Uncovering the Potential of CD40 Agonism to Enhance Immune Checkpoint Blockade

In this issue of Clinical Cancer Research, Weiss and colleagues present the results of a single-arm phase II trial evaluating the combination of the CD40 agonist sotigalimab and nivolumab in patients with advanced melanoma following confirmed disease progression on a PD-1 inhibitor (1). These results are intriguing given the unmet need in melanoma for the approximately 50% of patients who develop resistance to combination immune checkpoint blockade (2) and the mechanistic rationale for agonizing co-stimulatory receptors, especially on APCs, to reinvigorate suppressed antitumor immune responses.

Despite impressive single agent activity in preclinical murine models, drug development of agonistic mAb toward immune-stimulatory receptors has been complex. In 2006, the initial phase I study of a super-agonistic antibody (TGN1412) targeting CD28, a T-cell co-stimulatory receptor, led to massive cytokine storm and multiorgan failure in six healthy human volunteers (3). Around the same time in 2005, the humanized mAb (urelumab) targeting 4–1BB, another T-cell co-stimulatory receptor, was suspended due to fatal hepatotoxicity in 2 patients (4). Subsequent to that, multiple agonistic mAbs against targets such ICOS, OX40, GITR, and other co-stimulatory receptors were investigated alone and with other agents without a suggestion of robust clinical activity.

CD40 is a co-stimulatory receptor in the TNF superfamily that is mainly expressed on APCs such as dendritic cells (DC), macrophages, and B lymphocytes. APCs play an important role in the delivery of appropriate costimulatory signals that lead to activation of effector T cells. Binding of CD40 by CD40L (also known as CD154), activates multiple innate immune cells, and specifically promotes maturation, or “licensing” of DCs to enable them to express multiple other costimulatory receptors for optimal T-cell and general immune activation. Agonistic anti-CD40 can mimic the CD40L-CD40 interaction that activates type I interferon signaling and drives antigen specific antitumor T-cell responses.

There are currently at least seven agnostic anti-CD40 mAb in early phase clinical development: sotigalimab (APX005M), ChiLob7/4, ADC-1013, ABBV-927, and SEA-CD40 are of the IgG1 isotype and their agonism depends on cross-linking of FcR. Selicrelumab and CDX-1140 are of the IgG2 isotype that directly activates the CD40 molecule without the need for cross-linking. These anti-CD40 agonistic mAb also differ in the binding specificity to the CD40 receptor, which can also influence the degree of receptor activation (5).

Selicrelumab is the most extensively studied CD40 agonist to date, with a monotherapy toxicity profile complicated by cytokine release syndrome (CRS), decrease in peripheral lymphocytes, monocytes, and platelets, and elevated serum liver transaminases and total bilirubin. Clinical activity of this agent was modest and thought to be limited by narrow therapeutic window. The agent did demonstrate partial responses in 14% of patients in a phase I study, which included 27% of patients with melanoma (6). Notably, one metastatic melanoma patient remained in complete response (CR) for more than 9 years (7). Due to minimal monotherapy activity, subsequent early phase trials for advanced solid tumors explored the combination of selicrelumab with other immune checkpoint therapy, chemotherapy, anti-VEGF therapy, or CSF-1R inhibition. The combination of selicrelumab and tremelimumab for metastatic melanoma showed an ORR of 27.2%, with 10% CR and 18% PR, with the most common adverse event (AE) being CRS (8). Though not randomized, this response rate is higher than what would have been expected for anti-CTLA4 Ab alone. These results were never followed up however as the trial data were reported just as PD-1 antibodies entered clinical usage. The first-in-human study for sotigalimab in advanced tumors was reported in 2017 (9), demonstrating what was felt to be a wider therapeutic window than previously described CD40 agonists. The dose of 0.3 mg/kg was selected as the recommended phase II dose, representing the dose with maximum pharmacodynamic effects without grade >2 toxicities.

Weiss and colleagues (1) evaluated sotigalimab in combination nivolumab in a multicenter, open-label phase II trial for patients with anti–PD-1 refractory metastatic melanoma (mostly cutaneous, with small numbers of mucosal and acral subtypes). Five of the total 33 evaluable patients experienced partial response (ORR 15%), with duration of response lasting between 18.4 to 45.9+ months. 80% of the responders did not need further systemic therapy. The median duration of the treatment for patients with PR was 11.2 months with most responding patients experiencing greater than 50% target lesion reduction. Overall, the incidences of immune-related adverse events (irAE) from the combination were low and not higher than that expected for nivolumab alone. Grade 3 AE attributed to sotigalimab alone were 1 patient with AST/ALT elevation, and 1 patient with pyrexia, while G3 AEs attributed to the combination were 2 patients with AST/ALT elevation and 1 patient with elevated lipase/amylase. There were low incidences of significant infusion reactions, and no reported CRS. Overall, the response rate from the combination of sotigalimab and nivolumab compared favorably to that of nivolumab plus relatlimab (ORR 11%) in a similar PD-1 refractory patient population (10). Given this, higher response rates might be attainable if the combinations of sotigalimab with single or dual immune checkpoint therapies were evaluated in the frontline setting.

As only a subset of patients with melanoma benefit from combination therapy, there is an increasing urgency to identify predictors of benefit specific combinations. In the case of CD40, we and others hypothesized that there may be distinct differences in APCs and other immune cells in circulation for responders versus nonresponders to this therapy. Indeed, prior immune correlative studies in patients with metastatic pancreatic cancer identified that patients with > 1-year overall survival (OS) versus < 1 year OS after treatment with sotigalimab + chemotherapy had higher pretreatment and on-treatment frequencies of circulating cross-presenting DCs (CD1c⁺CD141⁺); this emerged as the strongest of all the predictive immune biomarkers (11). Other immune biomarkers in circulation associated with increased survival included higher on-treatment frequencies of conventional DCs, higher on-treatment concentration of soluble proteins associated with DC maturation (CD83, ICOSL), and higher pretreatment frequencies of circulating HLA-DR⁺CCR7⁺ B cells (11). The association between survival and on-treatment DC subsets points to more efficient cross-presentation with CD40 agonism and is congruent with previous studies showing induction of cross-presenting DCs and epitope spreading (12, 13). Furthermore, B cell subsets were associated with survival as expected with CD40 agonism. This may suggest a role for CD40 agonism in driving generation of germinal centers (14). These promising circulating immune biomarkers will need further validation in prospective trials to determine whether they can be used to select patients who will most likely benefit from CD40 agonistic therapy.

If the preliminary results presented by Weiss and colleagues are maintained in more definitive trials, it would be likely the case that the application of CD40 agonism would extend beyond melanoma. Due to its main biologic effect as a putative myeloid immune checkpoint, the possibility exists that CD40 agonism may have the potential to enhance antitumor immunity with anti- PD-1 and generate clinical effects even in non—T-cell–inflamed or intermediately inflamed/immune-excluded tumors. Indeed, several preclinical tumor models (including those of poorly immunogenic tumors such as pancreatic cancer, mesothelioma, breast cancer etc.) have shown synergistic effects of CD40 agonism in combination with treatment approaches such as chemotherapy, radiotherapy, and other immune based treatments (Fig. 1; ref. 5). Although the PRINCE study evaluating sotigalimab ± nivolumab + chemotherapy vs. nivolumab + chemotherapy for all-comer treatment-naïve metastatic pancreatic adenocarcinoma did not show significant improvement in 1-year OS for the sotigalimab-containing regimen, the accompanying aforementioned immune correlative studies identified distinct differences in circulating DC, B cells, and CD4⁺ T cells between patients having > 1 year vs. < 1 year OS (11). This suggests the importance of designing a prospective immune biomarker-driven study to help determine in advance the likely responders to the CD40 agonistic therapy.

In sum, there appears to be therapeutic potential for CD40 agonism in solid tumors though how to leverage this optimally has yet to be defined. The data from Weiss and colleagues suggest that in a tumor type known to be responsive to anti–PD-1, adding agonistic CD40 mAb can stimulate responses in otherwise refractory patients. These data deserve further exploration in a larger group of patients with PD-1 refractory melanoma and/or in a randomized trial of treatment- naïve patients. CD40 agonism has the potential to enhance the activity of other therapeutic modalities by modification of DC activity and priming of the tumor microenvironment. Whereas other mAbs against co-stimulatory receptors have seemingly run their clinical development course, a clear priority remains to determine the potential activity of CD40 agonists in combination with immune checkpoint blockade across the spectrum of immunologically “hot” to “cold” tumors.

R.C. Wu reports personal fees from Tempus and grants from Iovance, Xilio Therapeutics, and Regeneron outside the submitted work. J.J. Luke reports DSMB participation with AbbVie, Agenus, Immutep, and Evaxion; Scientific Advisory Board membership at (no stock) 7 Hills, Affivant, BioCytics, Bright Peak, Exo, Fstar, Inzen, RefleXion, Xilio, (stock) Actym, Alphamab Oncology, Arch Oncology, Duke Street Bio, Kanaph, Mavu, NeoTx, Onc.AI, OncoNano, physIQ, Pyxis, Saros, STipe, and Tempest; consultancy with compensation from AbbVie, Agenus, Alnylam, AstraZeneca, Atomwise, Bayer, Bristol-Myers Squibb, Castle, Checkmate, Codiak, Crown, Cugene, Curadev, Day One, Eisai, EMD Serono, Endeavor, Flame, G1 Therapeutics, Genentech, Gilead, Glenmark, HotSpot, Kadmon, Ko Bio Labs, Krystal, KSQ, Janssen, Ikena, Inzen, Immatics, Immunocore, Incyte, Instil, IO Biotech, LegoChem, Macrogenics, Merck, Mersana, Nektar, Novartis, Partner, Pfizer, Pioneering Medicines, PsiOxus, Regeneron, Replimmune, Ribon, Roivant, Servier, STINGthera, Sumoitomo, Synlogic, Synthekine, and Teva; research support from (all to institution for clinical trials unless noted) AbbVie, Astellas, AstraZeneca, Bristol-Myers Squibb, Corvus, Day One, EMD Serono, Fstar, Genmab, Hot Spot, Ikena, Immatics, Incyte, Kadmon, KAHR, Macrogenics, Merck, Moderna, Nektar, Next Cure, Novartis, Numab, Palleon, Pfizer, Replimmune, Rubius, Servier, Scholar Rock, Synlogic, Takeda, Trishula, Tizona, and Xencor; in addition, J.J. Luke reports patents for US-11638728 (Microbiome Biomarkers for Anti–PD-1/PD-L1 Responsiveness: Diagnostic, Prognostic and Therapeutic Uses Thereof).

R.C. Wu acknowledges National Institutes of Health 5K12CA133250–14 and P30 CA016058–47. J.J. Luke acknowledges NIH R01DE031729–01A1, UM1CA186690–06, P50CA254865–01A1, and P30CA047904–32.