Harnessing the Potential of CD40 Agonism in Cancer Therapy

Abstract

CD40 is a member of the tumor necrosis factor (TNF) receptor superfamily of receptors expressed on a variety of cell types. The CD40-CD40L interaction gives rise to many immune events, including the licensing of dendritic cells to activate CD8⁺ effector T cells, as well as the facilitation of B cell activation, proliferation, and differentiation. In malignant cells, the expression of CD40 varies among cancer types, mediating cellular proliferation, apoptosis, survival and the secretion of cytokines and chemokines. Agonistic human anti-CD40 antibodies are emerging as an option for cancer treatment, and early-phase clinical trials explored its monotherapy or combination with radiotherapy, chemotherapy, immune checkpoint blockade, and other immunomodulatory approaches. In this review, we present the current understanding of the mechanism of action for CD40, along with results from the clinical development of agonistic human CD40 antibodies in cancer treatment (selicrelumab, CDX-1140, APX005M, mitazalimab, 2141-V11, SEA-CD40, LVGN7409, and bispecific antibodies). This review also examines the safety profile of CD40 agonists in both preclinical and clinical settings, highlighting optimized dosage levels, potential adverse effects, and strategies to mitigate them.

Graphical Abstract

1. Introduction

Multiple signals are necessary for naïve T cells to differentiate into effector T cells in an adaptive immune response. Antigen-presenting cells (APCs), including dendritic cells (DCs), B cells, and macrophages, are essential for the delivery of primary signals to T cell receptors (TCRs) and secondary signals from costimulatory molecules. Additional tertiary signals include the secretion of cytokines, which modulate the activation and skew the differentiation of effector cells, including T cells [1].

One key coordinator in the activation process of effector cells is CD40, a costimulatory receptor molecule belonging to the tumor necrosis factor (TNF) receptor superfamily [1]. CD40 is mainly expressed in B lymphocytes, DCs, and monocytes [2]. The ligand of CD40, CD40L (also known as CD154), is primarily expressed by activated T cells. To maintain immune homeostasis, naïve T regulatory cells express low basal levels of CD40L, which can be upregulated upon activation [3]. In addition, ligation of CD40 by activated CD4⁺ T cells “license” DCs for CD8⁺ effector T cell activation [4]. The CD40 ligation on DCs also upregulates the cell surface expression of major histocompatibility complex (MHC), and the production of proinflammatory cytokines, such as IL-12 [5]. With viral infection, CD40L upregulation on DCs may directly prime CD8⁺ T cells independent of CD4⁺ T cells, promoting cytotoxic T lymphocyte (CTLs) responses [6]. For B cells, CD40 ligation upregulates costimulatory molecules, for example, B7-1/2, which reciprocally prime effector T cell activation [4]. CD40L is also expressed on platelets, and under inflammatory conditions, is induced in natural killer cells [7], mast cells, and basophils [8].

CD40 activation may modulate the tumor microenvironment (TME) independent of canonical innate immune sensors, such as Toll-like receptors (TLRs), inflammasomes, and stimulator of interferon genes (STING) [9]. CD40 knockout mice exhibited impaired T cell priming of B cells to class switching or to germinal center formation [10], and also exhibited significantly higher occurrence of spontaneous tumors [11]. In epithelial cancer and melanoma cells, CD40L ligation to CD40 mediates immunogenic cell death that leads to the activation of dendritic cells and CTLs in the TME [12, 13]. CD40 activation also extends efficiency to immune checkpoint blockade (ICB) in “cold” tumors that are typically resistant to checkpoint inhibitors, such as pancreatic cancer and melanoma tumors [13, 14]. Thus, CD40-CD40L interaction plays a critical role in activating the anti-tumor immune responses, which sparked CD40 agonism as a promising target for cancer immunotherapy. More than 23 years ago, agonistic CD40 antibodies were shown to mimic CD40L and initiate downstream anti-tumor immune signaling, particularly activation of CTLs [15]. Although toxicity issues hindered this agonist approach as a monotherapy, CD40 monoclonal antibodies have recently demonstrated their effectiveness in vivo for cancer therapy, especially when used in conjunction with other treatments [16]. Together, CD40-activating monoclonal antibody promotes CD40 agonism as a potentially significant yet unexplored means to broaden the scope of existing cancer immunotherapy.

2. CD40/CD40L downstream signaling mechanism

CD40 largely relies on the recruitment of the TNF receptor-associated factor (TRAF) family to transduce a wide range of downstream signals (Figure 1). TRAF1, TRAF2, TRAF3, TRAF5 and TRAF6 have been reported to bind to CD40, among which TRAF2, TRAF3, and TRAF6 most significantly contribute to CD40 intracellular signaling [17]. The most well-documented CD40 signaling pathway is the non-canonical nuclear factor kappa-B (NF-κB) pathway. Induction of this pathway involves NF-κB-inducing kinase (NIK)-mediated activation of inhibitor of NF-κB kinase α (IKKα). Subsequently, IKKα triggers selective proteolysis of p100 to produce p52, which then promotes the transcription of target genes [18]. Another CD40 downstream signaling pathway is the canonical NF-κB pathway. With NF-κB activator 1 (Act1) acting as an adaptor, diverse immune receptors activate a trimeric IκB kinase (IKK) complex (IKKα/β/γ) through the kinase TGFβ-activated kinase 1 (TAK1) [19]. Subsequently, the IKK complex mediates NF-κB inhibitor (IκB) degradation, leading to nuclear translocations of p50/p65 and p65/p65 dimers [18]. In B cells, NF-κB signaling promotes activation, differentiation, and proliferation [17, 18]. While in DCs, activation of the non-canonical NF-κB pathway is critical for DC maturation [20] and NIK-mediated cross-priming of CTLs [21]. Other signal transduction pathways mediated by the CD40-TRAF complex include the three mitogen-activated protein kinase (MAPK) pathways - the extracellular signal-regulated kinases (ERKs) pathway, the c-Jun N-terminal kinases (JNKs) pathway, and the p38 pathway [19]. Differential activation of MAPKs leads to the activation of a unique set of transcription factors, resulting in distinct biological responses [4, 20]. In B cells, CD40-dependent activation of JNK and p38 can lead to proliferation, antibody production, and Ig switching [22]. In DCs, p38 MAPK signaling is crucial for IL-12 production that then promotes differentiation of T helper type 1 (Th1) lymphocytes [23, 24]. Another well-known downstream pathway of CD40-TRAF interaction is the phosphoinositide 3-kinase (PI3K) pathway. In B cells, CD40-PI3K regulates most cell fate decisions, including the development and differentiation of bone marrow precursors and B cell receptor (BCR)-mediated proliferation [25]. In monocytes, CD40-triggered PI3K-AKT signaling plays a role in regulating the production of TNF-α and Interleukin-10 (IL-10) [26]. In DCs, the CD40-TRAF6-mediated PI3K/AKT pathway promotes cell survival via suppressing apoptosis factors, such as BCL2-associated agonist of cell death (BAD), caspase 3, and caspase 9 [24].

Figure 1. Downstream signaling pathways of CD40/CD40L.

CD40 ligation initiates diverse downstream signaling pathways. Independent of TRAF mediation, CD40 activates the Janus kinase 3 (JAK3), phosphorylating and activating the signal transducer and activator of the transcription 5 (STAT5). Several pathways are TRAF-dependent. In the canonical nuclear factor kappa-B (NF-κB) pathway mediated by the NF-κB activator 1 (Act1) and kinase TGFβ-activated kinase 1 (TAK1), activation of the trimeric IκB kinase (IKK) complex (IKKα/β/γ) mediates NF-κB inhibitor (IκB) degradation, leading to nuclear translocations of the p50/p65 dimers. Induction of the non-canonical NF-κB pathway involves NF-κB-inducing kinase (NIK)-mediated activation of the inhibitor of NF-κB kinase α (IKKα), which triggers selective proteolysis of p100 to produce p52 and promotes transcription of p52/RelB. Three mitogen-activated protein kinase (MAPK) pathways are activated by CD40 ligation, including the extracellular signal-regulated kinases (ERKs) pathway, the c-Jun N-terminal kinases (JNKs) pathway, and the p38 pathway. The TRAF6-mediated phosphoinositide 3-kinase (PI3K)- protein kinase B (AKT) pathway also plays a crucial role via suppressing caspase-9/caspase3- associated apoptosis. CD40 ligation on cancer cells can lead to direct apoptosis via two distinct pathways downstream of JNK activation. In the intrinsic pathway, JNK activates the cytochrome c/caspase-9/caspase-3-associated intrinsic apoptosis in a caspase-8-dependent manner. In the extrinsic pathway, JNK triggers the intracellular TNF-related apoptosis-inducing ligand (TRAIL) induction, initiating caspase-10-associated apoptosis.

In addition to TRAF-mediated pathways, CD40 also activates the Janus kinase 3 (JAK3), which then phosphorylates and activates the signal transducer and activator of the transcription 5 (STAT5) [27]. In B cells, upon CD40 stimulation, JAK3 phosphorylation does not occur, indicating that JAK may not be crucial for CD40 signaling in B cells. However, monocytes induce JAK3 phosphorylation after CD40 engagement, inducing STAT5 dimerization and translocation to trigger targeted gene transcription [27]. In DCs, CD40-dependent JAK/STAT activation induces the Th1-polarizing cytokine IL-12, IFN-γ, and costimulatory molecule expressions, influencing the maturation of monocyte-derived dendritic cells [28].

In general, it is evident that the CD40/CD40L downstream signaling effects vary depending on cell types (Figure 1).

3. CD40/CD40L upstream signaling mechanism

Compared to CD40 downstream signaling pathways, less is known about the upstream signals regulating the CD40 expression [29] (Table 1). CD40 has a TATA-less promoter that contains four NF-κB sites and three IFN-γ activation sites (GAS) elements [30]. Mechanisms of induction of CD40 vary among cell types. In human macrophages and microglia, CD40 transcription is promoted through the canonical NF-κB pathway and STAT1 activity. LPS induces the production of IFNβ, which activates STAT-1α. Moreover, LPS induces the acetylation and phosphorylation of H3 and H4, facilitating recruitment of NF-κB, STAT-1α, and RNA polymerase II to the CD40 promoter in a time-dependent manner [30]. In human macrophages, IL-6 can induce CD40 expression through upregulating STAT3 and hypoxia-inducible factor (HIF)-1α [31]. In mouse macrophages, transcription factors such as STAT-1α and ETS (PU.1, SP1-B) have been shown to mediate CD40 expression after IFN-γ stimulation [32]. After LPS stimulation, CD40 expression is upregulated by NF-κB in a p65-dependent manner [33]. In human DCs, CD40 is transcriptionally regulated by RUNX1 and RUNX3 and by ribonucleotide binding proteins RPL26, RPL4, RPL8, and RPS9 [34, 35]. In contrast, arginine and serine-rich coiled-coil 2 (RSRC2) were shown to suppress CD40 expression [35]. In human B cells, AKNA, a human AT-hook protein, has been shown to bind to the A/T-rich promoter regions of CD40 and CD40 ligand (CD40L) to stimulate expression of both ligand and receptor [36]. Moreover, disruptions in the binding of EBF1 and STAT1 can significantly decrease CD40 promoter activity in human B cells, indicating their stimulatory roles in CD40 expression [37].

Table 1.

Transcription factors upregulating CD40

Cell Type	Origin	Transcription Factor	Induced by	References
Macrophages and Microglia	Human	NF-κB STAT-1α RNA polymerase II	LPS	[30]
Macrophages	Mouse	NF-κB PU.1, Spi-B STAT-1α Sp1	LPS IFNγ IFNγ or LPS-induced IFNβ N/A	[33] [32] [30] [32] [33]
Dendritic Cells	Human	RUNX1 and RUNX3 RPL26, RPL4, RPL8, RPS9	LPS or TNFα or SAG N/A	[34] [35]
B cell	Human	AKNA EBF1, STAT1	N/A	[36] [37]
Malignant Melanoma	Human	N/A	IFNγ, TNFα	[40]

NF-κB: nuclear factor kappa-B; STAT-1α: signal transducers and activators of transcription 1α; RUNX1: RUNX family transcription factor 1; RUNX3: RUNX family transcription factor 3; RPL26: ribosomal protein L26; RPL4: ribosomal protein L4; RPL8: ribosomal protein L8; RPS9: ribosomal protein S9; AKNA: AT-Hook transcription factor; EBF1: EBF transcription factor 1; STAT-1: signal transducers and activators of transcription 1; LPS: bacterial lipopolysaccharides; IFNγ: interferon gamma; IFNβ: interferon beta; TNFα: tumor necrosis factor α; SAG: superantigen.

Apart from inducing signals, several signals were found to repress CD40 activation in mouse models. While IFN-γ is a potent inducer of CD40 expression in macrophages, IL-4, a Th2-derived cytokine that acts antagonistically to IFN-γ, was found to suppress IFNγ-induced CD40 gene expression in macrophages and microglia, in a STAT6-dependent manner [38]. Moreover, another immunosuppressive cytokine named transforming growth factor-β (TGF-β) was reported to inhibit IFN-γ-induced CD40 in microglia by enhancing the degradation of CD40 mRNA [39]. In mouse macrophages, SP1 was shown as a critical regulator by which phosphorylation can downregulate basal CD40 gene expression [33].

In human malignant melanoma cell lines, CD40 expression is enhanced by IFN-γ and TNF-α [40]. However, the transcription factors regulating CD40 expression in malignant cells remain poorly defined.

4. Role of CD40 activation in immune cells

On immature DCs, CD40 ligation stimulates the production of inflammatory cytokines, such as IL-1α, IL-1β, TNF-α, IL-6, and IL-8 [1, 41] (Figure 2). CD40 activation also upregulates the expression of costimulatory molecules such as CD80, CD86, and adhesion molecules such as CD54 (ICAM-1) and CD58 (LFA-30) [24, 41]. The costimulatory signals can “license” DCs to present exogenous antigens in the context of MHC-I molecules, and thus cross-present to CD8⁺ T cells[4]. Moreover, it has recently been shown that the CD40L-expressing CD8⁺ T cells may upregulate IL-12 production by DCs via CD40L-CD40 interactions, which forms a positive-feedback loop to promote their own proliferation and differentiation [42]. Mature DCs and macrophages secrete IL-12, which upon CD40L stimulation, results in sustained polarization to Th1 response and activation of NK-cell antitumor responses [23]. In macrophages, CD40 ligation stimulates a tumoristatic effect through increasing IFN-γ, TNF-α, T-cell dependent nitric oxide production, and antibody-dependent cellular cytotoxicity (ADCC). Also, CD40-activated macrophages can induce apoptosis in tumor cells in vitro, such as in mouse lymphoma cells (L5178Y) [43]. CD40 is constitutively expressed on platelets and its ligation is critical for platelet homeostasis and activation during inflammation [44]. In B cells, CD40 ligation promotes germinal center formation, where B cells undergo differentiation and proliferation through clonal expansion, somatic hypermutation, and immunoglobin class switching, resulting in high-affinity antibody production [17, 18].

Figure 2. The role of CD40-CD40L interaction on immune and malignant cells.

Schematic of the effect of CD40 ligation by membrane-bound (mCD40L) or soluble (sCD40L) form of CD40L on immune and cancer cells.

CD40 stimulation can also be generated by a soluble form of CD40L (sCD40L), which co-exists with the membrane-bound form of CD40L (mCD40L) [45]. The trimeric sCD40L can be generated by enzymatic cleavage of mCD40L by matrix metalloproteinases (MMPs) [45]. Activation by sCD40L triggers trimerization of the receptor, which can activate both the canonical and non-canonical NF-κB pathways [46]. The human native sCD40L is biologically active [47] and generates similar activities as mouse sCD40L, such as promoting B cell differentiation and proliferation [48]. In contrast, stimulation of endothelial cells by platelet-derived sCD40L promotes tumor angiogenesis, shedding light on potential negative effects with delivery of sCD40L as a therapeutic (discussed later) [49]. Early clinical testing of sCD40L did not produce positive results [50], which may due to the fact that sCD40L activity was dependent on FcγR-mediated cross-linking [51], and CD40-expressing immune cells were eliminated by ADCC [52]. Recently, hexavalent forms of sCD40L, such as the hexavalent receptor agonist (HERA), were developed to improve drug efficiency by circumventing the need for FcγR. HERA fusion proteins comprise three receptor-binding domains in a single chain arrangement linked to an Fc-silenced human IgG, resulting in a hexavalent molecule [52]. HERA-CD40L enhances activation of signal transduction in DCs via signaling through the NF-κB, AKT, p38, ERK1/2, JNK, and STAT1 pathways. This helps establish robust antitumor immune responses both in vitro and in vivo via increasing intratumor CD8⁺ T cell and promoting macrophage polarization from the pro-tumor M2 macrophages to antitumor M1 macrophages [52, 53].

5. Role of CD40 activation in malignant cells

CD40 is expressed in a variety of malignant cells, and based on tumor cell types, CD40-CD40L ligation differentially mediates cell survival, proliferation, and apoptosis [54, 55] (Figure 2). In malignant B cells, low-level constitutive CD40 engagement has been shown to promote sustained proliferation, such as in non-Hodgkin lymphoma (NHL) and chronic lymphocytic leukemia (CLL) cells [54, 56]. Related mechanisms include upregulating the expression of anti-apoptotic factors, such as Bcl-XL and TNF-α-induced protein 3 (TNF-A IP3, or A20) [57, 58]. In fact, TNF-A IP3 can suppress inflammatory and autoimmune responses induced by the NF-kB pathway in B cells, and is critical in maintaining normal B cell differentiation of the marginal zone B and B1 cell subsets [59]. Moreover, CD40 activation induces cytokines with pro-survival activities in B cells, including IL-6, TNF-α, and GM-CSF, that can potentially affect malignant B cell survival, migration, and interaction with other cells in the TME [60]. Paradoxically, CD40-CD40L ligation has also been reported to induce apoptosis in various malignant B cells, for example, in Burkitt lymphoma (BL) [55]. The dual roles of apoptosis and proliferation can be explained by a balance determined by two factors during T cell-B cell interaction. Firstly, higher relative abundance of Th1 cells compared to Th2 cells lead to higher secretion of cytokines such as IL-2 and IFN-γ. This, in turn, triggers the expression of APO-1/FAS on B cells, such as the BL cell line Ramos, thereby promoting apoptosis [61]. Secondly, the timing and quality of BCR stimulation by antigen is crucial. Compared to B cells capturing antigen without BCRs or with desensitized receptors, BCRs stimulated by a sudden burst of new foreign antigen receive full intracellular signaling. The presence of prior BCR signaling, the binding of FasL to Fas on B cells does not eliminate the B cells but instead promotes cell division and proliferation [62].

Similarly, under CD40 stimulation, a variety of carcinoma cells can be induced to undergo apoptosis, including CD40 expressing human malignant melanoma [40], multiple myeloma [63], B-cell lymphoma [64], bladder cancer [15], and colorectal cancer (CRC) [65]. Studies in the epithelial context have unraveled several potential mechanisms of CD40-mediated intrinsic pathway of apoptosis of carcinoma cells. Using membrane-bound CD40L, it is shown that cell death is induced through the upregulation of TRAF3 protein, which activates the JNK pathway, initiating cytochrome c/caspase-9/caspase-3-associated intrinsic apoptosis [65, 66]. Moreover, the TRAF6-mediated activation of caspase-8 is essential for this process [66]. In addition, an extrinsic pathway downstream of JNK activation was reported to operate alongside with the intrinsic apoptosis pathway in CRC, in which JNK triggers the induction of the intracellular TNF-related apoptosis-inducing ligand (TRAIL), initiating caspase-10-asscociated apoptosis. This concurrent action results in a faster apoptosis process compared to other types of carcinoma cells [67].

6. CD40 expression in tumor cells and regulation of the TME

CD40 expression dynamically varies among cancers. Our group investigated the change of CD40 expression in a pan-cancer analysis. Of the twenty-three cancer types screened, CD40 downregulation occurred in eleven, while upregulation occurred in five [68]. Previous studies have also supported the downregulation of CD40 during tumor progression in several cancers, including malignant melanomas and gastric tumors, in which the latter associated with decreased apoptosis and increased proliferation of the immunosuppressive myeloid-derived suppressor cells (MDSC) [40, 69]. Activation of CD40 in cancer cells has frequently been associated with antitumor immunity. Hill et al. (2005) showed that CD40 activation in human cervical carcinoma cells can activate NF-κB and MAPK signaling pathways, upregulate cell surface markers and intracellular molecules to facilitate antigen processing and presentation, and enhance tumor recognition by CTLs [70]. CD40/CD40L expression in melanoma was shown to correlate with not only type-I antitumor T-cell and APC responses in the TME but also a better predicted response to ICB therapy [68]. While CD40 activation or upregulation was beneficial to tumor killing, contrary results were found in lung and oesophageal cancer, where upregulated CD40 expression on cancer cells correlated with poor prognosis [71, 72]. These contradictory results emphasize the potential of CD40 as a specific immunological target and that strategies of CD40 therapeutics need to be customized depending on cancer types [73].

While it is difficult to predict prognosis from CD40 expression levels, DNA methylation of CD40 has been discovered as an innovative noninvasive biomarker for monitoring the prognosis and diagnosis of several types of cancers, including gastric, breast, colorectal, and prostate cancers [74-76]. Co-expressed genes of CD40 have also been examined in three immunotherapy-responsive and/or resistant cancer types (clear cell renal cell carcinomas, cutaneous melanoma, pancreatic adenocarcinomas) (reported in [77]). Eight genes correlated with CD40 expression are shared between all three cancer types: KIAA1949 (a potential tumor suppressor) [78], IL-27 receptor alpha (IL-27RA) (immune regulation) [79], proteasome subunit beta type-8 (PSMB8) (antigen presentation-related proteasome), tripartite motif containing 21 (TRIM21) (antibody receptor involved in various immune responses) [80], vesicle-associated membrane protein 5 (VAMP5) (involved in vesicle activities) [81], caspase 4 (inflammasome activator) [82], thymidine phosphorylase (TP) (involved in tumor proliferation and apoptosis) [83], and syntaxin 4 (STX4) (involved in membrane fusion) [77, 84]. These results entail that CD40 signaling activities are closely related to immune modulation, proliferation, apoptosis, and vesicular trafficking, thus providing insights into co-targets with CD40 agonist-based therapy.

7. CD40 agonists in clinical trials

Several properties modulate the function of agonistic human anti-CD40 antibodies: fragment crystallizable (Fc)-mediated interactions, isotype, epitope, and specificity.

Cross-linking of the IgG1 monoclonal antibody Fc region through interacting with Fc gamma receptors (FcγR) is crucial for effective CD40 downstream signaling events. The clustering of CD40 receptors amplifies downstream signaling triggered by CD40 activation, thus optimizing immunostimulatory effects. Mice possess four distinct IgG Fc receptors, including activating receptors FcγRI, FcγRIII, FcγRIV, and the inhibitory receptor FcγRIIB [51]. Surprisingly, the inhibitory FcγRIIB, that generally plays a suppressive role in the immune system, is indispensable for in vivo antitumor activities. However, in vitro, the activating FcγRs retained effective cross-linking effects, of which the strengths do not correlate with binding affinity. Together, these findings likely reflect the importance of the bioavailability of FcγRIIB in vivo as other activating FcγRs in vitro to enable cross-linking [24]. While FcγR-mediated crosslinking can be achieved, limited evidence supports efficient in vivo crosslinking effects for therapeutic purposes [85]. To maximize therapeutic potential, Fc affinity for FcγRIIB has therefore been optimized in monoclonal antibodies [85-88]. As expected, these Fc-enhanced anti-CD40 agonist antibodies were shown to induce robust antitumor efficacy both in vitro [86] and in vivo [88] and are currently being explored in early-phase clinical trials.

Unlike IgG1 monoclonal antibodies, IgG2 monoclonal antibodies impart agonistic and antitumor activities independent of FcγR [89]. Structural analysis revealed that the IgG2 isotype contains a more rigid hinge region due to differential disulfide bonding, contributing to improved clustering of CD40 receptors [89, 90]. The agonistic potency of IgG2 monoclonal antibodies was further corroborated by isotype switching both in vitro and in vivo. A clinically relevant human CD40 IgG4 antagonist antibody (bleselumab) bearing antagonistic epitopes was transformed in two formats: IgG1 and IgG2. While the IgG1 format retained its antagonism, the IgG2 format was converted to a superagonist, eliciting >4-fold CD8⁺ T cell expansion via receptor clustering in vivo. Although the underlying mechanism remains unknown, the conversion from an antagonist to a superagonist demonstrated that sufficient FcγR engagement could overcome antagonistic epitopes and improve antitumor activity [90].

More recently, it has been demonstrated that receptor epitope location combines with isotype to determine the biological activity of human anti-CD40 antibodies. As a TNFR, CD40 consists of 4 cysteine-rich domain (CRD) subunits, CRD1-4, that antibodies can selectively target to enhance agonist activity. Regardless of isotype difference and FcγR dependency, antibodies targeting the membrane distal CRD1 domain were shown to induce the highest agonistic activity, while those binding CRD2-4 display antagonistic properties. This is because antibodies that target epitopes distal to the membrane enable access to FcγRs domains with less steric hindrance [91].

Lastly, bispecific antibodies are rapidly growing as a new therapeutic for their ability to directly modulate the TME via targeting both CD40 and tumor antigens [92]. There are currently nine agonistic human anti-CD40 antibodies in clinical development: 4 IgG1 antibodies (APX005M, mitazalimab, 2141-V11, SEA-CD40), 2 IgG2 antibodies (selicrelumab and CDX-1140), one IgG mutant antibody (LVGN7409), and two bispecific antibodies (RO7300490, ABBV-428). This review also discusses other CD40 agonistic therapies currently in clinical stages, including one multidomain protein (MP0317) and three oncolytic viruses (NG-350A, LOAd703, MEM-288).

7.1. Selicrelumab

Selicrelumab (also known as CP-870,893, RO7009789) is a fully human IgG2 antibody that does not depend on crosslinking via the Fc domain to activate CD40 [93]. To date, selicrelumab is the most extensively studied CD40 agonist in clinical trials. It has been tested as monotherapy and in combination with radiotherapy, chemotherapy, immune checkpoint inhibitors (ICI), and anti-angiogenic therapy (reported in [77]). More recently, a phase 1 study evaluated neoadjuvant selicrelumab with or without chemotherapy (nab-paclitaxel and gemcitabine) before surgery in patients with pancreatic ductal adenocarcinoma (PDAC) (NCT02588443). The 1-year OS rate (Median + SE) was 81.8% + 11.8% for patients receiving selicrelumab monotherapy and 100% for patients treated by selicrelumab with chemotherapy, respectively. Selicrelumab was well-tolerated as related adverse events (AE) were mostly mild. Evidence of selicrelumab-induced TME modulation, including T-cell infiltration, DC maturation, macrophage repolarization, and decreased tumor stroma density, was found [94].

7.2. CDX-1140

CDX-140, a human IgG2 monoclonal antibody, drives NF-kB activation. It binds a domain that is outside the CD40L binding site, and its activation mechanism is independent of FcR cross-linking. It is shown to synergize with the naturally expressed CD40L [95]. Data from a phase I trial evaluating CDX-1140 with or without anti-PD-1 therapy (pembrolizumab) have shown a good safety profile at the dosage of 1.5 mg/kg of CDX-1140 in combination with 200 mg of pembrolizumab. Clinical benefits were observed, especially in patients with non-small cell lung cancer (NSCLC), since 80% (4 out of 5) evaluable patients experienced stable disease (SD) , while all had progressive disease (PD) on prior anti-PD-1/L1 based therapies (NCT03329950) [96].

7.3. APX005M

APX005M is a humanized rabbit IgG1 monoclonal CD40 antibody. With a point mutation in the Fc domain, APX005M has enhanced binding affinity with FcγRIIb, facilitating crosslinking of CD40 and mediating ADCC [86]. Multiple early-phase clinical trials evaluated APX005M in combination with immunotherapies. One phase II study evaluated APX005M in combination with chemotherapy (nab-paclitaxel and gemcitabine) with or without anti-PD-1 (nivolumab) in patients with metastatic pancreatic cancer. It was shown that APX005 did not significantly improve OS: in 105 patients, the primary endpoint of 1-year overall survival (OS) was met for nivo/chemo but not for APX005M /chemo or APX005M/nivo/chemo. However, survival after APX005M/chemo correlates with more significant intratumoral CD4⁺ T cell infiltration, circulating differentiated CD4⁺ T cells, and APCs [97] (NCT03214250). The first-in-human phase I study evaluated APX005M and an inhibitor of macrophage recruitment, colony-stimulating factor 1 receptor (CSF1R) inhibitor (cabiralizumab), with or without nivolumab in patients with melanoma, kidney cancer, or non–small cell lung cancer resistant to anti-PD-1/PD-L1. In this study, Weiss et al. (2021) demonstrated the safety and pharmacodynamic activity of the dual macrophage-polarizing therapy with or without nivolumab (NCT03502330) [98]. Recently, in a subsequent phase II study, Weiss et al. (2023) further investigated APX005M in combination with nivolumab in patients with metastatic melanoma following disease progression on anti-PD-1. The objective response rate (ORR) was 15%, and the median duration of response was at least 26 months, rendering the drug response safe, durable, and prolonged [99].

7.4. Mitazalimab

Mitazalimab (also known as ADC-1013 and JNJ-64457107) is a human IgG1 monoclonal antibody targeting CD40. It is engineered to have high potency and affinity in binding to CD40 in an FcγR-dependent manner, which benefits intratumoral administration to reduce dosage and adverse immune-related events [87]. Systemic administration of mitazalimab is also beneficial, although this method has been associated with lower antitumor efficacy and higher toxicities than intratumoral delivery [100]. In a phase I study in patients with advanced solid tumors, intravenous administration of 0.075-2 mg/kg of mitazalimab demonstrated a manageable safety profile when administered once every two weeks, both with or without corticosteroid pre-infusion [101]. Although the MTD and RP2D were not determined, the principal DLT reported was drug-induced liver injury (grade-3 ALT/AST + grade-2 bilirubin increase). Mitazalimab exhibits a favorable pharmacokinetic (PK) profile, as demonstrated by target-mediated drug disposition and rapid decline in serum concentrations [102]. Clinical, PK, and pharmacodynamics (PD) assessments demonstrate broad physiologic activation of CD40 after intravenous administration. In fact, through analyzing cell-free RNA from the whole blood of pre- and post-mitazalimab-treated patients, significant CD40-related transcriptional activity was identified in patients without corticoid pre-treatment [103]. Gene signatures related to the activation of B cell and myeloid/DC were upregulated post-treatment, confirming the mitazalimab-triggered CD40 agonism [103] (NCT02829099).

7.5. 2141-V11

2141-V11 is a fully human IgG1 CD40 antibody, a Fc-optimized version of selicrelumab in which the human IgG1 Fc is engineered with 5-point mutations that enhance Fc binding to FcγRIIB [88]. A phase I study evaluated 2141-V11 in patients with solid tumors with metastasis to the skin and demonstrated signs of antitumor activity in both local and distant lesions (NCT04059588). Another phase I study has been investigating the safety and dosage of 2141-V11 monotherapy in patients with non-muscle invasive bladder cancer (NMIBC) who did not respond to standard treatment (NCT05126472). Local intravesical delivery is used to mitigate systemic toxicity [104]. Lastly, phase I study has been examining 2141-V11 in patients with recurrent glioblastoma. Interim data have demonstrated that intratumoral administration of 2141-V11 in combination with immunotoxin (D2C7-IT) via convection-enhanced delivery (CED) was safe and somewhat effective: among the 5 patients enrolled, early signs of tumor response were observed at one patient treated at 0.70 mg, and two patients treated at 2.0 mg (NCT04547777) [105].

7.6. SEA-CD40

SEA-CD40 is a fully human, non-fucosylated IgG1 monoclonal CD40 antibody derived from the normally fucosylated CD40 monoclonal antibody, dacetuzumab. Non-fucosylated antibodies have enhanced binding to the FcγRIIIa receptor, potentially improving the agonistic efficacy. A first-in-human phase I trial evaluated intravenous SEA-CD40 monotherapy in patients with advanced solid tumors and lymphoma. SEA-CD40 infusion demonstrated potent immune activation via inducing cytokines associated with immune activation and trafficking, including interferon (IFN)-γ-inducible protein-10 (IP-10), monocyte chemoattractant protein-1 (MCP-1), monokine induced by IFN-γ (MIG), and macrophage inflammatory protein-1b (MIP-1β). Moreover, trafficking of innate and adaptive immune cells was evident, specifically T cells (CD4⁺ and CD8⁺), NK cells, and monocytes. Optimal immune activation was reached at 10–30 μg/kg via intravenous delivery. However, minimal antitumor effect of SEA-CD40 was illustrated with 1.8% (1/56) partial response in a patient with basal cell carcinoma and 1.8% (1/56) complete response in a patient with follicular lymphoma [106]. The Coveler group examined SEA-CD40 in combination with chemotherapy (nab-paclitaxel and gemcitabine) and pembrolizumab in patients with metastatic PDAC. Objective responses were observed in 48% of patients at 10 μg/kg and 38% at 30 μg/kg. The median duration of response was 5.7 months for both dosage groups. This might indicate that the higher dosage of 30 μg/kg can lead to adverse events in patients who are less tolerated, counteracting the potentially higher antitumor outcomes. Overall, the combination of SEA-CD40 at 10 μg/kg with chemotherapy and pembrolizumab demonstrates an acceptable safety profile and evidence of antitumor activity in patients with PDAC (NCT02376699) [107].

7.7. LVGN7409

LVGN7409 is an IgG mutant recombinant monoclonal CD40 antibody with an engineered Fc segment to bind FcγRIIB selectively. LVGN7409 activates CD40 in an Fc-FcγRIIB cross-link dependent manner and thus operates optimally in CD40 and FcγRIIB enriched TME [108]. Currently, there are two phase I studies of LVGN7409 in dose-establishment phases in patients with locally advanced, metastatic, or recurrent/refractory malignancy. The first-in-human phase I study of LVGN7409 has been evaluating LVGN7409 with or without anti-PD-1 antibody (LVGN3616) and/or CD137 agonist antibody (LVGN6051). Intravenous LVGN7409 monotherapy has been well tolerated up to 2 mg/kg (NCT04635995).

7.8. Bispecific antibodies

Bispecific antibodies can target both CD40 and tumor-associated antigens, therefore minimizing off-target effects by selectively modifying the TME.

RO7300490 is a novel bispecific antibody targeting CD40 and fibroblast activating protein (FAP), the latter of which is often over-expressed in tumor tissues and low in healthy tissues. Thus, RO7300490 enables direct remodeling of the TME while potentially avoiding toxicities associated with systemic CD40 activation [85]. Currently, one phase I study is evaluating RO7300490 with or without atezolizumab in patients with advanced solid tumors (NCT04857138).

ABBV-428 is another bispecific antibody that is designed to achieve localized activation of CD40 by interacting with the tumor antigen mesothelin, a cell-surface molecule that in cancers is highly expressed on epithelial mesotheliomas and carcinomas of the pancreas, lung, and ovary [109]. ABBV-428 targets mesothelin via a C-terminal single-chain variable fragment flanking Fc-modified human IgG1 and CD40 via an N-terminal single-chain variable fragment. Simultaneous binding to mesothelin and CD40 is required to activate CD40, leading to potential APC activation and immunomodulation of the Treg and M2-like macrophage in the TME [109]. The first-in-human phase I study evaluated ABBV-428 with or without nivolumab in patients with advanced solid tumors (NCT02955251). From interim data of patients receiving ABBV-428 monotherapy, the best clinical response was SD in 36% of patients treated at the recommended phase II dose (3.6 mg/kg). Toxicities of CD40 agonism, such as cytokine release syndrome were not observed, demonstrating an acceptable safety profile. However, ABBV-428 produced minimal clinical activity as there were no consistent changes from baseline in intratumoral CD8⁺ T cells, programmed death ligand-1 (PD-L1⁺) cells, or immune-related gene expression detected post-ABBV-428 treatment [109]. This could be explained by the low mesothelin expression on tumor cells, which would limit tumor accessibility and crosslinking of CD40 by ABBV-428 [110].

7.9. Others

7.9.1. MP0317

A novel therapy, MP0317, is a multidomain-designed ankyrin repeat protein (DARPin^®) agonistic to CD40. DARPin^® drug candidates are created based on naturally occurring ankyrin repeat domains, which possess properties ideal for uses as scaffold drugs, such as high thermal-dynamic, storage stability, and high solubility. MP0317 is exclusively active in the presence of both FAP and CD40 and thus is expected to preferentially activate T cells in the TME and avoid systemic toxicities [111]. The first-in-human phase I study of MP0317 (NCT05098405) aimed to evaluate its safety/tolerability profile, PK/PD characteristics, and recommended dose for combination therapy in patients with advanced solid tumors. Interim clinical data have confirmed that MP0317 is safe and well-tolerated. However, the lack of proinflammatory circulating cytokines (TNFα, IL-2, IL-6, and IL-8) raised a concern for its antitumor efficacy. However, evidence of colocalization of MP0317 with FAP and CD40 from paired pre- and on-treatment tumor biopsies suggested tumor-localized CD40 interaction [112].

7.9.2. Oncolytic Viruses

Oncolytic virus immunotherapy (OVT) utilizes natural or genetically modified viruses to selectively propagate in cancer cells, leaving healthy cells unharmed. OVT acts by two principles: it directly infects and lyses tumor cells, subsequently triggering antitumor immunity by activating innate and adaptive immune responses. Oncolytic viruses were encoded to express costimulatory molecules that are needed for T cell activation to enhance their antitumor effects [113]. Oncolytic adenovirus (OAd) is one of the most frequently employed OVTs, which utilizes adenovirus, a non-enveloped, double-stranded linear DNA virus that binds and infects tumor cells via the adenoviral fiber knob [114]. This review focuses on one CD40 agonist OAd currently in clinical trials, NG-350A.

NG-350A is an adenoviral tumor-specific immunogene vector that expresses a potent, fully human agonistic anti-CD40 antibody. The tumor-specific immunogene vector bears transgenes selectively to tumors and was clinically validated for intravenous delivery. This platform enables NG-350A to selectively replicate in tumor cells [115]. NG-350A re-educates the immunosuppressive TME while avoiding systemic toxicity [116]. The first-in-human phase 1a study evaluated NG-350A with or without pembrolizumab in patients with metastatic/advanced epithelial tumors (NCT03852511). No objective responses were observed; however, SD was found in 50% of patients treated with NG-350A monotherapy. The treatments were well-tolerated, and elevated peak cytokine responses higher than those of systemic anti-CD40 Abs have demonstrated local immunological tumor impact without CD40-related systemic toxicity [117].

Challenges remain for the success of OAd-based cancer therapies. Scant evidence has shown that OAd could effectively eliminate the tumor stem cells, in which their self-renewal capacity and differentiation potential have contributed to the metastasis of tumor cells and also resistance to radiation or chemotherapy [118].

8. Toxicity

While promising in principle, CD40 agonists have been associated with different levels of toxicities both in the clinic and animal models. As shown in Table 2 of recent clinical trials, several CD40 agonistic antibodies with abundant clinical records – CDX-1140 [96], APX005M [119], mitazalimab [120] – were reported to be well tolerated, with most treatment-related adverse events (TRAE) being grade 1 or 2. In two phase I trials with intravenously delivered selicrelumab and LVGN7409, respectively, CRS and alterations in liver function (elevation in ALT and AST) were transient [94, 121].

Table 2.

Clinical studies of CD40 agonistic antibodies in cancer treatment, alone or in combination with other treatments.

Therapy	Interventions	Route of Delivery	Phase	Cancer Type	Study/ NCT number	Status	Dosage (mg /kg)	Outcome	Side Effects
Selicrelumab	w/w.o nab-paclitaxel and + atezolizumab	Intra venous,(neo	I	PDAC	NCT02588443 [94]	Completed	0.2	OS=23.4 mo, median DFS=1	Neoadjuvant therapy (n=16): grade 1 CRS (56%). Adjuvant therapy (n=13): grade
		Intratumoral	Ib	B cell lymphoma	NCT03892525	Terminated	N/A	N/A	N/A
CDX-1140	w/w.o pembrolizumab	Infusion	I	Advanced solid tumor	NCT03329950 [96]	Completed	0.72 or 1.5 Q3 W	CR (4%) SD (36%)	At 1.5 mg/kg dose level (n=21): arthralgia (62%), fatigue (62%), nausea (48%), diarrhea (48%), vomiting (43%), myalgia (43%), fever (38%), chills (38%), AST increase (38%), bilirubin increase (24%), ALT increase (19%), CRS (19%)
	+ pembrolizumab	Infusion	I/Ib	Epithelial cancer	NCT04520711	Completed	N/A	N/A	N/A
	+ TCRT + radiotherapy + CDX-301 + Poly-ICLC	Intratumoral	I	Metastatic solid tumors	NCT04616248	Completed	N/A	N/A	N/A
	+ CDX-301	Intravenous injection	I	TNBC	NCT05029999	Recruiting	1.5 Q4 W	N/A	N/A
	+ odetiglucan	N/A	I	PDAC	NCT05484011	Recruiting	0.7 2 or 1.5 Q3 W	N/A	N/A
	+ bevacizumab + pembrolizumab + capecitabine + oxaliplatin + pembrolizumab	N/A	II	Ovarian cancer	NCT05231122	Not yet recruiting	N/A	N/A	N/A
		Intravenous	I/II	Biliary tract cancer	NCT05849480	Not yet recruiting	0.36-1.5 Q3 W	N/A	N/A
APX 005 M	+ nab-paclitaxel and gemcitabine w/w.o nivolumab	Intravenous	II	Pancreatic cance	NCT03214250 [97]	Completed	0.1 and 0.3 Q4 W	Primary endpoint of 1yr OS met for nivo/chemo (57.7%) but not for sotiga/chemo (48.1%) or sotiga/nivo/chemo (41.3%)	n=105: CRS, infusion reactions, thrombocytopenia and elevated liver function (87% total). CRS in nivo/chemo (0%), sotiga/chemo (24%) and sotiga/nivo/chemo (34%)
	+ domvanalimab + zimberelimab	Infusion	Ib/II		NCT05419479	Recruiting	N/A	N/A	N/A
	monotherapy	N/A	I	Pediatric braintumors	NCT03389802	Ongoing	0.1 Q3 W	N/A	N/A
	+ cabirali zumab w/w.o nivolumab	Intravenous infusion	Metastatic melanoma, Kidney cancer, NSCLC	NCT03502330 [98]	Ongoing	0.03 Q2 W	SD (8%) DP (62%)	n=26: asymptomatic elevations of lactate dehydrogenase (100%), creatine kinase (96%), AST (96%), and ALT (73%); periorbital edema (65%); fatigue (50%); DLT of acute respiratory distress syndrome (3.8%)
	w/w.o radiotherapy	Infusion	II		NCT04337931	Completed	N/A	N/A	N/A
	+ nivolumab	N/A	II	Metastastic melanoma	[99]	Ongoing	0.3 Q3 W	ORR(15%) PR(13%) SD(36.8%)	n=38: at least one AE (89%); grade 3 AE related to APX005M (13%): pyrexia, chills, nausea, fatigue, pruritus, elevated liver function, rash, vomiting, headache, arthralgia, asthenia, myalgia, and diarrhea
Mitazalimab	w/w.o corticos teroid pre-infusion	Intravenous	I	Advanced solid tumor	NCT02829099 [102]	Completed	0.075-2 Q2 W	Mitazalimab has shown to be safe and well tolerated at at doses up to 1.2 mg/kg	n=95: gatigue (44.2%), pyrexia (38.9%), pruritus (38.9%), chills (27.4%), and headache (26.3%). grade-3 or higher TEAEs (56.8%): gamma-glutamyl transferase increase (7.4%), AST increase (5.3%), and anemia (5.3%)
	+ mFOLFIRINOX	Intravenous	Ib/II	mPDAC	NCT04888312 [120]	Ongoing	0.45 - 0.9	ORR (52.2%)	n=5, grade 1 or 2 AEs (80%)
	+ MesoPher + modified FOLFIRINOX	Intravenous after chemo	I	Metastic pancreatic cancer	NCT05650918	Rrecruiting	0.075-0.15 - 0.3-0.6 or 1.2	N/A	N/A
2141-V11	monotherapy	Intratumoral	I	Solid tumor	NCT04059588 [128]	Completed	N/A	N/A	N/A
	monotherapy	Intravesical instillation	I	Nonmuscle invasive bladder cancer	NCT05126472 [104]	Recruiting	0.012-0.033-0.12 - 1.2 Q1 W	N/A	N/A
	+ D2C7-IT	Intratumoral, convection enhanced delivery (CED)	I	Malignant glioma	NCT04547777 [105]	Recruiting	0.012, 0.033, 0.12 0.35	Early signs of tumor response (25%)	n=8: headache (grade 3=12.5%; grade 2=25%); paresthesia (grade 3=12.5%; grade 2=12.5%); dysphasia (grade 3 =12.5%); pyramidal tract disorder (grade 3=12.5%; grade 2=12.5%); and depressed level of consciousness (grade 2=12.5%)
SEA-CD40	monotherapy	Intravenous injection	I	Advanced solid tumors, Lymphoma	NCT02376699 [106]	Terminated	0.0006, 0.003, 0.01, 0.03, 0.045, 0.06	0.03mg/kg was well tolerated	n=67: infusion/hypersensitivity reactions (73%)
	+ nab-paclitaxel and gemcitabine + pembrolizumab	N/A	I	mPDAC	NCT02376699 [107]	0.01 or 0.03 Q4 W	OR at 0.01 mg/kg (48%) OR at 0.03 mg/kg (38%)	n=61: fatigue (84%), nausea (74%), and neutropenia (67%); grade ≤3 TEAEs: neutropenia (61%), anemia (33%), and thrombocytopenia (20%); TEAE-caused discontinuation of treatment (10%)
	+ carboplatin + pemetrexed + pembrolizumab	N/A	II	NSCLC Melanoma	NCT04993677	Ongoing	N/A	N/A	N/A
LVGN7409	w/w.o LVGN3616 w/w.o LVGN3616 + LVGN6051	Intravenous infusion	I	Locally advanced, metastatic or recurrent/refract orymalignancy	NCT04635995 [121]	Recruiting	0.01-0.1, then conventional “3 + 3” design from 0.3	SD(44%)	n=12: infusion related reaction (58%), amylase increased, lipase increase, ALT increase and AST increase (42% each), bilirubin increase, WBC or neutrophil decrease, nausea, and cytokine release syndrome (8.3% each); grade 3 TRAEs (17%): amylase/lipase increase and ALT/AST increase (8.3% each)
	monotherapy	Intravenous infusion	I		NCT05152212	Recruiting	N/A	N/A	N/A
RO7300490	w/w.o atezolizumab	I	Intravenous	Advanced solid tumor	NCT04857138	Recruiting	N/A	N/A	N/A
ABBV-428	w/w.o atezolizumab	Intravenous infusion	I	Advanced solid tumor	NCT02955251 [109]	Completed	0.01-3.6	SD at 3.6 mg/kg (36%)	n=59: infusion-related reactions (12%)

PDAC: pancreatic ductal adenocarcinoma; mPDAC: metastatic pancreatic ductal adenocarcinoma; TNBC: triple-negative breast cancer; NSCLC: non-small cell lung cancer; OS: overall survival; CR: complete response; PR: partial response; SD: stable disease; DFS: disease-free survival; AE: adverse events; TRAE: treatment-related adverse event; CRS: cytokine release syndrome; ALT: aminotransferase; AST: aspartate aminotransferase; WBC: white blood cells.

In contrast, several clinical trials have reported severe side effects, specifically grade 3 or higher treatment-emergent adverse events (TEAEs) and non-transient alterations in liver function. In a terminated phase I study of intravenous SEA-CD40 (0.01 or 0.03 mg/kg) in combination with chemotherapy and pembrolizumab in patients with PDAC, grade 3 or higher TEAEs occurred, including neutropenia (61%), anemia (33%), and thrombocytopenia (20%). TEAEs that lead to treatment discontinuation were reported in 10% of patients across two dosage groups [107]. In another phase I study of mitazalimab with or without corticosteroids in patients with advanced solid tumors, grade-3 or higher TEAEs were reported in 56.8% and 60% in cohorts with and without corticosteroids, respectively. Other grade-3 or higher TEAEs include gamma-glutamyl transferase increase (7.4%), AST increase (5.3%), and anemia (5.3%) [102].

To control toxicity, clinical methods have been modified. Firstly, local intratumoral delivery of CD40 agonist antibodies has been utilized to induce an abscopal effect [122], potentially reduce toxicity while retaining the antitumor effect [100, 123]. Recent studies using intratumoral delivery include treating malignant solid tumors with 2141-V11 (NCT04616248) and CDX-1140 (NCT04059588), respectively, as well as treating malignant glioma (MG) with 2141-V11 by convection-enhanced delivery (CED) (NCT04547777) [105].

Another possible approach to down-tuning treatment-induced toxicity is the sequence of treatments between CD40 agonism therapy, chemotherapy, and surgery. Using mouse models of PDAC, Byrne et al. showed that CD40 agonist antibody was lethal in a CSF1R-dependent manner when administered 48h before chemotherapy [9]. Moreover, delivering chemotherapy five days after CD40 antibody was shown to allow time for tumor stroma disruption, improving subsequent tumor killing [124]. Therefore, an alternative schedule between chemotherapy and CD40 agonists has been utilized in the clinic; for example, chemotherapy would be given on days 1, 8, and 15, and selicrelumab on day 3 [94]. This chemo-CD40 agonist antibody – chemo “sandwich” dosing schedule also aligned with previous preclinical [9] and clinical [125] studies, as well as current trials with SEA-CD40 [107], CDX1140 (NCT05849480), and APX005M [97]. Furthermore, the sequence with surgery was shown to mediate toxicity. In a phase 1 study in patients with PDAC, both neoadjuvant and adjuvant selicrelumab with or without chemotherapy were evaluated. In the neoadjuvant group, selicrelumab was given two days after chemotherapy and before surgery [94]. TRAEs were primarily mild at grade 1 or 2, with no CRS or serious adverse events observed. However, the adjuvant therapy was associated with grade 2 CRS, a series of grade 3 or higher AEs, and severe adverse events. The mechanism behind these AEs remains to be discovered.

While dosages selected for clinical trials vary from 0.01 mg/kg to 3.6 mg/kg (most frequently at 0.01-0.3 mg/kg) depending on the CD40 agonist used (Table 2), in vivo mouse studies using CD40 rat anti-mouse IgG2a mAb (clone FGK45) have commonly used 1.5-10 mg/kg (30-200μg/dose), with 5 mg/kg (100μg/dose) being the most frequently employed dosage across treatment groups (Table 3) [126-130]. The minimum dosage of 1.5 mg/kg in mouse studies converts to exactly 0.01 mg/kg in human after metabolic normalization, validating the safety of initial dosages used in current clinical studies. However, a lower dosage comparable to that in the clinic at 0.1 mg/kg occurred in one humanized mouse model, which carries human Fcγ receptors (FcγRs) and human CD40 (hFcγR/hCD40) instead of their mouse homologs. The CD40 agonist used was 2141-V11. As mentioned above, it is a fully human anti-CD40 agonist antibody with enhanced binding to FcγRIIB. Establishing the effect of FcγRIIB enhancement under physiological conditions, this mouse model was able to mirror and predict dose-limiting toxicities occurring in patients. While the MTD of the parental antibody 2141-IgG2 was found to be 0.2 mg/kg, doses higher than 0.1 mg/kg of 2141-V11 led to profound transaminitis and hepatotoxicity, and evidence of intravascular thrombi and hepatocyte necrosis was also found in mice treated with concentrations equal or above 0.25 mg/kg. This model also provides insights into the constraints of using anti-CD40 agonistic antibodies in humans – when administered systematically, they may struggle to attain an optimal therapeutic concentration due to dose-limiting toxicities [128].

Table 3.

Preclinical studies of CD40 agonistic antibodies in cancer treatment, alone or in combination with other treatments.

Therapy	Interventions	Route of Delivery	Cancer Type	Dosage (mg/kg )	Outcome	Side Effects	Reference
Anti-mouse CD40 mAb (FGK45)	w/w.o nab-nabpaclitaxel and gemcitabine (CD40/Gem or Gem/CD40) w/w.o anti-CSF-1R	IP	PDA	5 or 15	No reduction in tumor volume. aCD40 was lethal when administered before chemotherapy (40%). Lethal hepatotoxicity was abrogated when macrophage activation was blocked using anti-CSF-1R mAb	42% mice treated with 0.3mg/kg followed by Gem died vs. 58.1% in the 0.1 mg/kg group. In all mice: multifocal granulomatous and neutrophilic infiltration; significant increase in ALT and AST. Higher hepatotoxicity in CD40/Gem group.	[9]
	Diphtheria toxin (DT) (Treg depletion) w/w.o TNF neutralization	IP	Mammary carcinoma (4T1.2) and Colon adenocarcinoma (MC38)	5	Mice with Treg depletion had higher TNF and IL-6 production. Concomitant TNF neutralization may reduce irAEs but also antitumor efficacy	DT + aCD40: significant weight loss, increase in clinical score, ALT and the proinflammatory cytokines. Concomitant anti-TNF abrogated weight loss, liver damage and colitis, resulting in an improved clinical score.	[126]
	+ Radiotherapy + PDL1	NDES or IP	Mammary carcinoma (4T1)	5	Localized immunotherapy via NDES + radiotherapy demonstrated an abscopal effect, augmented CD8+ T cells	IP vs. NDES CD40/PDL1: significant weight loss, worsened pathologic liver inflammation.	[127]
	w/w.o anti-PD-1, anti-CTLA4	IP	Colon adenocarcinoma (MC38)	5	IL-12 and IFN-γ interdependently drive tissue-damaging effects in the liver and other tissues in a neutrophil-dependent way	IrAEs at liver and the GI tract. Portal and prominent lobular hepatitis and broad, confluent areas of necrosis. Colon crypt hyperplasia.	[131]
	+ IL-2 w/w.o T-cell adoptive transfer	IP	Melanoma (B16) and Colon adenocarcinom a (MC38)	10	CD40 agonist expanded transferred T cell in vivo and improved antitumor effects	N/A	[130]
	monotherapy	IP	Melanoma (B16-OVAMO4)	1.5	Antitumor activity was significantly higher for mice treated with CD40 antibody with enhanced FcγRIIB binding	N/A	[51]
	monotherapy	IP	Breast cancer (4T1) and Hepatocellular carcinoma (RIL-175)	5	CD40 agonist matured murine and human hepatic MDSCs, relieving immunosuppression.	Liver damage occurred within 24h: ALT and AST elevations, and ROS-mediated hepatotoxicity.	[129]
	monotherapy	subcutanous or peritumoral or intravenous	Bladder cancer (MB49)	0.5, 1.5, 5	local peritumoral at low dose (0.01 and 0.03 mg.kg) improved survival and reduced toxicity than the intravenous systemic treatment		[100]
Human anti-CD40 agonist antibody	monotherapy	Intratumoral injectionor IP	Colon adenocarcinoma (MC38)	0.1	Intratumoral delivery led to a significantly lower tumor burden.	Dosage > 0.1 mg/led to profound increase in AST, ALT, and hepatotoxicity. Increasing dose worsened thrombocytopenia.	[128]

NDES: nanofluidic drug-eluting seed; IrAE: immune-related adverse events; ALT: aminotransferase; AST: aspartate aminotransferase

To improve antitumor efficacy while minimizing toxicities, in vivo studies have been exploring tumor-targeted CD40 agonism. Our group has shown that the inhibition of RAS/PI3K/RAF-mediated pathways directly induces CD40 expression in melanoma cells, triggering immunogenic cell death [13]. Recently, a novel bispecific CD40 agonistic antibody named CEA-CD40 was developed. It stimulates CD40 selectively in the presence of carcinoembryonic antigen (CEA), a glycoprotein exclusively expressed on tumor cells, thereby activating the delivery of tumor antigen derived from CEA⁺ tumor-derived extracellular vesicles by DCs to trigger tumor-specific T cell cross-priming [131]. To further reduce systemic toxicities, in vivo studies have shifted to evaluate the effect of local delivery on the toxicity and antitumor effect. Subcutaneous delivery [100], intratumoral injection [128], and nanofluidic drug-eluting seed (NDES) [127] were shown to allow lower toxicity and better efficacy than intraperitoneal injection (Table 3). Using mouse models, recent studies have unraveled the key players mediating CRS and liver toxicity: Kupffer cells, neutrophils, macrophages, and platelets. Mechanistically, tissue-resident Kupffer cells sense lymphocyte-derived IFN-γ and produce IL-12 [132]. IFN-γ and IL-12 subsequently stimulate neutrophils that in turn upregulate and secrete TNF [133], and cause granulocytic inflammation, which results in liver toxicity. It has been further shown that CD40-mediated hepatotoxicity is associated with an accumulation of triple-positive (MHCII⁺, CD14⁺, and CD11b⁺) macrophages in the liver in an IL-12p40-dependent manner [132]. Last but not least, a causal relationship between liver macrophages and platelets has been described in anti-CD40 treatment-related hepatotoxicity, whereby the FcγRIIB expression by Kupffer cells might lead to crosslinking of CD40 antibodies and subsequent platelet activation [134]. Therefore, to reduce toxicity, mechanism-based interventions have been employed, such as blocking macrophage activation [9] and neutralizing TNF [126]. However, antitumor efficacy could be compromised with this approach [126]. OVT “armed” with different adjuvants can be used to enhance antitumor efficacy while minimizing systemic toxicity. Arming oncolytic viruses with T cell costimulatory molecules, such as CD40, OX40, ICAM-1, and B7-1, can enhance APC function. In addition, a variety of cytokines that modulate antitumoral responses, such as GM-CSF, IL-2 and IFN-γ, have been armed with oncolytic viruses to promote proapoptotic lymphocyte functions in the TME, and have demonstrated higher antitumor efficacy with no or repressed systemic toxicity in the clinic [113].

9. Future directions

Several strategies can be incorporated with CD40 agonism therapy in cancer treatment, including cytokine therapies, targeted therapies of the RAS/RAF/PI3K pathway, and chimeric antigen receptor (CAR) T therapies.

9.1. Cytokine therapy

CD40 agonism can be combined with cytokine therapies targeting the MDSCs, macrophages, or NK cells, potentially mounting complementary and synergistic antitumor effects. For example, the C-X-C chemokine receptor 2 (CXCR2) signaling axis plays an essential role in the recruitment of neutrophils and immunosuppressive MDSCs into the TME [135]. The deletion of CXCR2 in a Braf^V600E/PTEN^−/− melanoma mouse model skewed the TME toward antitumor immunity via reducing MDSCs and promoting CD8⁺ T cell infiltration via B-cell derived C-X-C Motif Chemokine Ligand 11 (CXCL11) [136, 137]. A CXCR2 antagonist, such as SX-682, was reported to significantly decrease the expression of chemokines associated with MDSC recruitment and suppress progenitor cells via upregulating Tfcp2l1, a transcription factor regulating tumor suppression and stemness [137]. By combining CXCR2 antagonism with CD40 agonism, multiple aspects of the TME can potentially be complementarily targeted, with the former inhibiting the recruitment of immunosuppressive cells and the latter enhancing immunosurveillance.

Macrophage-based cytokine therapies can be further expanded to combine with CD40 agonism. Most TMEs are encompassed with the immunosuppressive tumor-associated macrophages (TAM) of the M2 type, which in several cancer types has been associated with the expression of the immunoregulatory IL-10 [138], promoting tumor growth and correlating with poor outcomes [139]. Therefore, the key to macrophage-based therapies has been modulating or reversing the M2 macrophages towards the M1 macrophages to overturn the immunosuppressive milieu [139]. One example that acts by such a mechanism is the clinically relevant CSF-1 inhibitor, as mentioned above [119]. Blockade of the immunosuppressive IL-10 or TGF-β may also modulate TAMs, decrease tumor stemness [140], and restore their antitumor effect, specifically by promoting IL-12 expression and activating type I antitumor immune responses [141]. Combining macrophage-modulating cytokine therapy with CD40 agonism may be beneficial, firstly because of their complementarily mechanisms of action, as reducing immunosuppressive TAMs while increasing immunosurveillance. Secondly, a perpetuating synergistic effect may arise since CD40 ligation on macrophages can activate their production of proinflammatory cytokines, including IFN-γ, which upregulate CD40 expression on the macrophages and lead to downstream autocrine tumoristatic effects [32].

CD40 agonism may also be combined with cytokine therapies targeting NK cells, which are expressed at low frequency in tumors compared to myeloid and major lymphoid cells [142]. Interleukin-2 (IL-2) and interleukin-5 (IL-5) are two stimulatory molecules for NK cells [142]. Clinical data have shown that the recombinant forms of IL-2 [143] and IL-5 [144] can induce robust activation of CD8⁺ T cells and NK cells within the TME in patients with solid tumors. Combining NK cell therapy that enhances the direct killing of tumor cells with CD40 agonism can activate both the innate and adaptive immune systems, thus mounting a more comprehensive antitumor response. In addition, a synergistic effect may result since activated NK cells produce IFN-γ and TNF, which in turn, upregulate CD40 expression on macrophages and promote type I antitumor immune responses [32].

The anti-apoptotic protein TNF-A IP3 (A20) was reported to protect malignant B cells against cell death under CD40 stimulation [59]. IL-17 signal is known to maintain the baseline production of the anti-inflammatory TNF-Α IP3 (A20) and restrain the aberrant activation of JNK and NF-kB pathways in melanoma and breast cancer cells [145]. In CRC, IL-17 receptor deletion correlates with tumor invasion, growth, metastasis, and poor clinical outcome [146]. The paradoxical role of IL-17 and IL-17-producing Th17 cells in cancers may interfere the therapeutic outcome of CD40 agonism [147], and thus, finetuning the IL-17/A20 signaling with CD40 agonism therapy is intriguing for future investigation.

9.2. Targeted therapy of the RAS/RAF/PI3K pathways

CD40 agonism can also be combined with targeted therapy in cancer patients exhibiting mutations in the RAS/RAF/PI3K pathways. RAS mutations occupy approximately 19% of all cancers, among which KRAS is the most frequently mutated isoform [148]. RAS-mutant tumors are thought to be highly aggressive and recurrent, correlating with specific clinicopathological features, such as the shifted location of metastasis, lower tumor-infiltrating lymphocyte grade, and shorter OS in certain types of cancers. Therefore, the inhibition of RAS – most frequently directly or through downstream proteins – has become a burgeoning field. A KRAS^G12C inhibitor named AMG510 exhibited encouraging results both in vivo [149] and in patients with NSCLC and colorectal cancer [150]. In breast cancer where the RAS-downstream PI3K/mTOR pathway is frequently altered, our group showed that PI3K/mTOR inhibition (gedatolisib) plus microtubule poison paclitaxel (PTX) halted the PyMT breast tumor growth and improved response to ICB [151]. In addition, dual inhibition of the PI3K/AKT/mTOR and the RAF-downstream MAPK/MEK/ERK pathway was mechanistically warranted: both being downstream of receptor tyrosine kinase (RTK) activation of RAS, such that the inhibition of only one pathway would cause compensatory enhancement of the other, offsetting ideal tumor suppression [152]. Therapeutic effects of the dual inhibition of PI3K and MAPK pathways were also confirmed by preclinical studies in breast cancer and melanoma models [152]. The combination of CD40 agonism with targeted therapies of the RAS/RAF/PI3K pathways, specifically of MEK, was reported to produce a synergistic effect without impeding T cell functions in murine melanoma (B16-OVA) and colon cancer (MC-38). Mechanistically, the MEK inhibitor (GDC-0973) produced pro-immunogenic effects by inhibiting Tregs, M2 macrophages, and MDSCs, while the CD40 agonist complementarily promoted antigen presentations of DC, CD8⁺ T cells and M1 macrophages [153]. Such combinative therapies shed light on the personalization of treatment tailored to cancer patients harboring RAS/RAF/PI3K mutations.

9.3. CAR T therapy

Engineered CD40 agonists with selective binding to tumor cells can be combined with CAR T therapies. Adoptive transfer of CAR T cells with constitutive expression of CD40L was reported with robust antitumor efficacy by licensing APCs, which then mobilized endogenous tumor-recognizing T cells and enhanced cytokine productions [154]. The addition of the tumor-specific CD40 agonists may complement the proinflammatory responses by CAR T cells by inducing direct apoptosis of tumor cells, creating a synergistic antitumor effect. It would be intriguing to explore whether CD40 agonists lacking selective affinity for tumor cells might bind and augment the recruitment of CAR T cells. Alternatively, there's the question of whether CD40 agonists could saturate CD40L-expressing CAR T cells, potentially diminishing their effectiveness.

9.4. Advanced methodologies and models

CD40 agonism can be explored in various preclinical models with quantitative analysis to improve the prediction of drug responses in patients. The patient-derived xenograft (PDX) established in humanized mice was recently optimized for evaluating both targeted therapy and immune therapy: human tumor tissues were transplanted and expanded in the humanized mice that were established by the engraftment of CD34⁺ hematopoietic stem cells. The translational utility was maximized in this model since human tumor tissues were grown in humanized mice competent for mounting antitumor responses [155]. Moreover, high-throughput screening containing an extensive collection of PDXs with diverse mutations enabled insights into the association between genotypes and drug efficacy and facilitated the exploration of predictive biomarkers [156]. In addition, the prediction of clinical trial responses from preclinical models can be further improved by multiplex immunohistochemistry/immunofluorescence and spatial imaging analysis of cells in the TME, which visualize desired biomarkers in the TME of the tumor tissue section and perform quantitative assessment of their spatial distribution [157].

Since the successful generation of organoids about a decade ago, patient-derived tumor organoids have become a favorable preclinical model. Organoids are 3D-cultured in vitro from patient-derived tumor tissues in an extracellular matrix, enabling mimicking of the genetic, phenotypic, and histopathologic characteristics of the original tumor, such as in colon, lung and breast cancer [158]. However, the variation in the physical structures of organoids poses challenges to reproducibility, which is alleviated by organoid-on-a-chip models. These models culture organoids in specific scaffolds that mimic the physiological aspects of the target organ, and allow coculture with immune cells via microfluidics technology, thereby enabling a uniform, monitorable interplay between the organoid and the human-like physiological environment [158]. Bridging the gap between patient-derived tissues, animal, and cell models by simulating human biological conditions, these preclinical models combined with quantitative analysis would enable faster, more cost-effective testing of CD40 agonists and generate more reliable and personalized prediction of drug responses.

Conclusions

The intricate role of CD40-CD40L signaling in a variety of immune cells and cancer types underscore its significance as a therapeutic target in cancer treatment. While the clinical development of CD40 agonists shows remarkable potential, continued research and careful consideration of dosage optimization and adverse effects management are crucial for harnessing the full therapeutic potential of CD40 agonism in the fight against cancer.

Highlights.

• CD40L-CD40 signaling plays an intricate role in a variety of immune and malignant cells and the regulation of the tumor microenvironment.

• CD40 is linked to various biological and immunological functions.

• Preclinical studies and clinical development of CD40 agonism therapy underscore its significance in cancer treatment.

• Careful consideration of dosage optimization and adverse effects management is crucial for the future design of CD40 agonism cancer therapy.

• Several cancer treatment strategies, including cytokine therapies, targeted therapies of the RAS/RAF/PI3K pathway, and chimeric antigen receptor (CAR) T therapies, can potentially be incorporated with CD40 agonism therapy.

Abbreviations

Act1

activator 1

ADCC

antibody-dependent cellular cytotoxicity

AE

adverse event

AKNA

AT-Hook transcription factor

ALT

alanine aminotransferase

AST

aspartate aminotransferase

BAD

BCL2-associated agonist of cell death

CR

complete response

CRS

cytokine release syndrome

DFS

disease-free survival

EBF1

EBF transcription factor 1

ERK

extracellular signal-regulated kinase

IFNβ

interferon beta

IFNγ

interferon gamma

IKK

IκB kinase

IKKα

inhibitor of NF-κB kinase α

IrAE

immune-related adverse event

IκB

NF-κB inhibitor

JAK3

Janus kinase 3

JNK

c-Jun N-terminal kinase

LPS

lipopolysaccharides

MAPK

mitogen-activated protein kinase

mPDAC

metastatic pancreatic ductal adenocarcinoma

NDES

nanofluidic drug-eluting seed

NF-κB

nuclear factor kappa-B

NSCLC

non-small cell lung cancer

OS

overall survival

PDAC

pancreatic ductal adenocarcinoma

PI3K

phosphoinositide 3-kinase

PR

partial response

RPL26

ribosomal protein L26

RPL4

ribosomal protein L4

RPL8

ribosomal protein L8

RPS9

ribosomal protein S9

RUNX1

RUNX family transcription factor 1

RUNX3

RUNX family transcription factor 3

SAG

superantigen

SD

stable disease

STAT-1

signal transducers and activators of transcription 1

STAT-1α

signal transducers and activators of transcription 1α

STAT5

signal transducer and activator of the transcription 5

TAK1

TGFβ-activated kinase 1

TNBC

triple-negative breast cancer

TNFα

tumor necrosis factor α

TRAE

treatment-related adverse event

TRAF

TNF receptor-associated factor

WBC

white blood cells

Biographies

Yang Zhou

Yang Zhou is a senior undergraduate at Vanderbilt University, majoring in Molecular & Cellular Biology. Her research in the Lab at Vanderbilt's Department of Pharmacology focuses on investigating CD40 overexpression and targeted therapy in melanoma. She is a fellow of the SyBBURE Searle Undergraduate Research Program. She is also a Student Member of the American Association for Cancer Research. She serves as an Editorial Board Member of Vanderbilt's largest student-run STEM newspaper, Vanguard. She was also a News Editor Intern for the Chinese platform of Scientific American. With a deep passion for immunology, she aspires to pursue graduate studies in this field, aiming to contribute to bench-to-bedside translation.

Ann Richmond, Ph.D.

Ann Richmond is an Ingram Professor of Cancer Research, Director for the Program in Cancer Biology, Professor of Pharmacology and Dermatology at Vanderbilt University School of Medicine and at the TVHS Department of Veterans Affairs. Early in her career her lab identified and characterized one of the first chemokines, now known as CXCL1 and went on to characterize the role of this and other chemokines and their receptors in wound healing, inflammation, tumor growth and metastasis. Her research goals are to develop and translate new therapeutic approaches for treatment of malignant melanoma and breast cancer. With her research team she has developed new strategies for treatment of tumors that respond poorly to immune checkpoint inhibitors (ICI). They have shown that combining inhibitors of the PI3K/AKT pathway in breast cancer, or inhibitors of the RAS/PI3K/MEK pathways can significantly increase response to ICI therapy and in some instances result in tumor regression. She has been continuously funded by the NIH and the Department of Veterans Affairs throughout her career. Recent work from her group lead to an investigator initiated clinical trial to determine whether the small molecule inhibitor rigosertib will restore sensitivity to immune checkpoint inhibitors of tumors that have developed resistance to these therapies. She is a fellow of the American Association for Advancement of Science, was awarded the William S Middleton Award in 2016 for outstanding achievement in biomedical research by the Department of Veterans Affairs and she received the Society of Leukocyte Biology Legacy Award in 2019. Her research has been continuously supported by grants from the NIH and the Department of Veteran’s Affairs throughout her career. She has published over 200 peer-reviewed papers and trained over 50 graduate students and postdoctoral fellows over her career. She is currently the Director for the Program in Cancer Biology at Vanderbilt University School of Medicine. She serves on numerous editorial boards, NIH advisory boards, and grant review boards.

Chi Yan, Ph.D.

Dr. Chi Yan is an Assistant Professor in the Department of Pharmacology at Vanderbilt University School of Medicine, Nashville, TN, USA. He obtained his Ph.D. in Cancer Immunology from Dalhousie University, Canada, and his M.Sc. in Molecular Biology and Bioinformatics from Saint Mary’s University, Canada. Dr. Yan is a recipient of the prestigious Graduate Fellowship from CRTP-BHCRI in Cancer Research at CIHR, Canada, and the Graduate Fellowship from IWK Health Center, Canada. The goal of Dr. Yan’s research program is to better understand the interaction between tumor cells and immune cells so that one can improve current therapeutic strategies for the treatment of melanoma and breast cancer patients. Dr. Yan’s ongoing research at Vanderbilt University aims to enhance antitumor immune responses by combining personalized targeted and immune therapies. As an accomplished young scientist, Dr. Yan’s work has resulted in investigator-initiated clinical trial NCT05764395 and over 30 publications, including 24 first- or senior-authored articles, in highly-ranked journals that are considered among the top in the field, e.g., Cytokine & Growth Factor Reviews (2024), Molecular Cancer (June 2021, Nov 2021, and 2023), Cancers (2023), npj Precision Oncology (2022), Cancer Immunology Research (2021), and Journal of Clinical Investigation (2020) etc. In addition, Dr. Yan has peer-reviewed over 60 manuscripts for biomedical science journals, including but not limited to the prestigious Nature, Molecular Cancer, Journal of Clinical Investigation, Clinical Cancer Research, and Cancer Research etc. Dr. Yan has also served as Review Editor for Frontiers in Immunology, Frontiers in Oncology, Frontiers in Cell and Developmental Biology, and Frontiers in Pharmacology. Dr. Yan was invited to present his research work at local, national, and international conferences, including Innovation in Breast Cancer Symposium 2022 (Madrid, Spain & Virtual) and the AAI Annual Meeting - IMMUNOLOGY2022 with the awarded AAI Early Career Faculty Travel Grant, etc. Dr. Yan has recently been selected as the Vanderbilt University Early-Career Scientist nominee for the global 2024 Takeda Innovators in Science Award - Cancer Immunology. Dr. Yan received a Henry J Lloyd Foundation Melanoma Research Career Development Grant (PI), Vanderbilt-Ingram Cancer Center Support Grant (Co-PI), and is actively managing and serving as the Co-PI of two ongoing NIH R01 grants (CA116021 and CA243326). Dr. Yan served as the direct mentor of Yang Zhou (first author), and is the corresponding author of the published review in Cytokine & Growth Factor Reviews (2024).