DISRUPTING CANCER ANGIOGENESIS AND IMMUNE CHECKPOINT NETWORKS FOR IMPROVED TUMOR IMMUNITY

Abstract

Immune checkpoint inhibitors (ICIs) have advanced the field of cancer immunotherapy in patients by sustaining effector immune cell activity within the tumor microenvironment. However, the approach in general is still faced with issues related to ICI response duration/resistance, treatment eligibility, and safety, which indicates a need for further refinements. As immune checkpoint upregulation is inextricably linked to cancer-induced angiogenesis, newer clinical efforts have demonstrated the feasibility of disrupting both tumor-promoting networks to mediate enhanced immune-driven protection. This review focuses on such key evidence stipulating the necessity of co-applying ICI and anti-angiogenic strategies in cancer patients, with particular interest in highlighting newer engineered antibody approaches that may provide theoretically superior multi-pronged and safe therapeutic combinations.

GRAPHICAL ABSTRACT

Images created with Biorender.com.

Introduction

The tumor protective role of immune effector cells (such as cytotoxic CD8+ T cells) has historically been appreciated for some time, although the clinical translation of basic concepts worked out in preclinical models took decades longer to materialize from large-scale clinical trials [1]. Arguably, a larger perception of “success” surrounding cancer immunotherapy followed the appreciation that components of the tumor microenvironment (TME) instigated immunosuppression by engaging immune checkpoints (ICs) expressed by components of the immune system [2]. If the brakes (so to speak) of lymphocytes could be released (through IC inhibitors [ICIs]), anti-tumor T cell activation/proliferation could be sustained during early and late phase priming events in lymphoid tissues and the periphery, respectively, to better engage and destroy malignant cells [3]. Early clinical trials confirmed such suggestions through blocking the ICs CTLA-4 and PD-1 and eventually paved the way for the paradigm shifting immunotherapeutic strategies now commonplace in patients receiving FDA approved ICI-based regimens (as outlined in Figure 1).

Figure 1. Immune checkpoint blockade unleashes effector T cell activities in the tumor microenvironment.

(Panel A) A naïve T cell is first activated in a lymph node by receiving stimulatory signals (e.g., upon TCR:peptide/MHC and CD28:CD80/CD86 ligations) from an antigen-presenting cell. Activated T cells that infiltrate a tumor lesion are then capable of killing cancer cells in an antigen-specific manner. (Panel B) Cognate ligands engaging immune checkpoint molecules such as CTLA-4 or PD-1 on activated T cells will prevent the proper effector functions of these cells. (Panel C) The therapeutic use of antibodies that neutralize CTLA-4, PD-1, or PD-L1 binding will sustain T cell activation in the tumor microenvironment, which improves cytotoxic responses and protection against cancer growth. Abbreviations used: TCR, T cell receptor; MHC, major histocompatibility complex. Images created with Biorender.com.

In this review we focus on the major IC classes that work to subvert immune-based responses against cancer. In particular, although ICIs have provided select patients phenomenal clinical responses, further improvements to this generalized strategy are needed to improve patient eligibility, safety, and long-term efficacy. One such promising area involves optimizing ICIs with tumor vascular-disabling agents. Angiogenesis is certainly a well-known/-established impediment to therapeutics since unrestrained blood vessel development helps establish invasive and immunosuppressive tumor lesions. Yet, cancer angiogenesis appears suited to also fuel IC-based suppression of tumor infiltrating T cells, which opens new possibilities for novel therapeutic interventions.

Immune checkpoints (ICs)

The major IC classes on which FDA-approved therapeutics have focused, thus far, are CTLA-4, PD-1, and PD-L1 [4]. Under general circumstances, these ICs serve to prevent autoimmunity and limit the extent/duration of immune responses following antigen stimulation (e.g., during a viral infection) [5]. While immune regulation is an essential and important host function, expression and engagement of ICs on tumor cells and T cells can actually prevent immune-directed clearance mechanisms by inducing an immunosuppressive state within the TME that helps cancer lesions grow and progress [6].

CTLA-4

CTLA-4 (CD152) is a cell-surface receptor that is expressed on activated T cells. Following T cell receptor interaction with its cognate peptide/MHC complex and co-stimulation through CD28, intracellularly-stored CTLA-4 traffics to the cell surface to further suppress co-stimulation [7]. CTLA-4 binds CD80 (B7–1) and CD86 (B7–2)-bearing antigen presenting cells (APCs), which interrupts CD80/CD86 interaction with T cell-derived CD28. In addition to dampening these extrinsic activation signals, CTLA-4 also exerts inhibitory effects inside the T cell, such as abrogating transcription factors such as NF-kB, NFAT, and AP-1 that help govern T cell activation, proliferation, and cytokine production [8, 9].

PD-1

PD-1 (CD279) regulates T cell activation through interaction with PD-L1 (B7-H1) or PD-L2 (B7-DC). Following ligand binding, PD-1 attenuates T cell activation by recruiting the tyrosine phosphatase SHP2, which dephosphorylates nearby effector proteins (e.g., ZAP70, CD28) to consequently reduce cytokine production and T cell proliferation [10, 11]. CD28 appears to be a preferential target of SHP2, indicating that PD-1 and CTLA-4 exhibit similar molecular mechanisms of T cell attenuation, although their functions are non-redundant based on PD-1/CTLA-4 upregulation during distinct phases of T cell activation (early phase [CTLA-4] vs. late phase [PD-1]), location of cognate ligands (many cell types [PD-L1/L2] vs. mostly APCs [CD80/86]), as well as intracellular signaling pathways that are largely distinct [12, 13]. PD-1 is also crucial for the maintenance of peripheral T cell tolerance, as evidenced by the development of a lupus-like autoimmune condition in mice following deletion of Pdcd1, which encodes PD-1 [14].

PD-L1

PD-L1 is constitutively expressed by various immune cells including T cells, dendritic cells (DCs), B cells, and macrophages along with nonhematopoietic cells such as vascular endothelial cells, epithelial cells, hepatocytes, and astrocytes [15]. PD-L1 expression can also be induced by interferons (especially IFN-γ) that arise as a result of immune activity [16]. Therefore, given the widespread expression profile of PD-L1 outside lymphoid tissues, the PD-1/PD-L1 axis plays a crucial role in systemically regulating CD8+ and Th1 CD4+ T cell function in order to maintain peripheral tolerance against self antigens and prevent autoimmune pathologies [7]. Although PD-L1 is the principal ligand for PD-1, PD-L2 is solely expressed by APCs and also binds PD-1 (albeit with higher affinity than PD-L1) [17]. During a Th2-biased immune response, however, PD-L2 is upregulated and can prominently alter T cell activity. PD-L1 is also capable of binding the CD80 costimulatory ligand, demonstrating some level of duplicative properties between the CTLA-4 and PD-1/PD-L1 pathways [18].

Ultimately IC-based mechanisms are important for maintaining normal immune cell/tissue homeostasis, but these regulatory molecules are frequently co-opted by tumors for evasion of immune system defenses involving, for example, CD8+ T cells [19]. Tumor cells generally circumvent immune detection through various means including MHC class I downregulation, mutations that confer resistance to cell death induction, and recruitment of immunosuppressive stromal cells, with each situation requiring a unique therapeutic intervention to potentially overcome [20]. For effective and efficient immune responses against cancer cells, MHC-presented peptide plays a vital role in engaging antigen-specific T cells. Progressing tumors (65–90%, depending on the type of cancer) tend to lose expression of various MHC class I molecules that allows escape from the destructive power of CD8+ T cells [21]. Spontaneous mutations arising in tumor cells may also confer survival benefits despite the prevalence of immune cells and/or cancer therapies. For example, acquired beta-tubulin mutations can disrupt microtubule stability and garner resistance to chemotherapeutic agents like paclitaxel [22, 23]. The stromal cell composition of the TME can additionally influence the immunosuppressive nature of cancer lesions. Distinct immune cell subtypes such as regulatory T cells (Tregs), myeloid-derived suppressor cells (MDSCs), and tumor-associated macrophages (TAMs) are typically recruited to tumor lesions from the surrounding tissue or circulation and help fuel tumor growth and immunosuppression through the expression of pro-angiogenic molecules (e.g., vascular endothelial growth factor [VEGF]) and immunomodulatory compounds (like cytokines [e.g., IL-10, TGF-beta], reactive oxygen species, and IDO) [24]. However, a relatively more recent appreciation for tumor escape is the upregulation of ICs within the TME. CTLA-4 is often expressed by Tregs, exhausted T cells, and tumor cells [25, 26]. PD-L1 is also evident on tumor cells, immune cells (such as MDSCs and TAMs), and vascular endothelial cells [27].

Strategies to mitigate ICs

To counter IC-instigated inhibitory signals in the TME, ICIs (most notably instituted by neutralizing monoclonal antibodies) can permit continued anti-tumor responses [28], despite IC prevalence and sustained antigen exposure [29]. PD-1 blockade diminishes effector T cells binding TME-retained PD-L1/PD-L2 to instead remain functionally active. Neutralizing CTLA-4 antibodies likewise may help instigate tumor rejection by facilitating T cell priming/activation in lymphoid tissues, namely, by preventing T cell exposure to inhibitory signals from APCs [30]. But, it has also been proposed that Tregs may bear the brunt of CTLA-4 blockade, effectively reversing immunosuppression within the TME. Tregs bearing anti-CTLA-4 antibodies appear to be selectively depleted in preclinical studies though an FcγRIV-dependent mechanism involving FcγR-expressing macrophages [31, 32]. The effects of directly blocking PD-L1 are generally expected to mimic that of neutralizing PD-1 on T cells, but system-specific deviations may arise based on the use of distinct ICI antibodies (and their isotypes) [33]. For example, PD-1 inhibition promoted tumor protection in an FcγR-independent manner in a murine colon cancer model while an anti-PD-L1 antibody in the same model augmented anti-tumor responses by eliminating PD-L1-expressing myeloid cells in the TME (via activating FcγRs) [34]. Additionally, cis interactions between PD-L1 and CD80 and/or prevention of PD-L1/PD-L2 engagement could be responsible for improved therapeutic outcomes with PD-1 blockade in certain scenarios as opposed to the strategy of only targeting PD-L1 (although head-to-head comparisons have not been formally studied clinically) [35, 36]. In vitro studies also suggest PD-L1:CD80 cis-heterodimerization prevents T cell-derived CTLA-4 interaction but permits CD28 engagement of CD80, implicating the need for anti-CTLA-4 neutralization if PD-L1 will also be targeted [37].

Clinical experience: CTLA-4 blockade

Despite potential modeling variations, a pivotal moment in the history of cancer immunotherapy came with the approval of the first ICI, ipilimumab (Yervoy®), for treatment of unresectable, late-stage melanoma in 2011. Ipilimumab is a humanized IgG1 antibody that was developed against CTLA-4. In the crucial phase 3 trial, ipilimumab significantly improved both 1-year (45.6% vs. 25.3%) and 2-year (23.5% vs. 13.7%) survival rates when compared to a gp100 vaccine control (outlined in Figure 2) [38]. Aside from extending overall survival (OS), post-trial observations revealed that some patients experienced prolonged durable responses, with 5-year survival rates ranging from 13–25% depending on the trial and ipilimumab-based protocol instituted [39]. Notably, select trial participants who experienced partial tumor regressions eventually transitioned to complete responses (CRs) months to years after receiving the therapy. In some cases, CR rates increased to 17% in patients receiving ipilimumab + IL-2 [39].

Figure 2. Patient overall survival from select clinical trials leading to FDA approved immune checkpoint inhibitor-based regimens.

(Panel A) Checkmate067 phase 3 trial (nivolumab + ipilimumab focus) – OS was a primary endpoint. (Panel B) IMvigor210 phase 2 trial (atezolizumab focus) – OS was a secondary endpoint (Panel C) Keynote-006 phase 3 trial (pembrolizumab focus) – OS was a primary endpoint. (Panel D) MDX010–20 phase 3 trial (ipilimumab focus) – OS was a primary endpoint. Abbreviation used: OS, overall survival.

Clinical experience: PD-1 blockade

While ipilimumab is currently the only approved antibody against CTLA-4, several ICIs are in clinical use for the PD-1/PD-L1 IC pathway. PD-1 blocking antibodies include nivolumab (Opdivo®), pembrolizumab (Keytruda®), and cemiplimab (Libtayo®) (summarized in Figure 2 and compiled in Table 1) [4]. Initial FDA approval of nivolumab came after it improved objective response rates (ORRs) over investigator’s choice of chemotherapy from 10.6% to 31.7% in 120 patients (previously progressing on ipilimumab and BRAF inhibition) enrolled in the Checkmate-037 trial [40]. Pembrolizumab, like nivolumab, is a humanized IgG4 antibody against PD-1 [41]. The open-label, phase 1b KEYNOTE-001 trial enrolled 655 patients with advanced melanoma (151 of which were treatment-naïve and 496 individuals having received prior treatments) and dosed pembrolizumab accordingly until the occurrence of disease progression, intolerable adverse events (AEs), or investigator determination of withdrawal [42]. After a median exposure of 5.6 months to the drug, a 5-year follow up revealed an estimated OS rate of 34% in all patients and 41% in treatment-naïve patients and a 5-year estimated progression free survival (PFS) rate of 21% and 29% in these patient cohorts, respectively [43, 44]. Later trials (KEYNOTE-002 and KEYNOTE-006) reported similar durable response rates ranging from 30–40% [45, 46]. Such notable successes helped lead nivolumab and pembrolizumab to approvals for numerous other solid and hematological malignancies [47, 48]. Lastly, cemiplimab was first approved for advanced cutaneous squamous cell carcinoma (CSCC) in 2018 after trial participants experienced an ORR of 47% against metastatic disease [49] and 44% against locally advanced CSCC [50]. The drug has since been approved for advanced basal cell carcinoma [51] and non-small cell lung cancer (NSCLC) [52] – both events occurring in 2021.

Table 1.

Clinical trials helping lead to FDA approved ICI-based regimens

Drug(s)	Target(s)	Trial (Clinicaltrials.gov registry)	Phase	Cancer indication(s)	Treatment group(s)	Major outcome(s)
Ipilimumab (Yervoy®)	CTLA-4	MDX010–20 (NCT00094653)	3	Unresectable, late-stage melanoma	ipilimumab + gp100 vs. ipilimumab vs. gp100	Median OS (months): 10 vs. 10.1 vs. 6.4
Nivolumab (Opdivo®)	PD-1	Checkmate-037 (NCT01721746)	3	Unresectable or metastatic melanoma with ICI relapse	nivolumab vs. investigator’s choice of chemotherapy	ORR: 31.7% vs. 10.6%; Median OS (months): 12.3 vs. 8
Pembrolizumab (Keytruda®)	PD-1	KEYNOTE-006 (NCT01866319)	3	Unresectable or metastatic ICI-naïve melanoma	pembrolizumab Q3W vs. pembrolizumab Q2W vs. ipilimumab Q3W	Median PFS (months): 4.1 vs. 5.5 vs. 2.8; OS (at 1 year): 68.4% vs. 74.1% vs. 58.2%
Cemiplimab (Libtayo®)	PD-1	R2810-ONC-1540 (NCT02760498)	2	Metastatic and locally advanced CSCC	cemiplimab	ORRs: 44–47%
Avelumab (Bavencio®)	PD-L1	JAVELIN Merkel 200 (NCT02155647)	2	Merkel cell carcinoma	avelumab	ORR: 31.8%
Durvalumab (Imfinzi®)	PD-L1	PACIFIC (NCT02125461)	3	Advanced and unresectable NSCLC	durvalumab vs. placebo	OS (at 2 years): 66.3% vs. 55.6%
Atezolizumab (Tecentriq®)	PD-L1	IMvigor210 (NCT02951767)	2	Locally advanced or metastatic UC that progressed on platinum-containing chemotherapy	atezolizumab in UC with 1% - <5% PD-L1 vs. atezolizumab in UC with ≥5% PD-L1	ORR (all subtypes): 19%; ORR (1% - <5% PD-L1): 22%; ORR (≥5% PD-L1): 27%
Nivolumab + Ipilimumab	CTLA-4 + PD-1	Checkmate067 (NCT01844505)	3	Previously untreated and advanced melanoma	nivolumab + ipilimumab vs. nivolumab vs. ipilimumab	OS (at 3 years): 58% vs. 52% vs. 34%

Abbreviations used: CSCC, cutaneous squamous cell carcinoma; ICI, immune checkpoint inhibitor; NSCLC, non-small cell lung cancer; ORR, objective response rate; OS, overall survival; PFS, progression free survival; Q2W, once every 2 weeks; Q3W, once every 3 weeks; UC, urothelial carcinoma

Clinical experience: PD-L1 blockade

PD-L1 inhibitors available for clinical use include avelumab (Bavencio®), durvalumab (Imfinzi®), and atezolizumab (Tecentriq®) (see Figure 2 and Table 1) [4]. Avelumab was approved in 2017 based on the JAVELIN Merkel 200 trial for treatment of metastatic chemotherapy-resistant Merkel cell carcinoma (MCC) after trial patients experienced an ORR of 31.8% [53]. Notably, avelumab is the first FDA approved treatment for metastatic MCC. Durvalumab was initially granted accelerated approval from the FDA in 2017 for patients with PD-L1+ inoperable or metastatic urothelial bladder cancer (mUC) that had progressed following treatment with a platinum-based course of therapy [54] although this was later withdrawn in 2020 following a phase 3 trial that did not meet its primary endpoints [55]. A separate phase 3 study (PACIFIC trial), however, in patients with surgically unresectable stage III NSCLC demonstrated significant improvements in patients treated with durvalumab as compared to placebo in PFS (17.2 months vs. 5.6 months), OS at 2 years (66.3% vs. 55.6%), and time to distant metastases or death (28.3 months vs. 16.2 months) prompting FDA approval in 2018 [56]. Atezolizumab received FDA approval in 2016 for the treatment of locally advanced or mUC in patients who progressed on platinum-containing chemotherapy based on the results of the IMvigor210 trial [57]. The following year, atezolizumab achieved approval for treatment of mUC in patients not eligible for chemotherapy regardless of PD-L1 tumor expression; however, in 2018, the FDA clarified that atezolizumab should only be used in patients [i] with tumor expression of PD-L1 ≥ 5% or [ii] who are not eligible for a platinum-containing chemotherapy regimen regardless of PD-L1 expression [4]. In addition to other indications (such as NSCLC, small cell lung cancer, hepatocellular carcinoma [HCC], and melanoma), atezolizumab received accelerated approval in 2019 for the treatment of metastatic triple-negative breast cancer (TNBC) harboring PD-L1 expression based on results of the phase 3 Impassion 130 study [58]. Briefly, patient cohorts receiving atezolizumab plus nab-paclitaxel demonstrated significantly increased PFS of 7.2 months compared to placebo plus nab-paclitaxel treatment of 5.5-months. Median OS also increased from 17.6 to 21.3 months in atezolizumab-treated patients overall and from 15.5 to 25.0 months in individuals with PD-L1 positive tumors.

Targetable ICs on the horizon

Although anti-CTLA-4 and anti-PD-1/PD-L1 therapies have conveyed impressive clinical results in subsets of patients, research is ongoing to identify other targetable ICs that are capable of sustaining immune-directed attacks against cancer lesions. Recent promising preclinical and early phase clinical results have led to intense interest in assessing other distinct ICs including LAG-3, TIM-3, TIGIT, and VISTA in phase 2 and 3 trials (overviewed in Table 2) [59, 60]. LAG-3 is absent on naïve CD4+ and CD8+ T cells, with expression being induced following antigen stimulation. Interestingly, LAG-3 presence can also fluctuate on other immune cells such as Tregs, NK cells, B cells, and plasmacytoid DCs [59]. The downstream effects of LAG-3 blockade in various studies have been modest [61], but co-expression of PD-1 and LAG-3 on exhausted CD8+T cells has kindled interest in combining PD-1/LAG-3 neutralizing antibodies as a potential means to overcome PD-1 blockade resistance [62–64]. Accordingly, simultaneous disruption of PD-1 and LAG-3 has demonstrated improved effects in reversing T cell exhaustion over either ICI alone in preclinical TNBC, melanoma, and colorectal cancer (CRC) models [65]. TIM-3 expression was initially recognized on CD4+ Th1 and CD8+ T cells, and is classified as an IC based on an ability to mediate T cell exhaustion through binding either Gal-9, CEACAM-1, HMGB1, or phosphatidylserine [66]. In the context of cancer, TIM-3 has been utilized as a marker for dysfunctional subsets of tumor-infiltrating CD8+ T cells that also express PD-1, with dysfunction being further identified by poor effector cytokine secretion [67, 68]. Like LAG-3, TIM-3 is frequently co-expressed with PD-1 on T cells and is observed to be upregulated following blockade of PD-1. TIM-3 is also frequently expressed by Tregs [69]. In mouse models of melanoma and CRC, TIM-3 and PD-1 neutralization resulted in an improved re-establishment of T cell functionality [68, 70–72]. TIGIT is expressed by activated CD8+ and CD4+ T cells, NK cells, Tregs, and follicular helper T cells [59] and interacts with CD155 (PVR) or CD112 (PVRL2), which are commonly present on cancer cells and APCs in the TME [66]. In mouse tumor models, TIGIT impairs T cell proliferation and function by binding CD155 on DCs, triggering an internal signaling cascade that results in reduced production of IL-12 and increased secretion of IL-10 alongside an overwhelmingly tolerogenic phenotype [73, 74]. TIGIT neutralization has demonstrated therapeutic benefits in combination with PD-1/PD-L1 blockade [75]. Additionally, preclinical studies have noted enhanced anti-tumor immune responses when TIGIT inhibition is combined with TIM-3 blockade [76]. VISTA (also known as PD-1H or B7-H5) is a recently identified coinhibitory protein of the CD28/B7 family [77]. With significant homology to both PD-1 and PD-L1, VISTA is commonly expressed on myeloid cells, naïve CD4+/CD8+ T cells, and Foxp3+ CD4+ Tregs [78]. VISTA has been shown to suppress CD4+ and CD8+ T cell proliferation, effector cytokine production, and cytotoxicity, with preclinical studies demonstrating increased anti-tumor benefits following VISTA blockade [79–82]. Yet, it currently remains mechanistically unclear how VISTA’s functions as a ligand and receptor contribute to T cell inhibition [83]. Additionally, VISTA can promote proliferation of tumor-specific CD4+ Foxp3+ Tregs, further contributing to VISTA’s T cell inhibitory functions [84]. In a study of 16 patients with metastatic melanoma initially responding to (but later relapsing from) anti-PD1 or anti-PD1/CTLA-4 therapies, VISTA+ lymphocytes significantly increased in 67% post-disease progression biopsies, potentially implicating the protein as a marker for acquired resistance to common ICIs [85]. However, in other cancer types, VISTA expression has been found to be positively correlated to survival [86, 87]. No doubt, while VISTA may be a promising immunotherapeutic target, further research is required to elucidate its complex roles across various malignancies.

Table 2.

Notable ICI antibody combinations under clinical trial evaluation

Targets	Drugs	Phase	Cancer indication(s)	Clinicaltrials.gov registry
LAG-3 PD-1	Favezelimab Pembrolizumab	3	CRC	NCT05064059
LAG-3 PD-1	Relatlimab Nivolumab	2/3	Melanoma	NCT03470922
LAG-3 PD-1	LAG525 Spartalizumab	2	Advanced solid tumors	NCT03365791
TIM-3 PD-1	TSR-022 Dostarlimab	2	Melanoma	NCT04139902
TIM-3 PD-1	MBG453 PDR001	1/2	Advanced solid tumors	NCT02608268
TIM-3 PD-1	BMS-986258 Nivolumab	1/2	Advanced solid tumors	NCT03446040
TIGIT PD-1	Vibostolimab Pembrolizumab	3	NSCLC	NCT04738487
TIGIT PD-1	Domvanalimab Zimberelimab	2	NSCLC	NCT04262856
VISTA PD-1	W0180 Pembrolizumab	1	Advanced solid tumors	NCT04564417
TIM-3 PD-L1	LY3321367 LY3300054	1	Advanced solid tumors	NCT03099109
TIGIT PD-L1	Tiragolumab Atezolizumab	3	NSCLC	NCT04294810
TIGIT PD-1 CTLA-4	BMS-986207 Nivolumab Ipilimumab	1/2	Advanced solid tumors	NCT02913313
LAG-3 TIM-3 PD-1	INCAGN02385 INCAGN02390 INCMGA00012	1/2	Melanoma	NCT04370704

Abbreviations used: ICI, immune checkpoint inhibitor; CRC, colorectal cancer; NSCLC, non-small cell lung cancer

Clinical obstacles to ICIs

Despite clinical benefits in subsets of patients, resistance to ICIs is a prevailing issue [88]. Some individuals never demonstrate an observable clinical response to ICI therapy (i.e., inherent insensitivity), while other patients exhibit acquired resistance following a period of responsiveness. Broadly, successful ICI interventions are generally proportional to tumor mutational burden (TMB) and, as a by-product, neoantigen processing/presentation that can elicit cytotoxic T lymphocyte activity [89]. For example, treatment-naive patients with high microsatellite instability (MSI-H) or mismatch repair deficiency (dMMR) CRC (i.e., leading to increased TMBs) provided pembrolizumab experienced a more than doubled PFS (16.5 months) over individuals receiving a standard-of-care chemotherapy-containing regimen (8.2 months) in the phase 3 KEYNOTE-177 trial [90]. Yet, in microsatellite stable (MSS) or mismatch repair proficient CRCs (i.e., displaying low TMB), ICIs do not typically impart clinically beneficial endpoints [91]. Such observations also apply to high TMB malignancies such as melanoma and NSCLC, where carcinogen-driven events endow immunogenicity through expression of neoantigens [92, 93]. Tumor-extrinsic mechanisms may also contribute an important role toward ICI resistance. Immunosuppressive cells populated in the TME (such as Tregs, MDSCs, and TAMs) can supply an ever-present cocktail of inhibitory cytokines and receptors that downmodulate the function of CD8+ T cells and APCs, which, in turn, discourages cytotoxic-driven immune responses against cancer cells [94–96].

In cases where patients display an initial promising ICI response following treatment, a substantial segment of individuals (nearly 80%) will not attain durable clinical responses in many cancer types such as those arising in the lung, breast, colon, and prostate [97–99]. The explicit details surrounding acquired resistance are not currently fully understood, but several mechanisms have been proposed. One major culprit involves the upregulation of other ICs to compensate for the loss of CTLA-4 or PD-1/PD-L1 following blockade [100]. An additional evasive avenue for the TME includes the expansion of cells expressing gene variants that institute loss-of-function mutations in JAK family proteins [101] and beta-2-microglobulin (B2M) [102], that ultimately confer cancer cell survival. For example, a small study of whole-exome sequencing in four patients with metastatic melanoma who, after initial tumor regression on pembrolizumab, experienced delayed relapses revealed resistance-associated mutations. One patient’s post-progression tumor biopsy included a frameshift deletion in B2M, resulting in an absence of cell surface-retained MHC class I and inability of CD8+ T cells to recognize the material [103]. Two patients’ biopsies (from the same study) post-ICI progression exhibited inactivation of JAK1 or JAK2, which detrimentally impacts IFN-γ responses that play important roles in nurturing APC function and inhibiting the progression of TME components such as tumor cells and the supportive vasculature [103]. Interestingly, other studies have found epigenetic mechanisms that may similarly institute acquired resistance to ICI treatment [101, 104]. Lastly, cell-mediated resistance mechanisms have also been implicated in resuscitating tumorigenesis following ICI treatment. In a preclinical CRC study employing in vivo imaging to determine the fate of an anti-PD1 antibody after administration, the antibody successfully bound PD1+ tumor-infiltrating CD8+ T cells but was quickly removed from the T cell surface by TAMs. This effect appeared to be driven by Fc-FcyR binding interactions and was recapitulated in vitro using nivolumab and human primary CD8+ T cells and macrophages [105].

Since CTLA-4 and PD-1/PD-L1 neutralizing antibodies impart distinct immunologic effects, ICI cocktails have gained interest as a potential route to discourage resistance and increase/extend clinical responses. Currently, ipilimumab + nivolumab is the only FDA-approved combination for cancer indications that include melanoma, renal cell carcinoma, MSI-H/dMMR metastatic CRC (mCRC), HCC, NSCLC, and malignant pleural mesothelioma [106–109]. Dual checkpoint blockade tends to improve OS rates (58% in the ipilimumab/nivolumab group, 52% in the nivolumab cohort, and 34% in the ipilimumab group in untreated melanoma patients [CheckMate067 phase 3 trial] [110]) and PFS (11.5 months for combination treatment vs. 6.9 months or 2.9 months for nivolumab and ipilimumab monotherapies, respectively, in CheckMate067) where data are available to directly compare single and combined ICI agents. Yet, additive ICI regimens may drastically increase immune-related AEs [111]. In CheckMate067, 59% of ipilimumab/nivolumab patients reported treatment-related grade 3 or 4 AEs compared to individuals treated with nivolumab (21%) or ipilimumab (28%) [110]. To counteract such issues, recent experience supports smaller doses of ipilimumab (alongside other ICIs) that may help abrogate increased toxicities while still conferring enhanced clinical benefits [112, 113]. Newer ICI combinations (as also mentioned above) are also currently being evaluated in clinical trials and may provide additional benefits to patients (see Table 2 for further details).

Intriguingly, stromal cells in the TME including endothelial cells and fibroblasts can drastically influence ICI responsiveness indirectly through various means. As an example, co-administration of tumor-associated fibroblasts expressing FAP in a murine model of MSS CRC promoted immunosuppression and resistance to anti-PD1 treatment by recruiting MDSCs via CCL2 chemotaxis [114]. Tumor stroma exposed to TGFβ (expressed by cancer cells or fibroblasts) could also be implicated in resistance to ICI blockade by antagonizing T cell infiltration and cytotoxic differentiation, which appears relevant to ICI responding/nonresponding mUC patients from the IMvigor210 trial [115, 116]. Notably, in various murine models like CRC and TNBC, TGFβ inhibition significantly increased sensitivity to IC blockade to better promote anti-tumor immunity [116]. Endothelial cells (immersed in a rich milieu of pro-angiogenic factors) also downregulate important adhesion molecules like ICAM1, VCAM1, and E-selectin to prevent circulating leukocytes extravasating into the underlying tumor bed [117]. More directly, endothelial cells can express ICs (e.g. PD-L1/L2) or soluble mediators like VEGF that upregulate ICs on infiltrating T cells [118]. So, overall, the physical and therapeutic constraints imposed by the TME must also be considered within the confines of ICI-based regimens.

Tumor angiogenesis

The TME consists of accessory cells such as immune cells, fibroblasts, and blood and lymphatic vessels that promote cancer growth and progression. Angiogenesis tends to be a crucial TME factor that is paramount to tumor survival and expansion, and, in the absence of blood vessel development, primary (avascular) tumors will usually only grow to a limited size with poor metastatic potential [119]. On the other hand, prolific tumorigenesis requires a sufficient blood vessel network that enables transport of oxygen, nutrients, and waste to and from cancer cells. However, the disordered and rapid growth of cancer cell proliferation ultimately strains vascular function; in turn, leading to intratumoral hypoxia, acidosis, and interstitial pressure, which further facilitates the release of signals that promote angiogenesis [120]. This phenomenon for vessel growth is initially managed by a balance of pro- and anti-angiogenic factors, but upon an “angiogenic switch” being tripped, such equilibrium is disrupted to favor unfettered vascular growth [121]. Ultimately, the TME’s tortuous blood vessel architecture consists of “immature” blood vessels (i.e., endothelial cells with low pericyte coverage and deficient basement membranes) that are leaky and permit the extravasation of tumor cells at distant sites [122, 123] that may eventually propagate macroscopic disease [124, 125].

Given the crucial need for angiogenesis in tumor development (and the strict control of blood vessel genesis/maintenance in healthy adult tissues), irregular characteristics of the tumor vasculature have enticed development of specific therapeutic interventions [126, 127]. Over the past decade alone, the FDA has approved a variety of anti-angiogenic cancer drugs that largely fall into three classes: monoclonal antibodies, small molecule drugs (namely, tyrosine kinase inhibitors [TKIs]), and fusion proteins (summarized in Figure 3 and detailed in Table 3). These vessel-targeting therapies work through various discrete molecular mechanisms, but the end result is usually endothelial cell destruction or normalization (at least in the short-term). Tumor-derived vessel destruction culminates in the abrogation of a necessary infrastructure for tumor tissue survival, depriving cells of needed oxygen/nutrients and expunging metabolic waste. Yet, despite these potential benefits, vessel demolition may elicit evasive immune cell maneuvering (to rearm tumor vascularity, as an example) or potentiate the emergence and dissemination of aggressive hypoxic-derived tumor cell variants [128]. Tissue normalization, on the other hand, results in a culling of “immature” endothelial branches and stabilization of “mature” vessels (i.e., those displaying increased pericyte coverage) that serves to reinstitute vessel integrity and normal aspects of blood flow and oxygenation. Vessel normalization also tends to aid the improved TME perfusion and performance of small molecule drugs/chemotherapeutics or immune interventions, underscoring the enhanced protective index of appropriately designed combinatorial strategies for patients [126, 129, 130].

Figure 3. Vascular-targeting strategies mediate tumor protection by abrogating important tumor angiogenesis pathways.

Blood vessel development/maintenance within the tumor microenvironment helps promote tumor cell survival and invasion. Yet, FDA approved monoclonal antibodies (like bevacizumab or ramucirumab), tyrosine kinase inhibitors, and fusion protein (e.g., aflibercept) therapeutics work to specifically interact with vascular components to disrupt crucial cancer angiogenesis pathways. The resulting destruction of tumor-derived vessels deprives cancer cells of a needed support system to grow and progress. Abbreviations used: FGFR, fibroblast growth factor receptors; PDGFR, platelet-derived growth factor receptors; PGF, placental growth factor; VEGF, vascular endothelial growth factor; VEGFR, VEGF receptor; Images created with Biorender.com.

Table 3.

Pivotal clinical trials leading to FDA approval of select anti-angiogenic therapies for cancer

Drug	Class (Target[s])	Phase	Cancers indication(s)	Treatment groups	Major outcome(s)
Bevacizumab (Avastin®)	mAb (VEGF)	3	mCRC	bevacizumab + chemotherapy vs. placebo + chemotherapy	Median OS (months): 20.3 vs. 15.6; Median PFS (months): 10.6 vs. 6.2
Ramucirumab (Cyramza®)	mAb (VEGFR2)	3	Gastric cancer	ramucirumab vs. placebo	Median OS (months): 5.2 vs. 3.8; Median PFS (months): 2.1 vs. 1.3
Sorafenib (Nexavar®)	TKI (Multiple)	3	RCC	sorafenib vs. placebo	Median PFS (months): 5.5 vs. 2.8
Sunitinib (Sutent®)	TKI (Multiple)	3	RCC	sunitinib vs. interferon alfa	Median PFS (months): 11 vs. 5; ORR: 31% vs. 6%
Pazopanib (Votrient®)	TKI (Multiple)	3	RCC	pazopanib vs. placebo	Median PFS (months): 9.2 vs. 4.2; ORR: 30% vs. 3%
Axitinib (Inlyta®)	TKI (Multiple)	3	RCC	axitinib vs. sorafenib	Median PFS (months): 6.7 vs. 4.7
Regorafenib (Stivarga®)	TKI (Multiple)	3	mCRC	regorafenib vs. placebo	Median OS (months): 6.4 vs. 5.0; Median PFS (months): 2.0 vs. 1.7
Cabozantinib (Cabometyx®)	TKI (Multiple)	3	RCC	cabozantinib vs. everolimus	Median PFS (months): 7.4 vs. 3.8; ORR: 21% vs. 5%
Lenvatinib (Lenvima®)	TKI (Multiple)	3	Differentiated thyroid cancer	lenvatinib vs. placebo	Median PFS (months): 18.3 vs. 3.6; ORR: 64.8% vs. 1.5%
Aflibercept (Zaltrap)	Fusion protein (VEGFs, PGF)	3	mCRC	aflibercept + chemotherapy vs. placebo + chemotherapy	Median OS (months): 13.5 vs. 12; Median PFS (months): 6.9 vs. 4.7

Abbreviations used: mAb, monoclonal antibody; mCRC, metastatic colorectal cancer; ORR, objective response rate; OS, overall survival; PFS, progression free survival; PGF, placental growth factor; RCC, renal cell carcinoma; TKI, tyrosine kinase inhibitor; VEGFs, vascular endothelial growth factors

Anti-angiogenic monoclonal antibodies are designed to bind vascular growth factors or cell-surface receptors that are major drivers of the pro-angiogenic cascade. Of the two antibodies approved, bevacizumab (Avastin®) was first in 2004 to demonstrate efficacy in patients with mCRC [131]. Bevacizumab acts by binding VEGF-A (an extremely potent angiogenesis inducer), which prevents the molecule from interacting with VEGFR2 on vascular endothelial cells, endothelial progenitor cells, and megakaryocytes [126]. Importantly, VEGF-A that is typically found in excess in the TME is neutralized and tumor vascularity ablated (at least in the interim) [132, 133]. Combining this effect with additional targeting strategies serves to increase tumor exposure to, for example, cytotoxic chemotherapies, presumably through a vessel normalization phenomenon [134, 135], that has demonstrated significant increases to both survival and response rates for indications like ovarian, NSCLC, renal cell carcinoma (RCC), and CRCs [136–144]. Ramucirumab (Cyramza®) is a distinct monoclonal anti-VEGFR2 antibody that directly interferes with VEGF binding, creating a similarly normalized vascular environment, wherein, traditional chemotherapies can better elicit anti-tumor effects. The FDA has approved ramucirumab for the treatment of NSCLC, mCRC, HCC, and gastric cancer [145].

The TKI class instigates tumor regressions by blocking important angiogenic-driven activation signals. Those tyrosine kinases implicated with the tumor angiogenic pathway often involve VEGF receptors (VEGFRs), platelet-derived growth factor receptors (PDGFRs), and fibroblast growth factor receptors (FGFRs). When targeted, TKIs interfere with kinase activity and downstream signaling events important for vascular proliferation, migration, and survival [146]. The TKI sorafenib (abrogating TKs such as VEGFRs, PDGFR-β, and FLT-3) was first FDA approved against advanced RCC [147]. Sorafenib’s activity against HCC also yielded an increased OS of 2.8 months (versus the placebo group) and its adoption as a first-line treatment in advanced cases [146, 148]. Sunitinib is another established TKI that is FDA approved for use against gastrointestinal stromal tumors (GIST), pancreatic neuroendocrine tumors, and metastatic RCC (mRCC) [149–151]. Sunitinib acts by interfering with TKs like VEGFRs, PDGFRs, and FLT-3 [152]. Further iteration of this class has been realized in second generation TKIs that demonstrate a more pharmacologically desirable effect (i.e., with reduced toxicities) [134]. Drugs such as regorafenib demonstrate potent anti-tumoral activity in their capacity to target kinases involved in the regulation of tumor angiogenesis (e.g., VEGFRs, FGFRs, PDGFRs, TIE2) and oncogenesis [153, 154]. Such drug development advances have yielded potent clinical activity, culminating in FDA approval for the treatment of mCRC, GIST, and HCC [155]. Uniquely, unlike the antibody class, TKI combination therapies incorporating chemotherapy typically improve PFS but at the expense of exacerbating severe AEs in patients that can lead to treatment discontinuation or death [156, 157].

The fusion protein class is currently represented by one compound – ziv-aflibercept (Zaltrap®). Ziv-aflibercept is composed of the extracellular domains of VEGFR-1 and VEGFR-2 that are fused to the Fc region of IgG1. The drug works as a decoy to immobilize/neutralize important angiogenic cues such as VEGF-A, VEGF-B, and placental growth factor, thereby, halting the pro-angiogenic effects of the VEGFR signaling pathway [158, 159]. Ziv-aflibercept was approved by the FDA in 2012 as a combination therapy with chemotherapy for the treatment of mCRC, after demonstrating a significantly improved OS and PFS in study participants [160, 161].

Co-applying anti-angiogenic and ICI strategies

In general, clinical experience tends to diverge from preclinical observations that anti-angiogenic antibodies induce tumor regressions as monotherapies. Vascular-disruptive small molecule drugs also usually only provide short-term patient benefits [162, 163]. Therefore, the potential for this class of therapeutics (at-large) is likely best realized in rationally designed combination therapies to make use of their unique effects on the TME. Since many pro-angiogenic molecules are associated with immunosuppressive effects, the administration of anti-angiogenic drugs should help sustain anti-tumor immune responses. Indeed, numerous preclinical models have demonstrated that combined anti-angiogenic and immunotherapeutic regimens resulted in normalized tumor vessels that promoted the enhanced infiltration and function of activated immune effector cells [135, 164, 165]. For example, VEGF neutralization by antibodies not only reduces tumor vascularity but also increases antigen-specific T cell infiltration and activity [166–169]. Small molecule drugs like axitinib and sunitinib also restore aberrant characteristics of the TME (e.g., diminishing hypoxia and Treg recruitment) to ultimately improve vaccine-driven T cell cytotoxicities [170, 171]. Interestingly, though anti-angiogenic interventions tend to upregulate ICs on tumor cells such as PD-L1 - presumably due to improved immune effector function within the TME (such as through IFN-γ signaling), their use in tandem with ICIs significantly reduced tumor growth/metastases and prolonged survival in cancer models related to the pancreas, breast, liver, and brain [172–177]. Consequently, given such corroborated evidence from preclinical systems, co-applied ICIs and anti-angiogenic therapies have made considerable progress in clinical trials in recent years, particularly for RCC, HCC and NSCLC (summarized in Table 4).

Table 4.

Examples of combined anti-angiogenic & ICI therapies

Drugs (ICI/anti-angio)	Trial (Clinicaltrials.gov registry)	Phase	Cancers indication	Treatment groups	Major outcome(s)
Atezolizumab + bevacizumab	IMbrave150 (NCT03434379)	3	HCC	ICI/anti-angio vs. sorafenib	Median PFS (months): 6.8 vs. 4.3; Median OS (at 1-year): 67.2% vs. 54.6%
	IMpower150 (NCT02366143)	3	NSCLC	ICI/anti-angio + carboplatin + paclitaxel vs. bevacizumab + carboplatin + paclitaxel	Median PFS (months): 8.3 vs. 6.8; Median OS (months): 19.2 vs. 14.7
	IMmotion151 (NCT02420821)	3	RCC	ICI/anti-angio vs. sunitinib	Median PFS (months): 11.2 vs. 7.7
Avelumab + axitinib	JAVELIN Renal 101 (NCT02684006)	3	RCC	ICI/anti-angio vs. sunitinib	Median PFS (months): 13.8 vs. 8.4
Pembrolizumab + axitinib	KEYNOTE-426 (NCT02853331)	3	mRCC	ICI/anti-angio vs. sunitinib	Median PFS (months): 15.1 vs. 11.1; ORR: 59.3% v. 35.7%

Abbreviations used: HCC, hepatocellular carcinoma; ICI, immune checkpoint inhibitor; mRCC, metastatic renal cell carcinoma; NSCLC, non-small cell lung cancer; ORR, objective response rate; OS, overall survival; PFS, progression free survival; RCC, renal cell carcinoma

In the phase 3 KEYNOTE-426 trial, mRCC patients treated with combined axitinib and pembrolizumab exhibited significantly increased PFS, 47% reduced risk of death, and an ORR of 59.3% compared with 35.7% in the sunitinib monotherapy group [178, 179]. The overall efficacy and toxicity profiles prompted FDA approval of axitinib/pembrolizumab as a first-line strategy for treatment-naïve mRCC patients in 2019 [180]. The phase 3 JAVELIN Renal 101 trial also reported significantly improved median PFS (13.8 months vs. 7 months) in advanced RCC patients taking avelumab plus axitinib compared to sunitinib alone [181, 182], which led the FDA to approve the regimen as a first-line treatment in 2019 [183]. Additionally, a separate phase 3 trial (IMmotion151) incorporating atezolizumab and bevacizumab provided significant increases to PFS (11.2 months vs. 7.7 months) for mRCC patients relative to sunitinib treatment [184, 185].

A variety of combination therapies have also been designed to make use of the observed preclinical relationship between VEGFR2 disruption and PD-L1 upregulation in HCC as a result of TME-produced IFN-γ [176, 186]. One such combination therapy involving lenvatinib (a multi-kinase angiogenesis inhibitor [e.g., targeting VEGFRs, FGFRs, and PDGFRα]) and pembrolizumab in a phase 1b evaluation demonstrated significantly improved ORRs while maintaining an acceptable safety profile in individuals with unresectable HCC [187, 188]. These results prompted the LEAP-002 phase 3 trial (NCT03713593) to evaluate lenvatinib/pembrolizumab as a viable first-line option for unresectable HCC [189] although the regimen achieved FDA approval in 2021 for mRCC [190]. Cabozantinib (e.g., blocking VEGFRs, FLT-3, and TIE2) represents another distinct kinase inhibitor with immunomodulatory potential and has been previously approved as a monotherapy against sorafenib-refractory HCC [191]. Accordingly, cabozantinib is currently being evaluated in combination with ICIs like atezolizumab in the phase 3 COSMIC-312 trial (NCT03755791) following early trial successes [192]. Lastly, a major advance for unresectable HCC came with the recent phase 3 IMbrave150 trial that demonstrated improved median PFS (6.8 months vs. 4.3 months) and OS at 1-year (67.2% vs. 54.6%) with acceptable safety profiles compared to the standard sorafenib treatment [193]. Atezolizumab/bevacizumab was approved by the FDA in 2020 as a first-line therapy for advanced HCC [194].

The treatment of NSCLC has similarly benefited from the application of combined anti-angiogenic/ICI therapeutic strategies. The Impower150 phase 3 trial comparing combined therapeutic strategies. The Impower150 phase 3 trial comparing combined atezolizumab/bevacizumab/carboplatin/paclitaxel (ABCP) to bevacizumab/carboplatin/paclitaxel demonstrated significantly increased median PFS (8.3 months vs. 6.8 months) and OS (19.2 months vs. 14.7 months) with comparable AE rates [195]. These clinical endpoints were found to be significant regardless of PD-L1 expression and EGFR or ALK genetic mutation status [196]. As a result, the ABCP combination was approved by the FDA in 2018 as first-line treatment for patients with metastatic non-squamous NSCLC.

Yet, combination therapies are not without their shortcomings nor created equal. The endogenous/acquired resistance mechanisms in the TME (as explained earlier) can predispose individuals to inefficacious therapies while other patients may suffer from significant AEs as a result of drug combinations. Although such problems could be lessened/mitigated by the use of predictive circulating biomarkers and/or TME-based pharmacogenomics, they cannot entirely safeguard the patient experience [197, 198]. This has proven beneficial in the administration of atezolizumab and bevacizumab in comparison to sunitinib monotherapy (i.e., IMmotion151 trial), wherein, a significant increase in PFS was observed in the sarcomatoid variant of mRCC and PD-L1 expressing tumors [168, 184]. On the contrary, treatment with the TKI pazopanib (that is approved as a single-agent for RCC) alongside nivolumab or pembrolizumab was found to be associated with significant hepatoxicity in mRCC patients and not deemed safe for further study [199]. The addition of axitinib/pembrolizumab has also fostered high grade 3 and 4 elevations of hepatic enzymes without any attributable mechanistic causes (KEYNOTE-426 trial) [184, 199]. Additionally, combinations with ICI antibodies can even present problems such as in the case of dual PD-1/CTLA-4 blockade for melanoma patients where over 50% of individuals experienced a significant immune-related AE [200]. Certainly the bulk of such toxicities may be manageable in the short-term (through, for example, clinically inducing immune suppression), but longer lasting immune-mediated issues could arise in the end [201].

Sustained treatment efficacy of any multipronged regimen represents another towering challenge to the patient. Although this problem is inherent to oncology in general, the improvements experienced with many of the “successful” combination treatments described above (evidenced by enhanced PFS and OS metrics) does not typically translate into durable patient responses for a majority of individuals. Optimistically, many of the common challenges that face immunotherapeutic/anti-angiogenic therapies can be eventually answered by, for example, a deeper mechanistic understanding of the interplay between components of the TME alongside therapeutic interventions [164, 202]. Yet, a shorter path to action also involves identifying new and novel therapeutic agents that can effectively incite tumor regressions while theoretically circumventing known impediments to lasting protection. The appropriation of bispecific antibodies (bsAbs) may represent one such approach to safely optimizing complex anti-angiogenic and ICI combinations in patients. BsAbs retain distinct Fab binding sites directed at different epitopes and are generally classified to be either IgG-or non-IgG-like molecules [203]. Their modular design permits an array of immune effector mechanisms due to the potential interplay between targeted sites. In 2014, the FDA approved blinatumomab (Blincyto®) as the first bispecific T-cell engager, a non-IgG like bsAb for the treatment of Philadelphia chromosome-negative relapsed or refractory B-cell precursor acute lymphoblastic leukemia. Blinatumomab is composed of linked scFv domains that work to incite CD3+ T cells killing against tethered CD19+ tumor cells [204]. However, the common bridged engagement phenomenon of bsAbs is not limited to T cell interactions. An additional FDA-approved bsAb emicizumab (Hemlibra®) treats factor VIII deficiency in hemophilia A by connecting factors IXa and X [205, 206]. BsAb-based therapeutics may also mediate agonist (i.e., mimicking the activity of natural ligands and cofactors) and/or antagonist (i.e., inactivating or blocking receptors/ligands to blunt downstream cellular events) effects [207]. In May 2021, the FDA approved the new bsAb amivantamab (Rybrevant^™) for treatment of NSCLC. Amivantamab incites tumor destruction by binding and blocking the function of surface-retained MET and mutant EGFR [208, 209]. These aforementioned approved molecules ultimately help rationalize a burgeoning therapeutic pipeline of over 100 bsAbs in clinical development for cancer [203, 210, 211]. BsAbs may also be uniquely suited to neutralize multiple candidate ICI and/or angiogenic targets more effectively within the TME (and with drastically reduced toxicities) than previous clinical attempts using additive monotherapies (as hypothesized in Figure 4) – although this area is still in its infancy and requires clinical validation.

Figure 4. Potential for bispecific antibodies enhancing anti-angiogenic and ICI strategies for cancer.

(Panel A) Generation of a standard bsAb format that specifically targets tumor cells and subsequently engages/activates locoregional CD3+ T cells. The provided example is the emerging bsAb mosunetuzumab for lymphoma, where variable domains from disparate antibodies have been attached to an Fc silenced heavy chain. (Panel B) Hypothetical utility of bsAbs to perform complex neutralization schemes within the TME that drives potent immune-driven protection and circumvents cancer evasion strategies. Ultimately, bsAbs could be instituted (along with other targeting agents) to synergistically block ICs and/or vascular targets to better achieve tumor localization that minimizes toxicities in patients. Abbreviations used: BsAb, bispecific antibody; ICs, immune checkpoints; ICI, immune checkpoint inhibitor; TME, tumor microenvironment. Images created with Biorender.com.

Regarding tumor-derived angiogenesis, Boehringer Ingelheim is evaluating a humanized VEGF x Ang-2 bsAb (designated BI 836880) in patients with solid tumors. Previously, a separate murinized VEGF x Ang-2 bsAb demonstrated improved therapeutic efficacy versus either anti-angiogenic antibody alone in transplant tumor models, and combined PD-1 neutralization provided enhanced bsAb activity (since the bsAb upregulated PD-L1 on tumor vessels) [212]. Similarly, in a preclinical brain metastasis model, BI 836880 prolonged animal survival and reduced brain metastasis formation that was linked to vascular disruption [213]. The molecule has also demonstrated manageable safety profiles in early phase clinical trials when used alone (NCT02674152, NCT026895053) and anti-tumor activity in combination with PD-1 blockade in NSCLC patients, with studies ongoing (NCT0346842) [214]. A DLL4 x VEGF bsAb (dilpacimab) is also being developed by AbbVie. In preclinical testing, dilpacimab achieved superior in vivo tumor protection over individual monotherapies through disruption of tumor angiogenesis, and the inclusion of chemotherapy alongside dilpacimab provided enhanced therapeutic benefits [215]. Preliminary phase 1 testing (NCT01946074) supported a good safety profile, with clinical activity occurring in a subset of advanced solid tumor patients [216]. Dilpacimab is currently under phase 2 evaluation in individuals with mCRC (NCT03368859). There is also intense commercial interest in designing bsAbs that engage multiple ICs. Some prominent examples include co-targeting PD-1 x CTLA-4 (Macrogenics, AstraZeneca, Xencor), PD-1 x LAG-3 (Macrogenics, Roche), and PD-1 x PD-L1 (Eli Lilly). As tumor infiltrating lymphocytes tend to express both PD-1 and CTLA-4, combined IC abrogation from a bsAb within the TME is hypothesized to better sustain T cell effector functions, which has been confirmed preclinically [217, 218] and led to several ongoing clinical trials (NCT03761017, NCT04522323, NCT05005728). Similarly, a PD-1 x LAG-3 bsAb would aim to resuscitate exhausted T cells in cancer lesions of patients as currently being clinically investigated (e.g., NCT04082364, NCT04785820). Eli Lilly’s PD-1 x PD-L1 bsAb (designated LY3434172) works to bridge PD-1+ T cells and PD-L1+ tumor cells, thereby, preferentially saturating PD-1 on effector T cells in the TME (opposed to peripheral lymphocytes). Although LY3434172 is under phase 1 evaluation (NCT03936959), in preclinical studies, the bsAb promoted superior in vitro T cell functionality and in vivo anti-tumor responses that were not possible with individual or combined parent antibody treatments [219]. Lastly, a noteworthy bsAb (from the company Akesobio) under clinical trial evaluation outside the United States actually combines angiogenic and IC targeting. This unique tetrameric bsAb engages both VEGF and PD-1 and presupposes that since these immunosuppressive targets are upregulated in the TME, bsAb-directed blockade will enhance anti-tumor activities. Initial phase 1 testing has determined a safe maximum tolerated dose with encouraging ORRs in patients with solid cancers (NCT04047290) [220]. The company is currently recruiting patients for various expansion trials (e.g., NCT04900363). Ultimately, one could envision bsAbs and other standard targeted agents being utilized together to deliver safe and multi-potent regimens that instill lasting immunity by ablating tumor angiogenesis and IC consequences. The inclusion of bsAbs to antagonize and/or stimulate multiple targets while tethering disparate cells in close proximity has numerous clinical applications with a favorable outlook of improved treatment efficacy for the patient.

Conclusions

ICIs have helped position immunotherapy at the forefront of cancer treatment. It is clear that for select patients and malignancies, IC blockade serves to best elicit protective immune effector responses against cancer growth and progression. However, a number of daunting challenges still face the field in order to more effectively and safely provide durable responses in substantial patient segments. Namely, improved strategies most be adopted that overcome the immunosuppressive and vascularized properties of the TME that ingratiates initially intractable or relapsed lesions. Additive therapies (such as FDA approved co-applied anti-vascular drugs and ICIs) have already provided the proof-of-concept of improved efficacy for certain malignancies but can be severely held back by inefficiencies in tumor penetrance and increased adverse events. The emerging precedence for engineered antibodies such as bsAbs provides a potential solution for developing and integrating complex therapeutic regimens in individuals with cancer, and, indeed, relevant molecules are currently in early phases of clinical evaluation. In the coming years, it will be exciting to observe how these novel agents may help further springboard the already favorable outcomes of ICI/anti-angiogenic strategies in patients.