Overview of CD70 as a Potential Therapeutic Target for Osteosarcoma

Osteosarcoma is the most common primary malignant bone tumor among pediatric patients. The metaphysis of long extremity bones such as the distal femur, proximal humerus, and proximal tibia are common primary sites for osteosarcoma. Metastases most frequently occur in the lungs and occasionally elsewhere in a bone other than where the primary tumor originated or in lymph nodes (1). When osteosarcoma is metastatic, the 5-y overall survival in patients significantly decreases from 60–70% to ∼20% (2). Given the prognostic implications of metastatic disease, many clinical trials and novel therapies have recently been introduced to specifically target metastases. However, treatment options for relapsed/refractory osteosarcoma are still limited. Immunotherapy has shown promise in various tumor types (3); yet, its efficacy in solid tumors remains limited (4, 5). Immune cell therapies to treat osteosarcoma continue to be investigated; several clinical trials are evaluating the effect of chimeric Ag receptor (CAR) T cells targeting known cancer Ags (6) in osteosarcoma. However, the limited number of known tumor Ags in osteosarcoma (7) has restricted the number of trials applicable to this disease. Therefore, further identification of new targets may assist in the development of novel and more effective therapies.

Cluster of differentiation 70 (CD70) has been suggested as a potential target for osteosarcoma because it has limited to no expression in normal tissues, including vital organs, and its expression tends to be increased or atypical during cellular transformation in various cancer types (8). CD70 is a type II transmembrane protein that is a member of the TNF receptor family. It is known to be expressed on highly activated B/T lymphocytes, mature dendritic cells, and APCs in the intestinal mucosa. In the normal immune environment, its expression is very tightly regulated (9, 10). CD70 has also been found to be overexpressed in several malignancies, including osteosarcoma (9). It is the only known ligand for CD27, and when the two bind, they play a role in the proliferation of B and T lymphocytes (10–12). In particular, when CD70 and CD27 interact, CD4⁺ and CD8⁺ T cells proliferate and produce cytokines. Another result of CD70/CD27 interaction is B-cell differentiation, which leads to the activation of B cells, plasma cell differentiation, and germinal center formation. Hence, the overexpression and subsequent dysregulation of the CD70/CD27 pathway can lead to T-cell exhaustion (13) and potentially increase growth signaling to tumor cells.

Despite limited understanding of the CD70/CD27 pathway as well as the consequences of targeting CD70, existent preclinical information has led to the development of CD70-targeted immunotherapies for different cancer types, including osteosarcoma, that are currently under investigation. CD70 is known to be overexpressed in lymphomas, most notably on the Reed-Sternberg cells of Hodgkin lymphoma, but also in other B-cell malignancies (9) where its expression potentially promotes the proliferation of B-cell leukemia and lymphoma (14). In liquid tumors, both CD70 and CD27 are expressed on blast cells. Specifically in acute myeloid leukemia, higher levels of soluble CD27 on circulating blast cells have been associated with a poor prognosis, and the coexpression of CD27/CD70 was found to increase proliferation of tumor cells (15). CD70 is also expressed on solid tumors, most notably renal cell carcinoma, glioblastoma, astrocytoma, osteosarcoma, nasopharyngeal carcinomas, and thymic carcinoma (9). In osteosarcoma in particular, the expression of CD70 is higher in metastatic and recurrent tumors than in the primary bone tumors, and this expression has a heterogeneous pattern. However, CD27 does not seem to be present on osteosarcoma cells (16). Therefore, CD70 holds promise as a potential target for osteosarcoma because the available treatments for metastatic and recurrent osteosarcoma are scarce, and, compared with primary site tumors, CD70 expression is highest in the lung metastases, the leading cause of death in these patients. In this review, we discuss what is known about CD70 in osteosarcoma, current therapies targeting CD70, and their potential impact in osteosarcoma.

CD70 expression is variable between different tumor tissues and cell lines. Many methods for evaluating CD70 expression have been described previously (14, 17–20), including flow cytometry, immunohistochemistry, Western blot analysis, RT-PCR, and mass spectroscopy. Although flow cytometry and mass spectroscopy are useful in quantifying CD70 expression on cells in culture, immunohistochemistry identifies CD70 expression throughout tissues and is the most frequently used method (21). These methods seem particularly useful in osteosarcoma because CD70 expression has been shown to be heterogeneous throughout tumor tissues and varies within different cell lines (16).

Molecular sequencing, such as whole-genome/exon sequencing and RNA sequencing of tumor tissues, allows determination of the CD70 gene expression profile. For instance, analysis of publicly available data sets (GSE126209, GSE11414, GSE12865, GSE14359, GSE14827, GSE21257, GSE42352, and Aqeilan’s data set) demonstrated increased CD70 mRNA expression in patients with osteosarcoma and cell lines (22, 23) (Fig. 1A, 1B).

All methods are useful. However, it is important to recognize that CD70 expression analysis needs to be extended to include surrounding normal tissue to account for the potential off-target effect of therapy. Previous studies demonstrated negative expression of CD70 as determined by immunohistochemical staining in adjacent, normal lung tissues in a non–small cell lung cancer (NSCLC) model (24). Therefore, understanding the expression patterns of CD70 in individual tissues can help determine whether these expression levels and distribution influence therapeutic response or contribute to potential side effects.

Although CD70 is variably expressed among different malignancies, it is tightly regulated in immune cells, and it has been studied in many normal tissues to ensure targeting CD70 will not lead to unnecessary toxicities (10). It has been hypothesized that the overexpression of CD70 is a way for tumors to evade the immune system via T-cell exhaustion or lymphocyte apoptosis (25). Therefore, direct tumor targeting of CD70 seems to be a rational approach.

It is yet to be determined whether normal bone or lung cells express CD70. Therefore, one of the key elements when considering CD70 direct tumor targeting in osteosarcoma is understanding the expression level in normal bone and lung tissue to predict or avoid potential off-tumor effects. However, on the basis of previous publications, CD70 appears to be a promising target in recurrent and metastatic osteosarcoma, particularly because it is found to have higher expression in lung metastases and there is a pressing need for more effective therapies to treat osteosarcoma metastatic disease in the lungs.

CD70 is also known to play a role in the metastatic potential and immune system evasion by tumors, as outlined in Fig. 2. Immune evasion is aided by the CD70-mediated activation of regulatory T lymphocytes (Tregs) that express CD27 and are activated when bound to CD70. Tumors with high CD70 expression have increased forkhead box P3 (FoxP3)-activated intratumoral Tregs (26). Because Tregs expressing CD27 have an immunosuppressive function in the tumor microenvironment, strategies to target or block the CD70/CD27 axis are being developed. Additionally, in osteosarcoma, the poor prognostic marker phospholipase Cε1 appears to mediate immune escape via the CD70/CD27 signaling pathway, suggesting the potential benefit of targeting this pathway to improve the immune response (27).

In addition to modulating the tumor microenvironment to increase immune evasion, CD70 has also been associated with resistance to a traditional chemotherapeutic agent, cisplatin. In ovarian carcinoma, high CD70 expression was associated with a higher incidence of cisplatin resistance and also correlated with a worse prognosis (28). This is remarkable because cisplatin is part of the standard-of-care treatment for patients with osteosarcoma, highlighting the potential benefit that targeting CD70 may offer to these patients.

Although CD70 is tightly regulated among immune cells, it is expressed on activated T and B cells (29), and there is a possibility that these normal immune cells could be inadvertently targeted. Tuning the CD70/CD27 interaction could also affect the functionality of the innate immune system (30). Upregulated CD70 expression has been associated with T-cell exhaustion of both memory and effector T cells in non-Hodgkin lymphoma (31). Although it is important to establish whether CD70 expression is found on normal human osteoblasts and pneumocytes to account for potential side effects, it will also be important to consider the possibility that CD70-targeted therapy will have limited therapeutic efficacy because immune cells, like T cells and NK cells, could also be affected. Additionally, the CD70 expression pattern within tumor cells and tissues may influence the therapeutic response. For instance, in osteosarcoma, studies demonstrated that the expression of CD70 varied between several human osteosarcoma cell lines and tumor tissues (16). However, whether heterogeneity in expression may influence response to CD70-targeted therapy is not known. Also, because certain chemotherapeutic agents have the potential to alter CD70 expression, cytotoxicity might be affected. CD70 expression has also been correlated with increased tumor metastatic potential (29); whether this would be the case for osteosarcoma is unknown. Additionally, CD70 has reverse signaling properties in B cells (32), and it has been reported that treatment with an anti-CD70 mAb stimulates B-cell chronic lymphocytic leukemia cell growth (14). However, CD70 reverse signaling has been suggested to be involved in immunosurveillance of B-cell malignancies because it increased NK cell survival and effector function in a B-cell acute lymphoblastic leukemia model (33). Reverse signaling of CD70 in solid tumors is not yet well defined.

The correlation between CD70 expression and chemoresistance has been studied with conflicting results in separate studies and databases (34, 35), and this variability highlights the challenge of targeting a heterogeneous tumor with possible differences in tumor epigenetic profiles. Last, as previously demonstrated, the presence of a capsule that surrounds solid tumors may limit the cytolytic effect of CD70-targeted therapies due to the inability to penetrate the tumor (36). This may pose a challenge in osteosarcoma. In summary, even though identification and validation of the target is critical, it is crucial that targeted therapies be devoid of toxicities and that inconsistencies in response be accounted for based on the variable nature of the target.

CAR T cells targeting CD70 have been generated and tested in vitro against many different CD70-expressing tumor types. In one study, using a single-chain variable fragment containing an anti-CD70 Ab to create a CD70-directed CAR T cell, effective lysis was seen on different solid tumor targets with a range of CD70 expression (37). CD70 is expressed on activated immune cells, including T cells. CAR T constructs must take this into account to avoid overt fratricide and enable the CAR T cells to have maximally effective cytotoxicity (38). In a separate preclinical evaluation of CAR T cells targeting CD70, researchers found a significant improvement in T-cell specific reactivity in a renal tumor mouse model when a 41BB costimulatory domain was used compared with a CAR with a CD27 costimulatory domain (39). A similar study found using CD27 fused to the signaling domain of the TCR cell ζ produced a CD70-directed CAR T cell capable of inducing cell lysis of CD70-positive leukemia and lymphoma cells (40). These CD27-derived CD70 CAR T cells also reduced tumor burden and increased survival in lymphoma mouse models. Bivalent CAR T cells targeting both CD70 and B7-H3 tumor Ags on solid tumors had increased antitumor effects compared with unispecific CAR T cells (41). Finally, CD70-targeted CAR T cells have also been shown to improve survival in glioblastoma mouse models (10, 42). This is encouraging for potential future use in osteosarcoma.

Anti-CD70 Abs have also been developed and studied in different tumor types. A novel CD70 Ab was generated against the CD70-positive B-cell lymphoma cell line, BJAB, which was able to effectively block the CD70–CD27 interaction in these cells (14). Preincubation with B-cell chronic lymphocytic leukemia cells with this Ab also inhibited proliferation. Another anti-CD70 mAb generated from immunized llamas germlined to humans, ARGX-110 (cusatuzumab), demonstrated antitumor activity against CD70-positive hematologic tumor cell lines as determined by complement-dependent cytotoxicity (43). The encouraging results in vitro led to in vivo studies using an Epstein-Barr virus–positive Burkitt lymphoma mouse model where animals treated with ARGX-110 revealed a correlation between reduced soluble CD27 levels and improved survival.

Targeting the pathway also offers additional challenges. For instance, in solid tumor mouse models, it was confirmed that the CD27–CD70 interaction increases Treg proliferation, which reduces the tumor-specific T-cell response. However, the Treg frequency and growth of the tumors were found to be reduced in both CD27-deficient mice and mice treated with an anti-CD70 Ab (44), providing further evidence that targeting the CD27/CD70 pathway is a promising therapeutic approach in solid tumors, including osteosarcoma. In contrast, when the CD70/CD27 pathway was modulated in precursor B cell acute lymphoblastic leukemia, inhibition of CD70 reduced the cytotoxic capacity of antileukemic T cells (45). Therefore, understanding the direct effects of targeting CD70 on different tumor types is essential before these novel therapies are used clinically.

Ab-drug conjugates (ADCs) provide a mechanism to target selective tumor cells and enact cytotoxic effects. These have been tested in CD70-positive tumors using an anti-CD70 Ab conjugated to the antineoplastic agent monomethyl auristatin F (46). The resulting ADC, SGN-75, has been shown to successfully accumulate preferentially in the tumor as opposed to normal tissue. It was also confirmed to inhibit the growth of carcinomas and was further studied in clinical trials of CD70-positive tumors (47–49). Anti-CD70 ADCs as well as CD70-targeted CAR T cells and NK cells have also been efficacious against NSCLC cells with epidermal growth factor receptor mutations and acquired resistance to tyrosine kinase inhibitors (50).

Although anti-CD70 therapeutic options remain promising as monotherapies, with the recent findings of the immunomodulatory effects of certain chemotherapies, it has become evident that combinations of CD70-targeted agents with traditional chemotherapeutic agents could show benefit in a broader range of tumors (51). Preclinical studies targeting CD70 in combination with other therapies continue to show promise in the treatment of many tumor types. The combination of an anti-CD70 Ab with cisplatin resulted in greater NK cell–mediated cytotoxicity in NSCLC cells (52). Targeting both CD70 and programmed cell death receptor 1 (PD-1) together may also improve therapeutic efficacy because emerging data show both CD70 and PD-1 to have a role in solid tumor immune evasion (53, 54). In addition to chemotherapy and immunotherapy, it has been suggested that because radiation therapy can modulate CD70 expression in gliosarcoma, radiation with CD70-targeted agents could serve as novel, potential combination therapies (55). Finally, combining CD70 CAR-T cells with a poly(ADP-ribose) polymerase inhibitor was efficacious against renal cell carcinoma through activation of the cyclic GMP–AMP synthase/STING pathway (56).

Last, it seems important to recognize that chemotherapeutic agents could potentially change the tumor microenvironment and make it more susceptible to immune cell or Ab tumor targeting. One preclinical study compared the combination of cisplatin therapy with an anti-CD70 therapy in NSCLC (52) and demonstrated not only that cisplatin increased CD70 expression in the tumor cells but also that the combination of cisplatin treatment and anti-CD70 therapy was efficacious in all of the tested lung cancer cell lines in vitro. Thus, for an optimal treatment strategy, because cisplatin, like in NSCLC, is one of the first-line treatments in osteosarcoma, it would be important to understand the effects of cisplatin on CD70 expression in the tumor cells.

CD70 has been used as a relevant target to treat many tumor types, ranging from hematologic malignancies to refractory solid tumors. The aforementioned germline anti-CD70 mAb ARGX-110 (cusatuzumab) was studied in a phase I/II dose escalation study. ARGX-110 blocks the CD70/CD27 axis, which decreases the activated Tregs in the microenvironment and thereby stops the tumor’s immune system evasion. ARGX-110 kills tumor cells via its modified Fc region, which enhances Ab-dependent cell-mediated cytotoxicity (57). ARGX-110 was used in multiple CD70-positive neoplasms on the basis of previously described in vitro studies that showed a high-affinity binding of ARGX-110 to different CD70-positive cell lines and block of Treg activation (43). In the 26 patients with advanced stage CD70-positive malignancies studied, dose-limiting toxicity was not observed. Stable disease was found to be the best overall response, and the drug was relatively well tolerated, with the most common adverse effects being fatigue and infusion-related reactions (57).

SGN-70, an additional CD70-targeted agent, is an anti-CD70 humanized IgG1 mAb that showed in vitro and in vivo activity via Fc-mediated effector functions such as Ag-dependent cellular cytotoxicity, complement fixation, and opsonization. This was validated in CD70-positive carcinomas. Preclinical in vivo studies using SCID mice with multiple myeloma and disseminated lymphoma xenografts showed that SGN-70 decreased tumor burden and prolonged survival (8). The safety and tolerability of SGN-70 (referred to as SGN-CD70A in clinical trials) have been measured in patients with metastatic renal cell carcinoma and CD70-positive lymphomas (58, 59). The most common dose-limiting toxicity was thrombocytopenia, which prohibited further clinical investigation of SGN-70.

SGN-75, a CD70 ADC, was evaluated in a phase I dose-escalation study of patients with relapsed/refractory non-Hodgkin lymphoma and metastatic renal cell carcinoma. The dosing was adjusted for better tolerability, but there was targeted ablation of the CD70-positive lymphocytes. Antitumor activity was seen in both of these heavily pretreated patient populations. SGN-75 was generally well tolerated, and the patients with either a complete or a partial response had uniform CD70 expression within the tumor (48, 60), suggesting that the pattern of CD70 expression might constitute a key determinant of patient response to CD70-targeted therapies.

Clinical trials investigating anti-CD27 (CD70 ligand) therapies in combination with PD-1 receptor targets, tyrosine kinase inhibitors, and immune-modulating vaccines are underway in patients with renal cell carcinoma, glioblastoma, ovarian carcinoma, melanoma, and breast cancer (61). On the basis of this and the aforementioned preclinical studies, future clinical trials combining CD70-targeted agents with immunotherapy or chemotherapy could be useful for patients with osteosarcoma.

There is still information that has yet to be defined regarding the mechanism of action of CD70 and how this potentially affects the tumor cells, the immune cells, and the overall tumor microenvironment. It is known that certain normal human immune cells express CD70, and although their expression is transient and very well controlled in a normal immune environment, it is important to know if targeting CD70 could have an off-target effect on these cells (B, T, and mature dendritic cells), which could have an impact in immune surveillance. Furthermore, it is not clear whether CD27, the only known ligand for CD70 (30), is present or overexpressed in solid tumors, including osteosarcoma, and whether its expression could hamper the efficacy of a CD70-directed therapy. There is also limited information on the impact of CD70 expression levels on overall survival in many tumors, including osteosarcoma. Because CD70 expression is reported to be higher in the lung metastases of osteosarcoma than in primary bone tumors, understanding the correlation between CD70 expression and survival may help predict therapy outcome. It is also critical to understand how cytotoxic chemotherapy may alter the tumor microenvironment to make it more or less susceptible to CD70-targeted therapy and whether CD70 expression levels determine response. Therefore, future investigations are needed to determine the effect of standard-of-care chemotherapy such as cisplatin on CD70 expression in solid tumors and whether increased CD70 expression translates into better therapy response.

CD70 is expressed on many tumor types and plays a role in metastatic potential, immune system evasion, and resistance to chemotherapy (30). Therefore, it holds promise as a target for tumors with positive CD70 expression. CD70-targeted immunotherapies are beginning to emerge. CD70 CAR-T cells have been studied in vitro, and they were shown to recognize CD70-expressing glioblastoma tumors and mediate regression (10). Additional studies may focus on evaluating the efficacy of CD70 CAR-T cells on various other CD70-expressing tumors to further understand the therapy’s benefits and limitations in addition to the mechanisms of response/resistance.

CD70 expression has been correlated with survival in tumors such as ovarian and renal cell carcinoma. CD70 has also been shown to serve as a biomarker of progressive disease in renal cell carcinoma (9). However, there is still limited knowledge on the role CD70 plays in osteosarcoma. It is not known whether the level of CD70 expression in osteosarcoma correlates with prognosis. In a limited number of osteosarcoma samples from patients, the level of CD70 expression was higher in osteosarcoma metastatic disease in the lung than in primary disease in the bone (16). Overall, CD70-targeted therapy might have future beneficial implications for the treatment of patients with osteosarcoma. Additionally, the majority of patients eligible to receive targeted therapy have already been treated with multiple other cytotoxic therapies. Thus, knowing how these cytotoxic therapies influence the tumor microenvironment may aid our understanding of the potential differences in response to the CD70-directed therapies.

Treatment for metastatic osteosarcoma remains a challenge. Immune therapies have shown benefit in many tumors, but limited efficacy has been seen in osteosarcoma (62). Given the limited knowledge of the antigenic landscape of osteosarcoma, identification of CD70 as a tumor Ag expressed not only within the primary bone tumor but also in lung metastases offers an opportunity for the use of CD70-targeted immune cell therapies (CAR-T or CAR-NK) for the treatment of patients with metastatic osteosarcoma.

The authors have no financial conflicts of interest.

Fig. 2 was created using BioRender.com. We acknowledge Sarah Bronson and David Farris of the Research Medical Library and Scientific Publications at The University of Texas MD Anderson Cancer Center for contributions. We thank Kathy King for aid in preparing this article.