T lymphocytes include several functionally distinct cell populations that are essential for combating virally-infected and neoplastic cells. A large fraction of the T cells in healthy individuals are naive resting cells that require antigen priming to acquire effector functions. The priming process involves the engagement of multiple T cell surface receptors that activate signaling pathways, leading to T cell proliferation, differentiation and acquisition of effector functions. A signal from the T cell antigen receptor (TCR) is sine qua non for the activation process, as it provides the very early activation step and ensures that only T cell clones specific to the appropriate antigen are expanded. The TCR interacts with a peptide antigen-loaded major histocompatibility complex (MHC) molecule on the surface of antigen presenting cells (APCs). T cell activation and responsiveness are carefully regulated by a number of accessory co-stimulatory and co-inhibitory cell surface proteins, termed immune checkpoint receptors, that can be triggered by cognate ligands on the surface of APC or target cells. The balance between signals provided by the TCR and the immune checkpoint receptors determines the cell’s ability to respond and differentiate into a fully active immunocyte.
The immune checkpoint receptors are implicated in immune response initiation and the regulation of their intensity and duration. Two of the best studied T cell-specific co-inhibitory receptors are the cytotoxic T lymphocyte associated antigen-4 (CTLA-4) and programed cell death-1 (PD-1). The main function of these co-inhibitory receptors is to maintain self-tolerance and prevent autoimmunity. Furthermore, they participate in the termination of T cell-mediated immune responses and limit collateral tissue injury during anti-microbial immune responses.
Recent studies demonstrated that different types of cancers utilize immune checkpoints to their own benefit. By activating co-inhibitory receptors, these tumors can suppress T cell-mediated functions and promote their own escape from immune destruction. The appreciation of the regulatory role and the mode of operation of immune checkpoints in cancer diseases have led to the development of new strategies for cancer immunotherapy based on immune checkpoint blockade. Multiple clinical studies have demonstrated that immune checkpoint blocking antibodies can be highly efficient in inducing durable clinical responses in different types of malignancies.
Nevertheless, the clinical benefits of these antibodies have been limited to subsets of cancer patients and treatments that are frequently associated with a unique spectrum of toxicities, termed immune-related adverse events.
CTLA-4: Where it all begins
CTLA-4 (CD152) is the prototypic immune checkpoint receptor and the most studied in the context of cancer immunotherapy. It is a disulfide-linked homodimeric transmembrane glycoprotein expressed exclusively on T cells, and it participates in the repression of T cell proliferation, cell cycle progression and cytokine production[1,2]. The CTLA-4 protein is highly homologous to the co-stimulatory molecule CD28, and both receptors utilize the same ligands, CD80 (B7.1) and CD86 (B7.2), which are expressed on the surface of APCs[3-5].
CD28 co-stimulation is necessary for maximal T cell activation. It promotes T cell expansion via interleukin-2 (IL-2)-dependent and independent mechanisms, but requires a primary signal (signal 1) in the T cell, which is generated by TCR engagement with MHC-bound cognate antigens on APCs (Figure 1).
Figure 1 T cells are subjected to co-stimulatory and co-inhibitory signals through interactions of their plasma membrane receptors with ligands on the surface of antigen presenting cells (dendritic cells and macrophages), as well as tumor cells.
Targeting of the co-inhibitory receptors by specific antibodies block the receptor-coupled inhibitory signals and enhance T cell proliferation and cytokine production to augment overall antitumor T cell responses. A: A productive interaction of T cells with antigen presenting cells (APCs) results in rapid delivery of stimulatory signals, leading to T cell activation; B: T cell activation leads to surface expression of cytotoxic T lymphocyte associated antigen-4 (CTLA-4). Interactions between CTLA-4 and programed cell death-1 (PD-1) co-inhibitory immune checkpoint receptors with the CD80/86 and PD-L1 ligands, respectively, on the APC surface result in the delivery of inhibitory signals that downregulate T cell activation and cytokine production; C: Targeting of CTLA-4 and PD-1 co-inhibitory receptors on T cells or the PD-1 ligand PD-L1 on the APC (or tumor cells) can block the inhibitory signals, increase the intensity and efficacy of anti-tumor T cell responses, and induce durable clinical responses and therapeutic benefits in cancer patients.
In contrast to CD28, which is constitutively expressed on both resting and activated T cells, CTL4 is only marginally expressed on the outer surface of resting T cells and localizes predominantly in intracellular stores. In response to antigen stimulation, CTLA-4 is temporarily transported to the outer cell membrane of the memory and regulatory T cells. At this stage, CTLA-4 can out-compete CD28 for ligand binding and induce T cell suppression via several independent mechanisms[8-10]. One mechanism relates to the fact that CTLA-4 has a higher binding affinity for both ligands. Its transient expression on the surface of T cells prevents CD28 from interacting with CD80/CD86 and delivering co-stimulatory signals. Another mechanism reflects the ability of CTLA-4 to directly deliver inhibitory signals to effector T cells through its cytoplasmic tail, which associates with signaling proteins. Furthermore, CTLA-4 can suppress T cell responses by upregulating the activity of Treg cells. The engagement of CTLA-4 on Tregs delivers activating signals that enhance their suppressive activity and inhibit antitumor immunity[8,13].
The role of CTLA-4 in maintaining T-cell activation under control is well demonstrated in studies of CTLA-4-deficient mice generated by homologous recombination. These mice suffer from a CD4+ T cell-mediated lymphoproliferative disease that is driven by excessive CD28 signaling[14-16]. CTLA-4-deficient mice also exhibit impaired differentiation of Treg cells, resulting in massive lymphoproliferation, splenomegaly and lymphadenopathy, including multiorgan lymphocyte infiltration and tissue destruction. These results substantiate the inhibitory role of CTLA-4 in the regulation of both Th and Treg development and function and in immune homeostasis.
Studies on the biological role of CTLA-4 and its involvement in the suppression of T cell functions laid the groundwork for the development of new strategies for cancer immunotherapy, which are based on CTLA-4 blockade.
The most studied CTLA-4 blocking antibody that was used in clinical trials is Ipilimumab (Yervoy®, Bristol-Myers Squibb), which showed significant survival benefit for patients with advanced metastatic melanoma. In early 2011, Ipilimumab was the first anti-immune checkpoint antibody to be approved by the United States Food and Drug Administration (commonly known as the FDA) for the treatment of melanoma[18,19]. Ipilimumab enhances antitumor immunity by blocking the negative effects of CTLA-4 and augmenting effector cytotoxic T cell functions.
Clinical trials with a second anti-CTLA-4 antibody, tremelimumab (ticilimumab, CP-675,20, Pfizer and AstraZeneca), demonstrated its effectiveness in patients with melanoma, refractory metastatic colorectal cancer, and hepatocellular carcinoma.
Additional clinical trials using a variety of CTLA-4 blocking drugs, either alone or in combination therapy, are being conducted on patients with a wide range of tumors. Combination of anti-CTLA-4 antibodies with immunotherapy, chemotherapy, or radiotherapy was found to improve long-term survival of cancer patients suffering from different types of malignancies[23-26].
The PD-1 immune checkpoint receptor is expressed on activated T cells and helps to preserve self-tolerance and the prevention of autoimmunity. It was initially cloned in 1992 in a search for molecules involved in the negative selection of thymocytes that undergo programed cell death. The role of PD-1 as an immune checkpoint became clear in 2000 upon identification of its physiological ligand, programed cell death ligand-1 (PD-L1), which is expressed on APCs and on some tumor cells. These tumors can co-opt the PD-1/PD-L1 pathway by upregulating PD-L1 expression as a mechanism of immune resistance. Preclinical studies in animal models demonstrated that targeting the PD-1/PD-L1 pathway can serve as another promising immunotherapeutic strategy for augmenting endogenous T cell antitumor immunity.
In addition to its fundamental role in maintaining self-tolerance and preventing autoimmunity, the PD-1 (CD279) immune checkpoint receptor is also critical for terminating immune responses. Its absence or down-regulation may cause tolerance breakdown or induction of autoimmune responses. PD-1 is expressed on activated T lymphocytes, and upon engagement with its ligand, it delivers signals that counteract the TCR-induced signals and inhibit IL2 production and T cell proliferation.
PD-1-mediated intervention with TCR-coupled signals involves the inhibition of ZAP70 binding to CD3, which down-regulates ZAP70 phosphorylation and catalytic activity. This occurs concomitantly with a reduction in the phosphorylation and activity of protein kinase C-theta (PKCθ), which attenuates the NF-κB and AP-1 transcription factors. As a result, PD-1 signals can promote apoptosis of antigen-specific effector T-cells and induce opposite effects on Tregs, leading to their exemption from programmed cell death[32,33]. One possible explanation for these contrasting effects might rely on the spatially-regulated distinct functions of PKCθ in these two cell types[34-36]. Thus, PKCθ is recruited to the center of the immunological synapse of activated effector T-cells[34,35], but is sequestered away from the immunological synapse of Tregs, instead concentrating on the opposite pole. Therefore, it is highly probable that PKCθ activation in the temporally-organized distinct subcellular structures will result in PKCθ-mediated phosphorylation of different substrate proteins that regulate distinct biological processes.
Although signaling through PD-1 helps maintain homeostasis in the immune system, PD-1-induced signals in cancer patients may facilitate tumor progression by suppressing effector T cell functions, particularly the induction of effective anti-tumor immunity.
The realization that PD-1 blockade might help boost the immune system in cancer patients has led to the development of an array of PD-1 targeting drugs that have already demonstrated remarkable clinical success in experimental models and in cancer patients.
In December 2014, the first monoclonal anti-PD-1 antibody to gain FDA approval was nivolumab (Opdivo®, Bristol-Myers Squibb), which was used in a clinical trial (CA209037) in metastatic melanoma patients that had deteriorated after CTLA-4-blocking antibody treatment. The overall objective response rate of the PD-L1+ melanoma patients was approximately 40%. In 2015, nivolumab received FDA approval for the treatment of patients with metastatic squamous non-small cell lung cancer (commonly known as NSCLC) and metastatic renal cell carcinoma, whereby the treatment extended the overall survival rate of cancer patients in both study groups.
In recent years, nivolumab received additional FDA approval for the treatment of a wide range of cancers.
A combination therapy of nivolumab and ipilimumab, which simultaneously target PD-1 and CTLA-4, was also used in clinical studies and found to improve the outcome of patients with melanoma[38-40] and other types of solid tumors[41-43].
In August 2016, pembrolizumab (Keytruda®, Merck Sharp and Dohme Corp.) was the second anti-PD-1 antibody that received FDA approval for the treatment of metastatic head and neck squamous cell carcinoma patients. Several clinical studies are being conducted with pembrolizumab in patients with head and neck cancer (NCT02454179, NCT02707588), in addition to multiple clinical studies that evaluate the efficacy of newly designed PD-1-targeting drugs in a wide range of cancers.
PD-L1 [CD274; also known as B7 homolog 1(B7-H1)] is a 40 kDa type I transmembrane protein that serves as a ligand for PD-1. Under normal physiological conditions, PD-L1 plays a vital role in the regulation of Treg functions and suppression of immunity during pregnancy and autoimmune diseases[45,46].
Upregulation of PD-L1 on APCs delivers inhibitory signals to PD-1+ T cells, which keep immune responses at bay by terminating them once antigens have been eliminated.
However, some tumors can disrupt this equilibrium and manipulate the PD-1/PD-L1 checkpoint pathways by expressing PD-L1 and delivering signals that turn off effector T cells.
Observations that support this mechanism were made in several studies where high PD-L1 expression on the tumor cells correlated with reduced immune responses, increased tumor aggressiveness and relatively poor survival[48,49].
Checkpoint inhibitors that target PD-L1 can therefore prevent its interaction with PD-1, restore T-cell activation, and amplify antitumor immunity[47,50].
The first FDA-approved anti-PD-L1 antibody atezolizumab (Tecentriq®, Genentech) was authorized in 2016 for the treatment of metastatic non-small cell lung cancer and advanced and metastatic urothelial carcinoma[50,51]. The clinical results indicated that tumors with a high frequency of PD-L1+ tumor infiltrating immune cells demonstrated particularly high response rates.
Two additional anti-PD-L1 antibodies, avelumab (Bavencio®, Merck KGaA, Pfizer and Eli Lilly) and durvalumab (Imfinzi®, AstraZeneca), received FDA approval in 2017 for the treatment of Merkel-cell carcinoma and advanced bladder cancer, respectively. Additional PD-L1 blocking antibodies are currently being evaluated for the treatment of a wide range of cancers in multiple clinical trials[52-54].
Additional T cell-associated co-inhibitory pathways
Cancer immunotherapies directed against the CTLA-4- and PD-1/PD-L1 co-inhibitory receptors exhibited undisputed efficacy in selected types of cancer diseases, but many patients were nonresponsive to these therapies, and several tumor types are refractory to these therapies. To increase the repertoire of drug targets in different cancers, scientists are searching for additional immune checkpoint receptors and testing their usefulness as drug targets in cancer immunotherapy.
The T cell immunoglobulin and ITIM domain (TIGIT) protein is a novel inhibitory receptor discovered in a genomic search for genes expressed in T cells. Their protein domain structure represents a potential inhibitory receptor. The expression mechanism of TIGIT on the outer surface of T cells is somewhat similar to that of CTLA-4, as it is upregulated by T cell activation and is transient in nature.
Other inhibitory receptors are constitutively expressed on “exhausted” T cells in patients with chronic cancer, including lymphocyte activation gene 3 (LAG3; CD223) and T cell immunoglobulin and mucin-containing molecule-3 (TIM-3; CD366).
The newly identified inhibitory receptors are being analyzed for their effectiveness as blocking antibody targets, while other surface proteins, such as CD137, CD27, ICOS, and GITR activating receptors, are being evaluated for their effectiveness as agonistic targets that amplify anti-cancer immune responses.
TIGIT (also known as WUCAM, Vstm3, VSIG9) is part of the CD28 family-like receptors that are expressed on T cells and various other hematopoietic cells[55,58,59]. TIGIT agonists include CD155 (poliovirus receptor-PVR) and CD122 (PVRL2, nectin-2), which are expressed by immune and non-immune cells, as well as tumor cells. TIGIT is a co-inhibitory receptor with a role in tolerance induction and autoimmunity. In general, TIGIT-/- mice were found to be more sensitive to immunization with myelin oligodendrocyte glycoprotein peptide than wild-type mice, and they develop more severe experimental autoimmune encephalomyelitis (EAE). In contrast, TIGIT transgenic mice are less sensitive to immunization with myelin oligodendrocyte glycoprotein and develop reduced symptoms of EAE[58,60]. TIGIT-/- mice develop severe autoimmune responses in a model of collagen-induced arthritis and graft-versus-host disease, similar to the effects induced by anti-TIGIT blocking antibodies, indicating that TIGIT functions as an immunoreceptor that downregulates T cell-mediated immunity.
The mechanism of action of TIGIT and the membrane receptor CD226 is reminiscent of that of the CTLA-4/CD28 receptor pair, with TIGIT fulfilling the role of the co-inhibitory receptor that counterbalances co-stimulation mediated by CD226. Furthermore, TIGIT/CD226 expression kinetics are similar to those of CTLA-4/CD28, with the co-inhibitory receptor being expressed only following T cell activation, in contrast to the co-stimulatory receptor, which is constitutively expressed on resting T cells[55,58,59].
TIGIT expression on peripheral blood lymphocytes and in lymphoid organs of tumor bearing mice is relatively poor, but it is highly expressed on tumor infiltrating lymphocytes, including Tregs, in both mouse and humans[61,62].
The observations that TIGIT may impose negative effects on anti-tumor responses were made in studies in which tumor growth in TIGIT-deficient mice were retarded compared to tumor growth in wild type mice, and that co-blockade of TIGIT and PD-1 had an additive positive effect on lymphoid cell functions in melanoma patients. Further studies revealed that the progression of multiple myeloma in mice and humans was associated with high TIGIT expression in CD8+ T cells that exhibited reduced functions, while targeting of TIGIT with monoclonal antibodies increased the effector function of CD8+ T cells and prolonged the survival of multiple myeloma patients.
In another set of studies, TIGIT, but not CTLA-4 or PD-1, was found to be associated with natural killer (NK) cell exhaustion in mice and humans with colon cancer. Blockade of TIGIT prevented NK cell exhaustion and promoted NK cell-dependent and T cell-mediated anti-tumor immunity.
The knowledge accumulated thus far on the immune-inhibitory role of TIGIT in tumor bearing mice and human cancer patients suggests that future designed TIGIT-directed blocking strategies could lead to the development of highly effective and improved anti-cancer therapies.
A phase I clinical trial using anti-TIGIT monoclonal antibodies (MTIG7192A) in patients with advanced or metastatic tumors is in progress (NCT02794571).
Another phase I clinical trial (NCT03119428) in patients with locally advanced or metastatic solid tumors investigates the safety and pharmacokinetics of the OMP-313M32 antibody, which blocks TIGIT binding to PVR.
TIM-3 [also known as Hepatitis A virus cellular receptor 2 (HAVCR2)] belongs to the TIM family of cell surface receptor proteins, which consists of eight members (TIM-1-8) in mice and three members (TIM-1, TIM-3, and TIM-4) in humans. The individual TIM proteins differ in molecular structure and expression patterns, as well as in their regulatory functions and impact on T-cell responses. TIM-3 is expressed on T cells and additional hematopoietic cell types[67-69] and utilize the C-type lectin galectin-9 as its ligand.
Ligation of the TIM-3 receptor in vitro initiates signals that suppress T cell responses, and in vivo administration of anti-TIM-3 antibodies increase the severity of clinical symptoms in a mouse model of EAE. Because EAE is a T helper 1 (Th1)-dependent autoimmune disease, it has been suggested that TIM-3 plays a role in the induction of autoimmune diseases by regulating macrophage functions.
Additional studies demonstrated that both TIM-3-deficient mice and wild-type mice treated with a TIM-3-Ig fusion protein exhibited defects in the induction of antigen-specific tolerance[71,72]. These studies and others indicated that TIM-3 is an immune checkpoint receptor that functions to specifically limit the duration and magnitude of T cell-mediated immune responses and contributes to the overall maintenance of immune tolerance[71,72].
TIM-3 is expressed on a large fraction of T cells in cancer patients where it is predominantly upregulated in tumor-infiltrating lymphocytes[73-75]. In addition, TIM-3 expression was found on various types of cancer cells where increased expression was frequently associated with disease progression and shorter survival[73-75].
Preclinical studies in cancer patients demonstrated that TIM-3+ T cells exhibited the most suppressed phenotype and were among the most severely exhausted lymphocytes[76,77].
In addition, a large fraction of TIM-3+ T cells in cancer patients co-expressed PD-1, and the dual receptor-expressing T cells exhibited greater immune function impairments compared to T cells that expressed PD-1 alone[76-78].
These investigations and additional studies support the hypothesis that TIM-3- and PD-1-coupled signaling pathways cooperate to induce severe T cell anergy or active T cell suppression in cancer diseases.
In contrast to CTLA-4 and PD-1, TIM-3 does not utilize immunoreceptor tyrosine-based inhibition motifs (commonly known as ITIM) or immunoreceptor tyrosine-based switch motifs (also known as ITSM) to transduce its signals. It is therefore suggested that TIM-3 is unlikely to be functionally redundant with other ITIM/ITSM-containing checkpoint receptors and that targeting of TIM-3 might have additive or synergistic effects with those induced by anti-CTLA-4 or PD-1 antibodies.
In agreement with this hypothesis, targeting the TIM-3 pathway was found to have modest antitumor effects in various experimental models and preclinical cancer studies[79,80]. However, the combination of anti-TIM-3 and anti-PD-1 monoclonal antibodies had a synergistic effect in suppressing tumor growth[73,76,81,82].
Multiple clinical trials (NCT03489343; NCT02817633; NCT03099109; NCT03066648) using a variety of anti-TIM-3 monoclonal antibodies (such as LY3321367, MBG453 and TSR-022) alone or in combination with anti-PD-1 or anti-PD-L1 antibodies are currently in progress in different types of leukemia and solid tumors.
LAG3 is a CD4-related 70 kDa type I transmembrane glycoprotein that is expressed on activated CD4+ and CD8+ T lymphocytes. It binds MHC-II with a high affinity compared to CD4, suggesting that it might function as a CD4 competitor and negatively regulate TCR-induced signals leading to T cell activation[85,86].
A negative regulatory role of LAG3 was demonstrated in human peripheral blood cells, when anti-LAG-3 blocking antibody combined with superantigen stimulation led to the increased proliferation of CD4+ and CD8+ T cells. LAG3 was found to synergize with PD-1 in downregulating T cell functions and promoting immune evasion by cancer cells. Furthermore, dual anti-LAG3/anti-PD-1 antibody treatment cured most mice of established fibrosarcoma tumors that were otherwise resistant to single antibody treatment.
Several LAG3-targeted therapies are currently at various stages of preclinical and clinical development. Combination therapies of anti-LAG-3 and anti-PD-1 antibodies showed promising results in melanoma patients who had relapsed or were refractory to anti-PD-1/PD-L1 antibody therapy.
Multiple clinical trials (NCT02658981; NCT03489369; NCT02061761; NCT03250832) using a variety of LAG3 blocking reagents [such as TSR-033, Sym022, and BMS-986016 (relatlimab) monoclonal antibodies or a soluble LAG-3Ig fusion protein (IMP321)] alone or in combination with a variety of other drugs are currently in progress in different types of hematological malignancies and solid tumors.
Adverse effects of immune checkpoint therapy
Although immune checkpoint targeting has shown a great deal of promise in treating a wide range of cancers, drug-induced side-effects termed “immune-related adverse events’’ (irAEs) have been observed in numerous patients due to non-tumor-specific activation of T cells by the immune checkpoint blocking antibodies.
Side effects of immune checkpoint therapy range from mild to severe and may include fatigue, cough, fever, diarrhea, nausea, loss of appetite, skin rash, itching and nerve inflammation. In addition, the irAEs may affect multiple organs and systems, including the gastrointestinal tract, liver, kidney, central nervous system, endocrine glands, and the pulmonary, cardiovascular and hematological systems.
The frequency and severity of irAEs usually correlate with antibody dosage. The median time of onset of irAEs is about ten weeks after the start of treatment, although it varies with respect to the affected tissues and between individual patients and may occur even after cessation of immune checkpoint therapy.
While the occurrence of irAEs normally indicates that the immune checkpoint blockade has activated the patient’s immune system, its correlation with improved antitumor immunity remains controversial, and the general assumption is that irAEs are not required for obtaining an effective anti-tumor response.
Although the majority of irAEs are completely reversible, immunosuppression using glucocorticoids or other drugs is recommended in cases of severe adverse events. This temporary immunosuppression was found to have no influence on the overall survival rate.
The discovery of additional biomarkers and the development of new tools to predict the patient’s risk of developing irAEs will facilitate the application of the full therapeutic potential of immune checkpoint-targeting drugs.