Tumor localized agonistic anti-CD40 therapy and beyond

KEYWORDS:

• CD40
• Immunotherapy
• Monoclonal
• Cancer

1. Introduction

The field of anti-CD40 agonistic antibody therapies is still awaiting the breakthrough in the clinic. Anti-CD40 agonistic antibodies should in theory complement anti-PD1/PDL1 therapy as they provide the necessary CD40 stimulation to antigen-presenting cells, thereby priming T cells in the presence of tumor antigen presentation in the absence of helper T cells. The preclinical effect has been well documented and there are several models providing evidence of efficacy, mainly driven by CD8 T cell priming and T cell mediated anti-tumor responses. However, as an infusion therapeutic and monotherapy in the clinic, there has so far been limited success and research has shown that we need to rethink the clinical use of anti-CD40 therapies and as such the field is expanding with novel administration routes, combination strategies and multi-specific antibodies to improve therapeutic responses. In addition to their anti-tumor effects via T cells, CD40 agonistic therapies can mediate anti-tumor activity via macrophages along with direct cytotoxic anti-tumor effects on CD40 positive tumors, and can also attract/stimulate myeloid cells. As such the indication of choice, the tumor micro-environment and the route of administration of these antibodies can impact the resulting immune response and should be considered when entering clinical trial evaluation.

2. Recent advances in the field of agonistic anti-CD40 antibodies

CD40 is expressed on antigen-presenting cells; it is highly and constitutively expressed on B cells, whereas it is expressed at low levels on immature dendritic cells and macrophages and is upregulated upon their activation. CD40 ligation with CD40L on helper T cells is critical for antigen-presenting cell activation and proliferation and for immunoglobulin class-switch [1,2]. Therapeutic strategies to target CD40 have been mainly focusing on developing monoclonal agonistic antibodies, recombinant forms of CD40L or viral delivery of CD40L to the tumor area to induce tumor cells to express CD40L. The first agonistic antibody to enter the clinic was developed by Pfizer, the CP-870,893 (now pursued in the clinic by Hoffman-la Roche with the generic name selicrelumab). This is an IgG2 antibody developed from the Abgenix Xenomouse® platform that binds CD40 with nanomolar affinity [3] without competing with the CD40L binding site [4]. It induces activation of CD40 expressing immune cells and, as an infusion product, transiently reduces circulating CD40 expressing cells in blood [5]. The maximum tolerated dose (MTD) as a single-dose was determined to 0.2 mg/kg. Cytokine release syndrome (CRS) was the most common adverse event. Grade 3–4 toxicities were however not related to CRS but rather to hematological and hepatic toxicity. Out of four patients that displayed a partial response (PR) in this trial, three of these patients, all with progressive disease at start of the study, received 0.2 mg/kg and one patient received 0.3 mg/kg. Although this is a limited observation set, this may suggest that the MTD and the clinical effective dose are identical resulting in a narrow therapeutic window. Interestingly it has been shown in murine models that a tumor-localized injection of anti-CD40 can be as effective as a systemic infusion [6]. This is due to that CD40 agonistic monoclonal antibodies provide a costimulatory signal to a dendritic cell (signal 2), and T cell activation as an end result requires antigen-presentation (signal 1) to activate and expand the tumor-directed T cell. Thus, for an optimal therapeutic effect of CD40 agonistic antibodies and reduced toxicity, it is essential to activate immune cells where antigen-presentation takes place and limit systemic inflammatory responses [6,7]. The need for tumor-localized immune activation was also key for the development of mitazalimab (ADC-1013) as an intratumoral injectable product [8].

The field has also leaped into trying to understand the underlying agonistic mechanism of anti-CD40 antibodies with respect to the targeted epitope, isotype and Fc gamma receptor (FcgR) interaction profiles. It has been reported that antibodies targeting epitopes in the CRD1 region of CD40 are the agonistic antibodies as compared to antibodies targeting epitopes found in the CRD2-4 [4]. Professor Glennie and Dr White with colleagues have studied the role Fc receptor cross-linking and of hinge rigidity for the agonistic effect of CD40 antibodies, Professor Ravetch with colleagues have focused on humanized models to study the role of FcgR cross-linking to provide optimal agonist effect of these antibodies [4,9–12]. In September, this year, an article came out in Nature communications giving support to both the relevance of the CH1-hinge region of the IgG2 format for optimal rigidity to provide agonistic activity along with the importance of a rigid hinge for activity, also when FcγR cross-linking is part of the mode-of-action [13]. Flexible hinges, either designed or the natural flexible IgG3 format, showed no ability for the agonistic activity of a CD40 targeting antibody. It is clear that CD40 targeting by antibodies requires multimerization of the therapeutic and the target on the cell surface or intracellularly in order to provide the threshold of activation for down-stream CD40 activation. This is not surprising as the naturally occurring CD40 activation is mediated via trimer CD40L. An example of a second-generation anti-CD40 agonistic antibody with improved FcγRIIb cross-linking capacity is Apexigen’s APX005M, a humanized IgG1 antibody with the S267E mutation that improves FcgRIIb and abrogates FcgRIIa interactions. This antibody also blocks the natural CD40/CD40L interaction [14]. The antibody has now reached phase II studies.

3. Expert opinion

The field of CD40 agonistic antibodies has advanced in the knowledge of the design for optimal efficacy. However, the field is still focused on CD40 agonistic antibodies as infusion products and as such rely on combination therapy strategies. Unfortunately, a more potent agonist will not aid an efficient T cell driven anti-tumor response as long as we do not consider that these antibodies act as adjuvants and that they require antigen-presentation to efficiently trigger anti-tumor T cell responses. Immune activation without proper antigen-presentation will hamper the clinical impact of these antibodies, if the anti-tumor response is via a T cell-mediated effect. In some reports, cytotoxicity of CD40 agonistic antibodies have been reported to be driven by macrophages [15] and in those cases a systemic administration may be of value. Considering the importance of dendritic cells for antigen presentation, it is also plausible that systemic administration leads to a B cell-mediated antibody sequestering that shifts the antibody-mediated CD40 stimulation to germinal center formation or alternatively to a reported short-lived humoral immune response by extrafollicular B cell activity [16], rather than to an efficient dendritic cell/CD8 T cell activation. In light of what we know about the role of the hinge and Fc receptor cross-linking for optimal agonistic antibody activity, the effect on the B cell compartment may depend on the given agonistic antibody investigated in an individual study. Clinical use of selicrelumab has reported an increase in circulating CD27 + B cells in a clinical trial, suggestive of a B cell memory response in circulation [17]. Thus, our understanding of how the plethora of CD40 triggered immune responses in humans converge to a net immune response based on the sum of myeloid, B cell, and dendritic cell driven immune responses in a certain tissue/organ location requires more studies using the individual therapeutic antibodies that are currently in clinical testing. The complication of this is of course the lack of cross-reactivity of these antibodies along with the difference in Fc and immunoglobulin biology between the species. For the future success of CD40 directed agonistic immunotherapy, it will be critically important that the field advances the design of novel multifunctional antibodies for systemic administration with tumor-targeting properties along with formulation initiatives for optimal tumor-localized immunotherapy. Preferably the field will also further refine the biological readout systems to characterize these novel entities in detail, to provide an improved translational understanding on how to best utilize CD40 as a drug target for the patient’s benefit.

Declaration of interest

The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Reviewer Disclosures

Peer reviewers on this manuscript have no relevant financial relationships or otherwise to disclose.

Additional information

Funding

This paper was funded by Svenska Sällskapet för Medicinsk Forskning.