Immunomodulatory Antibody Therapy of Cancer: The Closer, the Better

In this issue of Clinical Cancer Research, Dai and colleagues (1) report that locally delivering combinations of immunomodulatory antibodies completely eradicates larger tumors in three murine tumor models. In the same issue, Mangsbo and colleagues (2) report that local injection of CD40 agonistic antibody to activate dendritic cells (DC) significantly suppresses tumor growth. Both preclinical observations suggest that a local route to deliver immunomodulatory antibodies with different mechanisms of action could be a better way to initiate, enhance, and maintain a strong local antitumor T-cell response. Both reports also demonstrate that intratumoral injection of immunomodulatory antibodies not only elicits rejection of local tumors but also results in systemic protective immunity against distant tumors or a rechallenge with the same tumors.

The search for an optimal route of delivery of cancer immunotherapy agents has been a research focus since Coley's first clinical attempt of injecting mixed bacterial toxin into primary tumor tissues. At that time, intratumoral injection was considered by Coley to be crucial to a patient's survival (3). Since then, local delivery of immunotherapy agents, such as vaccines and adjuvants, has been the main route of immunotherapy administration for decades. However, in the 1980s, the use of IL2 in patients with melanoma and renal tumors changed the landscape of cancer therapy. As researchers realized that a strong systemic antitumor immune response could eventually not only reject local tumors but also prevent tumor metastasis, intravenous injection of immunomodulatory agents remained a major route of cancer immunotherapy until now. However, some limitations of systemic delivery have been identified. For example, it is unknown to what extent systemically delivered therapeutic agents eventually accumulate at the tumor sites. This is especially important for immunomodulatory antibodies, because they not only restore antitumor T-cell responses at tumor sites but may also release brakes for anti-self T-cell responses in nonmalignant tissues or organs when they are delivered systemically. To reach a therapeutic threshold, usually a high dose and/or repeat delivery is required for systemic therapies, thus increasing the risk of adverse effects. Although personalized formulation of antibody therapy could help reduce the side effects and maximize therapeutic effects, an alternative route of delivery of cancer immunotherapy agents should be considered and evaluated.

Recently, studies analyzing immune responses within tumors and related regulatory mechanisms have provided new evidence encouraging researchers to consider the local or intratumoral delivery routes for cancer immunotherapies. Thompson and colleagues (4) reported that naïve CD8 T cells are primed within tumors by tumor cells or DCs and undergo differentiation to become effector CD8 T cells at tumor sites. Our group has observed an endogenous tumor-reactive CD8 T-cell response within tumors (5). We further showed that blocking the migration of lymphocytes from secondary lymphatic organs does not impair the accumulation of tumor-reactive CD8 T cells within tumor tissues, suggesting that resident naïve CD8 T cells are primed and expand locally at tumor sites (5).

Although local resident CD8 T cells are primed or activated at tumor sites (4, 5), they cannot reject tumors as they are subjected to local regulatory mechanisms. To promote local antitumor immune response, Dai and colleagues (1) and Mangsbo and colleagues (2) report that local delivery of immunomodulatory antibodies with nonredundant functions is superior to systemic delivery in cancer treatment (Fig. 1). In their studies, an agonist antibody to CD40 is used to activate DCs for inducing effector CD8 T-cell differentiation. CD137 (4-1BB) agonist antibody is used to activate the costimulatory function of CD137 on CD8 T cells, as CD137 promotes T-cell proliferation, function, and survival (6). Because upregulation of PD-1 limits the antitumor activity of endogenous, tumor-reactive CD8 T cells within tumors (5), anti–PD-1-blocking antibody was used to block the coinhibitory signaling of PD-1 and its ligands (B7-H1/PD-L1 or B7-DC/PD-L2) expressed by tumor cells. Anti–CTLA-4-blocking (or to some degree depleting) antibody was used to remove another barrier within tumors, i.e., regulatory T cells (Treg) that preferentially express CTLA-4 (7). The combination of these three antibodies has demonstrated promising therapeutic effects in a previous report from the same group (8). However, following triple-antibody therapy, the authors detected an increase of CD19⁺ cells in tumor-draining lymph nodes. Because regulatory B cells (Breg) express CD19 (9) and suppress Th1 cytokine production (10, 11), they tested whether an anti-CD19 antibody capable of depleting CD19⁺ Bregs would improve the efficacy of the other three antibodies. Indeed, local injection of anti–CD19-depleting antibody significantly improved the therapeutic effects of the multiple antibody therapy (1). Importantly, the local delivery of combinations of four mAbs induces a complete rejection of very large transplanted tumors. In addition, this local treatment causes a shift of a protumor Th2 to an antitumor Th1 profile and establishes a long-term protective immunity in the tumor-bearing host.

Intratumoral DCs could be another local target of immunomodulatory antibody therapy. A marked reduction of CD80 and CD86, two costimulatory molecules expressed by DCs, was found within tumors, and this correlated with intensive infiltration of Tregs (12). CD8 T cells activated by these CD80/CD86-low DCs express elevated levels of PD-1 and TIM-3, and are dysfunctional within tumors (12). To restore or enhance the impaired function of DCs, agonistic CD40 antibody has been utilized in several preclinical studies. In this issue, Mangsbo and colleagues (2) report a human agonistic antibody for CD40 in the treatment of human bladder tumors in a human CD40 transgenic mouse model. This new version of CD40 agonistic antibody is selected based on its high affinity to CD40 at both physiologic and low pH situations, given that most tumor milieu are acidic. Due to its high affinity to CD40, this new antibody has a low release rate from the injection site. This feature is very important for local use of an immunomodulatory antibody, as it may maximize therapeutic effects and minimize systemic side effects.

Although local or intratumoral administration of immunomodulatory antibodies is expected to increase the efficacy of therapy, it may not be practical for tumors located in deeper organs. Surgical implantation of slow-release agents or ultrasound-guided delivery could be considered as clinical options for intratumoral delivery of cancer immunotherapy agents. To avoid potential side effects, the retention of local injected antibody needs to be improved, along with the identification of target molecules expressed by local resident immune cells. In addition, a reliable way to monitor the effects of injected antibody on immune cells along with monitoring tumor progression is needed. To that end, analysis of downstream signaling pathways of immunomodulatory signals could provide helpful insights for evaluating the efficiency of therapeutic antibody. Nevertheless, the revival of the local route of delivery of cancer immunotherapy, as pioneered by Coley in the early 20^th century, could lead to better outcomes of anticancer therapy in the 21^st century.

No potential conflicts of interest were disclosed.

Conception and design: H. Dong

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): H. Dong

Writing, review, and/or revision of the manuscript: R.S. Dronca, H. Dong

R.S. Dronca was supported the National Center for Advancing Translational Sciences of the NIH under award number TR000136. H. Dong was supported by the NCI of the NIH under award number R01CA134345 and the National Institute of Allergy and Infectious Diseases of the NIH under award number R01AI095239.