Cluster of differentiation 40 (CD40), which belongs to the tumour necrosis factor (TNF)/TNF receptor (TNFR) family, is a 277-amino acid, 45–50 kDa transmembrane glycoprotein (1,2). Under physiological conditions, CD40 is primarily expressed on the membranes of various antigen-presenting cells (APCs), including dendritic cells (DCs), B-cells and monocyte-macrophages (3). It can also be expressed on nonimmune cells, including endothelial cells, epithelial cells, haematopoietic progenitors, and platelets (4,5) as well as tumour cells (6,7).

The natural ligand of CD40 (CD40L, otherwise known as CD154, TRAP or T-BAM), which also belongs to the TNF/TNFR family, is a 34–39 kDa type II integral membrane protein that is primarily expressed by active CD4⁺ T helper cells under inflammatory conditions. Previous studies have shown that the soluble form of CD40L, which lacks the transmembrane region, possesses biological activities similar to the transmembrane form (8,9).

The interaction between CD40 and CD40L is critical for the generation of a series of systemic immunizing inflammatory reactions. This includes the class switching and affinity maturation of immunoglobulins; secretion of cytokines; the survival, proliferation, differentiation and adhesion of B cells; and the development of memory B cell generation and germinal centres (10). In addition, CD40 intracellular signalling induces the apoptosis of many transformed cells both in vitro and in vivo, such as breast cancer and haematological malignancy cells (6,7).

With the inspiring successes of monoclonal antibodies (mAbs) targeting programmed cell death protein (PD-1)/programmed cell death ligand (PD-L1) in cancer therapy, immunomodulation by antibodies has been regarded as an attractive way of boosting anticancer responses (11), CD40 is an emerging immunotherapy target that plays a critical role in human immunity. Extensive preclinical research has proven that CD40 ligation stimulates antitumour immunity via several potential mechanisms across various lymphoid (12,13) and solid tumours (14–16).

To date, several anti-CD40 mAbs have been investigated by clinical studies. Here, we present an overview of the physiological and immunological context of CD40/CD40L, emphasizing the clinical outcomes of anti-CD40 mAbs for treating malignant disease.

The human CD40 gene is located on chromosome 20q11-13 and encodes a polypeptide chain of 277 amino acids, including an extracellular region of 193 amino acid residues, a transmembrane region of 22 amino acid residues and an intracytoplasmic region of 62 amino acid residues (17). As CD40 lacks intrinsic kinase activity in the cytoplasmic tail, the signal is mainly transferred by recruitment of TNF R-associated factors (TRAFs), which are cytoplasmic adapter molecules (18). TRAF trimers specifically recruit tyrosine and serine/threonine kinases to initiate rapid protein phosphorylation, which activates downstream signal transduction pathways, including the phosphoinositide 3-kinase (PI3K), p38/mitogen-activated protein kinase (38 MAPK), nuclear factor-κB (NF-κB) and c-Jun-NH2-kinase (JNK)/stress-activated protein kinase pathways (19–22).

The human CD40L gene is located on the X chromosome, Xq26.3-Xq27.1, and has five exons that span 12–13 kb, its DNA consists of a 783 bp reading frame encoding 261 amino acids, including an intracytoplasmic region of 22 amino acid residues, an extracellular region of 215 amino acid residues and a transmembrane region of 24 amino acid residues, which has a signalling and anchoring function (23,24). The CD40 ligand amino acid sequence has homology to the amino acid sequences of the extracellular regions of other TNF gene family members, such as TNF-α and tissue growth factor-β. CD40L has a carboxyl terminus located extracellularly and is a type II membrane protein that lacks an amino-terminal signal peptide. There are two soluble forms of CD40L (31 and 18 kDa), which retain the ability to bind CD40 and elicit signals in the form of homotrimers (25).

Unlike other TNFR costimulatory targets, CD40 is mainly expressed on APCs including B-cells, macrophages and monocytes (3). In addition, the CD40 is also present on non-immune cells, such as endothelial cells, epithelial cells, hematopoietic progenitor cells and platelets as well as various tumour cells, including malignant lymphoma cells, leukaemia cells and solid tumour cells (4–7).

The key physiological function of the CD40/CD40L pathway is mediated by CD40 ligation on APCs, especially dendritic cells (DCs). CD40L binding with CD40 leads to the activation of DCs, including improving their antigen presentation ability by upregulating the expression of other costimulatory molecules [major histocompatibility complex (MHC) class II, CD58, CD80/86 and CD70] and downregulating the expression of immunosuppressive molecules such as PD-L1 (19–21). It also develops a pro-survival signal and increases the release of various cytokines, including interleukin (IL)-1β, IL-6, IL-8, IL-12, TNF-α and interferon (IFN)-γ by DCs (26–28), and hence further enhances the cytotoxic response and prevents immune tolerance induction.

The binding of CD40L to CD40 also activates macrophages to directly kill tumour cells (29) and leads to the secretion of IL-1β, IL-6, IL-8, IL-12, TNF-α, IFN-γ and nitric oxide (27). These cytokines mediate the pro-inflammatory response and are crucial to macrophage cross-priming and cytotoxic function (30,31).

The ligation of CD40 activates resting B lymphocytes and causes these cells to differentiate into secretory plasmocytes or memory B lymphocytes but inhibits the growth and immunoglobulin production of active B lymphocytes (32). The differentiation path of activated B lymphocytes is partly dependent on the extent of CD40 activation. Short-term CD40L exposure can differentiate B cells into plasma cell lymphocytes, while long-term CD40L exposure can produce CD40⁺ memory B lymphocytes (33). A previous study found that CD40-stimulated B lymphocytes prolong the lifespan of memory B lymphocytes in vivo (34). In addition, patients with X-linked hyper-IgM immunodeficiency syndrome, which is caused by CD154 mutation/dysfunction, are unable to perform immunoglobulin class switching (35).

CD40 is expressed on ~100% of malignant B cell tumours (such as Hodgkin's lymphoma, non-Hodgkin's lymphoma, Burkitt lymphoma and multiple myeloma) (36). CD40 can also be identified on the cell surface of 70% of malignant epithelial tumours (including breast cancer, nasopharyngeal cancer and rectal cancer) (36).

There are diverse roles of CD40 pathway activation in cancer (12,37). Activation of the CD40 pathway has been proven to enhance the host antitumour immune response and/or directly induce tumour cell apoptosis in several models, especially in primary high-grade B cell lymphoma models of Burkitt's lymphoma, diffuse large B cell lymphoma (DLBCL) or Epstein-Barr virus-driven lymphoma (12,38). This pro-apoptotic effect of CD40 ligation is associated with the activation of cytotoxic ligands of the TNF superfamily (39). However, in some low-grade B cell malignancies, such as follicular lymphoma, chronic lymphocytic leukaemia and hairy cell leukaemia, stimulation of CD40 promotes malignant transformation, tumour proliferation, lymphomagenesis and resistance to chemotherapy (39–43) by inducing overexpression of survival proteins, such as Bcl-x and Bfl-1/A1, and downregulation of FLICE-inhibitory protein (44,45). Collectively, these studies have highlighted that activation of the CD40 pathway can exert either apoptotic or pro-survival effects depending on the type and differentiation state of the cancer cells involved (10). Therefore, care should be taken in clinical trials to exclude tumours for which preclinical trials have shown that activation of the CD40 pathway can lead to tumour progression.

Both preclinical experimental and clinical observations have demonstrated that tumour cells can interfere with an immune response and that deficient activation of antitumour immunity, such as decreased expression of MHC class I and/or adhesion or accessory/costimulatory molecules and a lack of tumour antigen presentation may facilitate tumour progression (42). Activation of the CD40 pathway can increase antigen presentation, and enhance cytotoxic activity and cytokine secretion, thereby enhancing the host's antitumour immune effect (46).

As a target for cancer treatment, CD40 can be activated by three approaches under clinical research (47): Recombinant human CD40L, adenovirus vector-expressed CD40L and agonistic anti-CD40 mAbs.

Due to its action on both the immune system and tumour cells, agonistic anti-CD40 antibodies have been studied as novel cancer immunotherapy targets, demonstrating potent antitumour immune responses in animal models and cancer patients (13,48–50). Each of the agonistic anti-CD40 mAbs has unique characteristics, including unique binding affinities to CD40, antibody isotypes, Fc modifications and agonistic effects (Table I). However, there is no consensus on which agonist is best for cancer therapy at present (51).

It is worth mentioning that the generation mechanism by which each of these antibodies generates agonism is not exactly the same. IgG1 Fc domain engineering was employed for APX005M based on the finding in a murine model that the potency of a CD40 agonist can be enhanced by increased binding affinity to FcγRIIB (52,53). In contrast, agonism of an IgG2 mAb, such as CP-870,893, is believed to be provided by its unique hinge conformation (54,55). A recent study demonstrated that hinge rigidity and selective FcγR binding affinity are both critical in regulating antibody agonistic function (56). In addition, other research has shown that the binding site is important in determining CD40 mAb agonistic activities, as membrane CRD1-binding displays stronger agonistic activities when the binding site is distal to the membrane than when it is proximal (57), and this relationship between binding epitope specificity and agonistic activity varies in TNFRs and needs to be resolved on a case-by-case basis (58).

To date, a variety of agonistic anti-CD40 mAbs are currently under investigation in clinical trials, as monotherapies or in combination with other agents (51,59–65). All these agonistic anti-CD40 mAbs are in the early developmental stage (Table II).

CDX-1140, developed by Celldex Therapeutics, Inc., is a human IgG2 antibody that stimulates CD40 signalling without the requirement for cross-linking or Fc receptor interactions (85). This drug is currently being investigated in a phase I clinical trial.

ABBV-927 (AbbVie, Inc.) is an anti-CD40/anti-mesothelin bispecific antibody that is being tested in phase I trials for the treatment of advanced solid tumours, including non-small cell lung cancer, squamous cell carcinoma of the head and neck, cutaneous malignant melanoma, and pancreatic adenocarcinoma, as monotherapy or in combination with other immunotherapies (anti-PD-1 and anti-OX40 antibodies) (86). The estimated primary completion date is 2023.

APX005M, developed by Apexigen, is a humanized mAb IgG1/k against CD40 (87). A preclinical study has demonstrated that APX005M binds to CD40 at the CD40L binding domain with a high affinity in mice (K_d=0.12 nM) and monkeys (K_d=0.37 nM). The first-in-human phase I study for dose determination was conducted in patients with solid tumours (clinical trial identifier: NCT02482168) in 2015. The trial was completed at the end of 2018 without the results being reported. In 2017, three phase Ib/II studies and one phase II study were launched to investigate the safety and potential efficacy of APX005M combined with immune checkpoint inhibitors or chemotherapy in a variety of solid tumours, including APX005M combined with pembrolizumab for treating metastatic melanoma (clinical trial identifier: NCT02706353), a combination of APX005M and nivolumab in the treatment of solid tumours (clinical trial identifier: NCT03123783), APX005M in combination with chemotherapy with or without nivolumab in the treatment of metastasized pancreatic adenocarcinoma (clinical trial identifier: NCT03214250), and APX005M in combination with concurrent chemoradiation for resectable oesophageal/gastro-oesophageal carcinoma (clinical trial identifier: NCT03165994). In 2018, three phase I studies were launched. One study (clinical trial identifier: NCT03389802) has investigated the potential to overcome resistance to PD-1/PD-L1 blockade immunotherapy by the combination of APX005M with cabiralizumab, an anti-CSF1R antagonist, with and without nivolumab in several solid tumours. Another study (clinical trial identifier: NCT03502330) investigated the therapeutic potential of APX005M for treating paediatric CNS tumours. In addition, an innovative study (clinical trial identifier: NCT03597282) investigated the potential synergistic effect of APX005M with a vaccine (NEO-PV-01) in patients with advanced melanoma. In 2019, a phase II randomized multicentre trial was initiated for neoadjuvant therapy with or without APX005M in patients with locally advanced rectal adenocarcinoma. These trials are still ongoing at time of writing.

At present, most agonistic anti-CD40 antibodies have been demonstrated to be well tolerated both as single agents and in combination in the treatment of solid tumours and haematological malignancies (47,50). Nevertheless, due to most agonistic anti-CD40 antibodies being in the early development stage, the safety information of these agents has not been fully investigated. The most common adverse events associated with CP-870,893, a fully humanized IgG2 antibody and a strong agonist, was transient grade 1 to grade 2 cytokine release syndrome (CRS). This syndrome, which occurs within minutes to hours after infusion, is characterized by a variety of combinations of chills, rigors, rash, nausea, fever, vomiting, muscle aches and back pain. In a human phase I study of CP-870,893, the CRS occurring in 55% of patients was considered to not be an anaphylactic or allergic reaction by the authors, as normal serum tryptase levels were observed in the subjects (66). There was an association of CP-870,890 with acute elevations in TNF-α and IL-6 in serum. In most cases, this syndrome can be resolved by treatment with doxylamine within 24 h. Similar to CP-870,890, dacetuzumab, a weakly-agonistic humanized IgG1 antibody, showed a high overall incidence of CRS within one day of infusion (41%) in the first-in-human trial, but the agents were generally tolerated well.

Another major safety issue of CP-870,893 is dose-related haematological toxicities, such as a decrease in peripheral lymphocytes, monocytes and platelets. Grade 3–4 lymphopenia and thrombocytopenia were observed in a clinical trial of CP-870,893. Grade 3 anaemia, neutropenia, and thrombocytopenia were observed in a human study of dacetuzumab. Weekly dosing of CP-870,893 may cause long-lasting lymphocytopenia.

Transient elevations in D-dimer levels and two pulmonary embolism cases (grade 4 thrombosis: Of which, one was from the single-dose schedule and one was from the weekly dose schedule) were observed in two phase I clinical trials of CP-870,893.

Aaspartate aminotransferase, alanine aminotransferase and total bilirubin were transiently increased after both CP-870,893 and dacetuzumab infusion, which suggests hepatoxicity of both agents. A study of the weekly dose schedule of CP-870,893 showed that liver enzyme abnormalities returned to baseline in most patients at the time of the next infusion. A human study of dacetuzumab showed that elevations in hepatic transaminases were asymptomatic and not associated with marked changes in bilirubin.

APC dysfunction remains a rate-limiting biological issue in many cancer environments. Targeting CD40 with agonists to enhance APC function and indirectly regulate host immune cells by recruiting innate immune effectors via ADCC/ADCP is considered a promising method to treat cancers. To date, early clinical trials have shown that anti-CD40 antibodies have limited clinical activity and no severe immune-related autoimmune-like toxicities, either as monotherapies or in combination with other treatments. Thus, combination strategies targeting of CD40 for cancer therapy require further investigation in clinical trials with careful designs.