role in precision immune oncology

Upregulation of immune checkpoints such as programmed cell death protein-1 (PD-1) and cytotoxic T lymphocyte-associated antigen-4 (CTLA-4) results in negative regulation of T-cell activation [1]. Inhibition of immune checkpoints with anti-PD-1 and anti-CTLA-4 antibodies is associated with anti-tumor response. However, benefit is limited to a subset of the patients, highlighting the need for identification of other signaling mechanism to harness anti-tumor activity by immune cells.

T-cell co-stimulatory receptors that belong to the tumor necrosis factor receptor superfamily (TNFRSF), including OX40 receptor (CD134; TNFRSF4), 4-1BB (CD137; TNFRSF9), and glucocorticoid-induced TNFR-related (GITR) protein (CD357; TNFRSF18), are potential targets for cancer immunotherapy [2, 3]. OX40 and its binding ligand OX40L (CD134L; TNFSF4; CD252) are novel immune therapeutic targets that augment the immune response. The gene for OX40 is located on chromosome 1p36 [4, 5]. OX40 expression usually peaks around 24–72 h after antigenic stimulation of the T cell receptor (TCR) by the major histocompatibility complex (MHC) [6‐8]. Binding of OX40 to OX40L induces signal transduction pathways to activate immune response and regulate T-cell activation, proliferation, differentiation, expansion, and survival. Evidence also suggests a role of OX40 in Th1 and Th2 response [9, 10].

OX40 and OX40L overexpression in CD4 + T cells also has a role in the pathogenesis of multiple autoimmune diseases [11, 12]. Graft-versus-host disease (GVHD) is a common complication of allogeneic hematopoietic stem cell transplantation and is associated with significant morbidity and mortality; a preclinical study showed OX40 and O40L interaction correlated with the induction and progression of acute GVHD [13]. Studies also suggest the potential role of the co-stimulatory OX40 signaling pathway in anti-viral immune response [14, 15].

In this article, we review the impact of the co-stimulatory molecules OX40 and OX40L, including their landscape in cancers and ongoing clinical trials.

Figure 1 illustrates the OX40-OX40L signaling pathway. The OX40 is a type 1 transmembrane glycoprotein mainly expressed by T lymphocytes. The costimulatory molecule OX40 (CD134) receptor and its ligand OX40L (CD 134L/CD252) belong to tumor necrosis factor superfamily [5]. The naïve T cell does not have OX40 and OX40L expression; however, antigenic stimulation of T-cell receptor (TCR) via the major histocompatibility complex (MHC) leads to T-lymphocyte activation, resulting in upregulation of OX40 expression [8]. Furthermore, activated T cells have upregulated CD28 expression, potentiating OX40 expression. In addition, interleukin (IL)-2 and IL-2 receptor (R) signaling is essential for satisfactory OX40 expression on activated T cells [16‐18]. An in vivo study by Verdeil et al. demonstrated increased OX40 expression in CD8 T cells with IL-2 via STAT5-mediated signaling in the setting of weak TCR stimulation [19]. In mouse models with CD4 T cells converted CD4-CD8- double negative T cells, IL-2 facilitated the upregulated expression of OX40, supporting the survival of double negative T cells [20]. In the nonalcoholic steatohepatitis mouse model, liver tissues showed overexpression of OX40 in CD4 T cells with exogenous IL-2 stimulation [21]. The OX40 receptor is expressed on activated T lymphocytes (CD4 + and CD8 +), activated natural killer cells, regulatory T cells, neutrophils, and dendritic cells. On the other hand, OX40L is primarily expressed by antigen-presenting cells (APCs) [22]. Interaction between OX40 with OX40L is known to have immunomodulatory functions on T cell survival and proliferation. The cytoplasmic domain of OX40 is involved in downstream signaling pathways by binding with the tumor necrosis factor receptor-associated factor family (TRAF) of intracellular proteins. TRAF 2, 3, and 5 proteins are implicated in signal transduction after OX40 stimulation, mediating activation of nuclear factor-κBs (NF-κB) pathway [23, 24]. Further activation of the IκB kinase (IKK) complex and Rel A/B upregulates anti-apoptotic genes such as Bcl-2 and Bcl-XL, increasing T cells’ survival and proliferation. The activation of phosphatidylinositol-3-kinase (PI3k) and protein kinase B (PKB [Akt]) induces anti-apoptotic proteins such as Bcl-2, Bcl-xl, Bfl-1, and survivin [6, 25]. OX40 signaling also reduces expression of Forkhead box protein-3 (FOXP3) and cytotoxic T lymphocyte-associated protein (CTLA-4), contributing to decreased function of the immunosuppressive regulatory T cells [26, 27]. However, the molecular mechanism for the expression of FOXP3 and induction of Treg cells is quite complex and includes TCR signaling, cytokine milieu, transcription factors (Foxo1, STAT5, SMAD3, RUNX, NF-κB, BATF3, BATF), immune checkpoints (CTLA-4, PD-1), and costimulatory molecules [28‐30]. Activation of the OX40 costimulatory receptor prevented the induction of naïve CD4 T cells to CD25 + FOX3 + T cells by inhibiting TGF-β signaling and increased cytokine production [interferon-gamma (IFN‐γ), IL-4, and IL-6] [31]. In contrast, Ruby et al. demonstrated enhanced Treg conversion with OX40 agonists by blocking cytokines, namely IFN‐γ and IL-4, in the mouse model [32]. Notably, the OX40 agonist did not affect Treg function in the in vitro and in vivo study model; it enhanced IL-2 production by CD4 + T cells, promoting increased Treg proliferation [33].

OX40 is expressed on tumor-infiltrating lymphocytes (TILs) in various malignancies such as ovarian, head and neck, non-small cell lung (NSCLC), breast, colorectal, hepatocellular, and gastric cancer [34]. Studies have shown conflicting results with OX40 expression in TILs regarding clinical relevance and prognosis. Low OX40 TIL expression in tumor samples from NSCLC patients (n = 139) was associated with longer overall survival and better prognosis; however, the study did not specify the subtype of T lymphocytes (effector CD4 + /CD8 + T cells or CD4 + regulatory T cells) with OX40 expression [35]. The authors also found a negative correlation between PD-1 expression and TILs OX40 and OX40L expression [35]. In another study, high OX40 in tumor immune infiltrate was found to have a favorable prognosis in patients with stage I-III NSCLC (n = 100), but authors did not specify the subset of T cells with high OX40 expression [36]. Similarly, in patients with stage I-III colorectal cancer (n = 50), high OX40 expression on CD8 + T lymphocytes showed better overall survival/favorable prognosis [37]. Interestingly, advanced colorectal cancer (n = 22) patients with high blood levels of soluble OX40 had a worse prognosis as compared to low level of soluble OX40 [38]. In ovarian cancer, high OX40 expression on tumor-infiltrating immune cells (authors did not specify subset of T lymphocytes) correlated with longer recurrence-free survival and better response to chemotherapy [39]. Immunohistochemistry-based high expression of OX40 on breast epithelial cells and cancer cells in surgically resected specimens in patients with ductal carcinoma in situ and invasive ductal carcinoma of the breast were associated with the clinically aggressive disease; however, no differences were observed with high expression of OX40L [40]. Interestingly, high expression of OX40L in platelets from breast cancer patients was associated with high tumor grade, immune cell activation, and the tendency for metastases [41]. The tumor microenvironment (TME) in hepatocellular cancer (HCC) patients with high OX40 expression in regulatory T cells (Treg) was associated with poor survival and high serum alfa-fetoprotein level [42]. Besides, high expression of LAG3, PD-1, TIM-3, CD8, and CD68 were correlated with high OX40 expression [42]. Increased expression of OX40 on Treg cells was associated with disease progression in patients with cutaneous squamous cell carcinoma [43]. Treg cells in TME of the head and neck cancer patients demonstrated high expression of OX40, PD-1, and CTLA-4 in one of the studies but no clinical outcome was reported [44]. In patients with acute myeloid leukemia (AML), high expression of OX40 on blast cells was correlated with shorter overall survival and progression-free survival, highlighting its potential role as an immune prognostic marker, whereas no association was found with OX40L expression [45]. Additionally, high RNA expression of the OX40 gene/TNFRSF4 gene from the TCGA database in AML patients was associated with TP53, FLT3, and NPM1 mutation and unfavorable clinical outcome [46]. Taken together, the data demonstrate that expression of the OX40 machinery may correlate with either better or worse prognosis, depending on the cancer/setting studied; moreover, some studies show conflicting results. This could be potentially explained by fact that some of the studies did not specify the type of T lymphocytes with OX40 expression. Studies that specifically evaluated the Treg OX40 expression in the TME were found to be associated with poor prognosis; hence, high OX40 expression in Treg cells appears to correlate with worse clinical outcomes. Subtyping of T lymphocytes with OX40 expression is crucial as Treg plays a key role in TME in immune suppression and facilitate tumor progression [47].

Early-phase clinical trials of OX40 agonists showed anti-tumor activity in advanced solid malignancies; however, the response rate was low with single-agent as well with combination treatment. The possible reasons for low response rates include limitations of the compounds administered, accrual of patients without known OX40 expression, dysregulation of other checkpoints that could potentially attenuate efficacy, inadequate signaling of OX40 downstream pathway, and an immune suppressive microenvironment with upregulated Treg cells.

Selecting patients based on high OX40 and low OX40 ligand expression in T cells in the tumor microenvironment may warrant exploration when targeting OX40 and OX40L, as different tumors have variable levels of expression of OX40-related machinery in T cells. Various methods are used to evaluate OX40 and OX40L expression, such as immunohistochemistry and reverse transcription-polymerase chain reaction/mRNA expression, with the possibility of heterogenous results in OX40/OX40L positivity in TILs.

To evaluate the OX40/OX40L expression across diverse solid malignancies, we performed a comprehensive analysis of OX40/OX40L expression across various solid tumor types in 514 patients diagnosed with advanced malignancy at the Moore Cancer Center at the University of San Diego (Supplemental Table 1). The percentile of the OX40/OX40L expression was based on transcript level in each patient which was ranked on a scale of 0–100, and classified as low (0–24), moderate (25–74), and high (75–100) and was normalized to 735 control tumors as previously described [76‐78].

The percentage of high OX40 expression (≥ 75th percentile RNA rank) was 23% (118/514) across all tumor types. OX40 high expression was variable between and within cancer types; lung cancer 30% (6/20), pancreatic cancer 27% (15/55), and colorectal cancer 22% (31/140), and breast cancer 22% (11/49). For descriptive purposes (Fig. 2), we sought to assess the pattern of OX40 and OX40L expression in various tumor types such as OX40 high plus OX40L low-moderate expression, OX40 low-moderate plus OX40L high, OX40 high plus OX40L high, and OX40 low-moderate plus OX40L low-moderate. The OX40 high plus OX40L low-moderate expression pattern was observed in 17% (87/514) of all cancer types; it might be reasonable to assume that this expression pattern could be most amenable to OX40 agonist activity. This expression pattern was most common in patients with stomach (36% of patients) and lung cancer (30%) (Fig. 2).

OX40, also known as CD134 or TNFRSF4, is a cell surface receptor, which is a member of the tumor necrosis factor receptor superfamily; it is found on the surface of activated T-cells, and it serves as co-stimulatory immune molecule. When T cells encounter antigens presented by pathogens or tumor cells, they become activated; OX40 is upregulated on these T cells and binds to its ligand [OX40L (CD134L; TNFSF4; CD252)], which is typically found on antigen-presenting cells, including but not limited to dendritic and B cells. The interaction between OX40 and OX40L promotes T-cell survival, proliferation, production of immunostimulatory cytokines, and maintenance of memory CD8 + T cells. Modifying the OX40-OX40L interaction can enhance the immune response to fight cancer or dampen it to treat autoimmune conditions [79]. Many experimental drugs that serve as OX40 agonists are in cancer clinical trials. OX40 agonists are being explored as monotherapy and in combination with other immunotherapy agents for cancer treatment. To date, however, while responses have been observed, they remain isolated to a minority of patients. PDL1 expression, high tumor mutational burden (≥ 10 mutations/megabase), and high microsatellite instability (MSI-H) are well known biomarkers that are associated with improved outcomes with immunotherapy [80‐84]. PD1 expression on tumor-infiltrating lymphocytes has also been recently shown to correlate with better outcome after anti-PD1/PDL1 agents are given to patients with cancer [85]. Results from a recent meta-analysis of clinical trials (over 19,000 patients) found that most immune-oncology clinical studies did not include biomarkers for patient selection, even though retrospective analysis showed that biomarkers were independently correlated with improved immunotherapy outcome [86]. Therefore, the role of OX40 and OX40L expression and its association with expression of other checkpoints and biomarkers such as TMB, MSS, PD1, PDL1, and LAG3 warrants exploration in the context of determining if patient selection for monotherapy OX40 agonists and for combination therapy with specific checkpoint inhibitors would enhance response rates and other outcome parameters. The role of OX40 might also be complicated as OX40 expression on Tregs may be immunosuppressive.

Transcriptomic profiling reveals that OX40 and OX40L expression is variable between tumors and that the pattern of high OX40 and low OX40L, which might theoretically be most amenable to OX40 agonist compounds, occurs in only 17% of cancer patients, most commonly in lung and breast cancers. The correlation between OX40 protein expression and RNA expression however remains unclear and merits investigation in future studies. A precision immune oncology approach, interrogating individual tumor expression patterns of OX40 and OX40L and other immunomodulatory effectors, may warrant exploration in future studies of both single-agent and combination regimens with OX40 agonists.