Oncologists and immunologists have worked for many years on immunologic approaches to treat patients with cancer, but only recently have there been breakthroughs that have been successfully used to treat patients (Fig. 1). A major reason for interest in immunologic treatments is that most established therapies for cancer rely on drugs (chemotherapy) or radiation that kill dividing cells or block cell division, and these treatments have harmful effects on normal proliferating cells. As a result, the treatment of cancers causes significant morbidity and mortality. Immune responses to tumors can theoretically be highly specific for tumor cells and will not injure most normal cells. Therefore, immunotherapy has the potential for being the most tumor-specific treatment that can be devised. Recent advances in identifying tumor antigens and mechanisms of tumor immune evasion, and methods for genetically modifying T cells so they are specific for antigens on tumors, have ushered in the era of tumor-specific immunotherapy. The breakthrough approaches now in practice that stimulate the immune response to control tumors are not entirely tumor antigen specific and do have side effects of damaging normal tissues. Nonetheless, these approaches provide great benefit to many patients.
Fig1. History of cancer immunotherapy. Some of the important discoveries in the field of cancer immunotherapy are summarized. BCG, Bacillus Calmette-Guérin; CAR, chimeric antigen receptor; CTLA-4, cytotoxic T lymphocyte–associated protein 4; DC, dendritic cell; HPV, human papillomavirus; IL-2, interleukin-2; PD-1, programmed cell death protein-1; PD-L1, PD ligand 1. Modified from Lesterhuis WJ, Haanen JB, Punt CJ. Cancer immunotherapy—revisited. Nat Rev Drug Disc. 2011;10:591.
A second major reason to explore immunologic approaches for treating tumors is that cytotoxic drugs have been unsuccessful in achieving durable benefits in most cancers that have spread in the body beyond their site of origin. Because long-lasting memory is a cardinal feature of adaptive immune responses and immunity is systemic, it is possible that once an effective adaptive immune response to a tumor is initiated, it will be sustained for a long time and will be effective throughout the body. Because of this feature of the immune response, there is hope that some immunotherapeutic approaches will achieve long-term cures.
In this section, we describe the different strategies for tumor immunotherapy that are currently used clinically or in development.
Passive Immunotherapy With Monoclonal Antibodies and Antibody-Like Molecules
Passive antibody therapy involves the transfer of tumor-specific antibodies into patients, which is a rapid and theoretically very specific approach but does not lead to long-lived immunity. Some monoclonal antibodies have been in use to treat cancers for over 20 years, and about 100 are now approved (Table 1). Furthermore, recombinant single polypeptide proteins with antibody-like antigen-binding sites, called single-chain variable fragments, specific for tumor antigens have been developed for cancer treatment.
• Some antitumor antibodies bind to cell surface molecules on tumor cells and engage host effector mechanisms that kill the tumor cells (Fig. 2A). Several antitumor antibodies that are now approved for the treatment of certain cancers work in this way. For example, as mentioned earlier, anti-CD20 is used for treating B-cell lymphomas, and it works by depleting all CD20-expressing cells, including B cell-derived lymphoma cells as well as normal B cells. The mechanisms of depletion may include NK cell–mediated antibody-dependent cellular cytotoxicity, opsonization and phagocytosis by macrophages, and complement-mediated lysis.
• Immunotoxins, or conjugated monoclonal antibodies, are antibodies specific for tumor antigens that are linked to a chemotherapy drug or a radioisotope (Fig. 2B). The rationale for these agents is they will allow high local concentrations of the cytotoxic drug or isotope to be delivered to the tumor cells because of the antibody specificity. Early attempts to treat tumors with antibody-drug conjugates often failed in clinical trials because of significant systemic toxicity due to the nonspecific accumulation of the toxic component in various tissues. This problem has in part been addressed by designing linkers between the drug and antibody that are resistant to proteolytical cleavage in the circulation, thereby decreasing the release of the toxins systemically. There are now many antibody-drug-conjugates approved for clinical use, including those specific for HER2/ NEU to treat breast cancers, CD30 to treat Hodgkin lymphoma, CD33 to treat acute myeloid leukemia, CD22 to treat some acute lymphoblastic leukemias, CD79b (Igβ, a component of the B-cell receptor complex) to treat some B-cell lymphomas, and others.
• Some monoclonal antibodies used in cancer therapy bind to growth factor receptors on cancer cells and interfere with the signaling required for tumor growth and survival (Fig. 2C). Anti-HER2/NEU is an approved monoclonal antibody used to treat breast cancers that overexpress the cell surface growth factor signaling molecule HER2/NEU. An antibody that binds and blocks the function of the epidermal growth factor receptor (EGFR) is approved for the treatment of metastatic colorectal cancers and head and neck cancers. Another antibody in clinical use for several cancers blocks vascular endothelial growth factor (VEGF), which stimulates the angiogenesis that is required to maintain tumor growth. In this case, the antibody does not bind to the tumor cell.
• Bispecific T-cell engagers (BiTEs) facilitate the targeting of host T cells of any specificity to attack tumor cells (Fig. 2D). These reagents are recombinant proteins engineered to express two different antigen binding sites, one specific for a tumor antigen and the second specific for a T-cell surface molecule, usually CD3. In BiTEs, each antigen binding site is composed of a single-chain variable fragment, which is a singe polypeptide containing tandem immunoglobulin (Ig) heavy- and light-chain variable domains. The presumed mechanism of action of BiTEs, based on in vitro studies, is the formation of immune synapses between the tumor cells and the T cells and the activation of the T cells by CD3 ligation. BiTEs specific for CD3/CD19 and CD3/ CD20 have been approved for treatment of B-cell leukemias and lymphomas, and both CD3/BCMA-specific and CD3/ GRPC5D-specific BiTEs are approved to treat myeloma. BiTEs specific for many other tumor antigens have been developed, including EpCAM, HER2/NEU, EGFR, CEA, folate receptor, and CD33, and are at various stages of clinical trials.
Table1. Selected Monoclonal Antibodies Approved for Cancer Therapy
Fig2. Mechanisms of action of antitumor antibodies. Many different monoclonal antibodies that bind directly to molecules on the surface of cancer cells are used as drugs to treat cancers. Some of the main mechanisms by which these antibodies work to kill cancer cells are shown. Other monoclonal antibodies used for cancer immunotherapy that target soluble growth factors or cell surface molecules on immune cells are not shown. CTL, cytotoxic T lymphocyte; FcR, Fc receptor; NK, natural killer cell; TCR, T-cell receptor.
Adoptive Cellular Therapy With Antitumor T Cells
Adoptive cellular immunotherapy is the transfer of immune cells that have antitumor reactivity into a tumor-bearing host. The immune cells are derived from a cancer patient’s blood or solid tumor and then are treated in various ways in vitro to expand their numbers and enhance their antitumor activity before reinfusion back into the patient.
Adoptive Cellular Therapy With Autologous Tumor-Specific T Cells
T cells specific for tumor antigens can be harvested from a patient’s tumor tissue or blood, expanded and activated in vitro, and infused back into the cancer patients (Fig. 3A). This approach has been used in various trials for many years, but has had limited success, probably because the cells that are isolated from patients contain a low frequency of potent tumor specific T cells. With the advent of the technologies discussed earlier to identify the neoantigens that drive tumor-specific T-cell responses in individual patients, there is renewed hope for adoptive therapy with T cells specific for these antigens. The approach will involve harvesting T cells from the blood or tumors of patients, stimulating the cells with tumor neoantigens in vitro to increase the numbers and functional activity of cells specific for the tumor, and then transferring the activated T cells back into the patient. There have already been some successes with small trials using this approach in patients with melanoma. Another T-cell adoptive therapy approach currently in development is to transduce a patient’s T cells to express T-cell receptors (TCRs) specific for a tumor antigen commonly expressed by a particular tumor type and displayed by a known MHC molecule. The genes encoding the TCR polypeptide chains can be derived from cloned T cells selected for their ability to recognize particular tumor antigens. The genes encoding the TCRs are incorporated into a lentiviral expression vector that can be used to infect T cells taken from any patient with that tumor type and with the same MHC molecule. The TCR genes may be optimized by introducing mutations that enhance the affinity of the TCRs for the antigen. Clinical trials are ongoing using such engineered T cells specific for tumor antigens, including the cancer-testis antigen NY-ESO and the Wilms tumor-1 antigen. Recently, T-cell therapy using TCRs specific for a tumor anti gen, MAGE-4, has been approved for the treatment of a rare soft tissue tumor called synovial sarcoma that is resistant to most types of immunotherapy.
Fig3. Adoptive T-cell therapy. (A) T cells isolated from the blood of a cancer patient are activated in vitro to expand their numbers and induce effector phenotypes and are then infused back into circulation of the same patient. Some of these T cells that are specific for tumor antigens and will be activated by and kill the tumor cells expressing those antigens. (B) Chimeric antigen receptors (CARs) are genetically engineered membrane-bound receptors that recognize tumor antigen by antibody-like binding sites and generate intra cellular signals that activate T cells. CAR T cells are generated by transducing a cancer patient’s blood T cells in vitro with viruses engineered to express the CARs and are expanded in vitro before infusion back into the patient. CAR T-cell therapy has been successful for treatment of certain B-cell–derived tumors including leukemias, lymphomas, and multiple myeloma. IL-2, Interleukin-2.
Chimeric Antigen Receptor T-Cell Therapy
Adoptive therapy using T cells expressing chimeric antigen receptors (CARs) has proved successful in some hematologic malignancies, and this approach is being developed for other tumors (see Fig. 3B). CARs are genetically engineered mem brane-bound receptors with tumor antigen-specific binding sites encoded by recombinant immunoglobulin (Ig) variable genes (i.e., single-chain variable fragments) and cytoplasmic tails containing signaling domains of both the TCR complex and T-cell costimulatory receptors (Fig. 4). The reason for using an antibody-like binding site as the tumor antigen recognition receptor is that this approach avoids the problem of the MHC restriction of TCRs, so the same CAR construct can be used for a particular tumor type in any patient. In addition, the CARs will be able to recognize tumor antigens even if the tumors stop expressing MHC molecules, which, as discussed earlier, is a frequent mechanism of immune evasion. The genetically engineered cytoplasmic tails of CARs contain signaling domains that normally serve critical roles in T-cell activation. Several variations of signaling constructs have been used so far in CARs developed at different centers, but all contain the TCR ζ chain immunoreceptor tyrosine-based activation motif motifs and the cytoplasmic signaling motifs of costimulatory receptors such as CD28 or CD137 (4-1BB). These signaling motifs con fer on the tumor-specific Ig-like receptor the ability to potently activate T cells.
Fig4. Chimeric antigen receptor (CAR). CARs are composed of an extracellular immunoglobulin single-chain variable fragment specific for a tumor antigen, and cytoplasmic signaling domains that activate T cells, such as the ζ chain immunoreceptor tyrosine-based activation motif (ITAMs) and motifs in the cytoplasmic tails of the costimulatory receptors CD28 and 4-1BB, which promote robust T-cell activation. VH , V region of one heavy chain; VL , V region of one light chain.
In current protocols, a cancer patient’s peripheral blood T cells are isolated, infected with CAR-encoding retroviral or lentiviral vectors, and then stimulated with anti-CD3 and/or anti-CD28 antibodies to expand their numbers. The expanded CAR-expressing T cells are then injected back into the patient (see Fig. 3B). Prior to transfer, the patients are usually treated with drugs that deplete their own lymphocytes, which maximizes the proliferation of the transferred CAR T cells. The transferred T cells undergo robust expansion in the patient, in response to tumor antigen recognition by the Ig domain of the CAR and activating signals provided by the signaling domains. The specificities of the endogenous TCRs on these T cells (which are still present) become irrelevant to the goal of killing tumor cells because all the CAR-expressing T cells can be activated by the tumor antigen that binds to the receptor encoded by the CAR gene. Tumor killing is achieved by both direct cytotoxic and cytokine-mediated mechanisms. CAR T cells specific for CD19, a pan-B-cell marker expressed on B-cell–derived tumor cells, is approved for treatment of various B-cell malignancies that are refractory to other treatments, including chronic lymphocytic leukemia, acute lymphoblastic leukemia and B-cell lymphomas, and CD20-specific CAR T cells are approved for B-cell lymphomas. CAR T cells specific for BCMA, a pan plasma-cell marker, have been approved for treatment of multiple myeloma, a plasma cell tumor. Memory CAR T cells may persist in the treated patients so that surveillance against tumor recurrence is maintained. The technologies to produce large numbers of CAR T cells for each patient in a short time have advanced greatly, and CAR T cells are now used in many medical centers.
There remain some significant roadblocks that will need to be overcome for successful expansion of the use of CAR T-cell therapy.
• The tumor antigens targeted by CAR T cells may be expressed by some normal cells, which also will be killed. In the case of the CARs in clinical use specific for CD19 and BCMA, normal B cells or plasma cells are killed, respectively, and hypogammaglobulinemia occurs in a significant number of patients. If necessary, patients can be supplemented with pooled Ig to make up for the lack of antibody-secreting cells. Because long-lived antibody-producing plasma cells found in adult bone marrow and mucosal tissues do not express CD19 they continue to provide antibody-mediated immunity in adult patients treated with CD19-specific CAR T cells.
• Another problem commonly encountered in CAR T-cell therapy is cytokine release syndrome, a dangerous adverse reaction that frequently occurs soon after adoptive transfer of the T cells into patients with a high tumor burden. In these patients, so many of the T cells become activated at the same time that an intense systemic inflammatory response occurs due to the cytokines secreted by the T cells, which in turn stimulate cytokine release from macrophages and other cell types. Some patients who develop this reaction have been successfully treated with anti-IL-6 receptor antibody alone or in combination with IL-1 blockade. The severity of this complication can be decreased by treating patients with cytotoxic chemotherapy to reduce the tumor burden before CAR T-cell transfer. This limits the magnitude of the CAR T-cell activation that occurs after transfer. Other patients have developed serious neurotoxicity after CAR T-cell infusion, which may be related to microvascular injury or the effects of secreted cytokines that get into the brain, and the risk for long-term damage to the central nervous system remains a concern, especially in children whose brains are incompletely developed.
• Tumors may lose the expression of the antigen being targeted by the CAR and recur. The recurrence rate ranges from 20–50% in different tumors. The selection pressure imposed by the CAR T cells likely promotes the emergence of antigen loss tumor variants. This problem may be reduced by simultaneously expressing CARs specific for two (or, theoretically, more) tumor antigens; the chance of a tumor clone losing multiple antigens is low.
• In some patients, transferred CAR T cells appear to become unresponsive over time and initially controlled tumors have recurred. The CAR T cells in these patients express markers of exhaustion, including high levels of PD-1, and resemble the exhausted T cells mentioned earlier. Several approaches have been explored to address this problem, including treating the CAR T-cell recipients with blocking antibodies specific for PD-1, or using genome editing methods to eliminate the PD-1 gene in the CAR T cells. To avoid the risk for autoimmunity induced by the PD-1-negative T cells, endogenous TCRs have also been eliminated from the T cells before transfer. This creates T cells that have only the introduced tumor-specific antigen receptor and do not become exhausted because they lack a major immune checkpoint molecule.
• Attempts to treat solid nonhematologic tumors with CAR T cells have been hampered by difficulties in getting the T cells to migrate into the tumors and become activated there. Strategies to genetically engineer the CAR T cells with enhanced migratory functions are in development, but additional efforts to modify the immunosuppressive TME will likely be necessary.
• Another challenge in the development of effective CAR T-cell treatment of common solid tumors is identifying optimal antigen targets that are not also expressed on normal cells whose destruction would cause serious toxicity. The reason for this may be that cell surface antigens, which have to be the targets of the CARs, are usually not tumor specific and are often differentiation antigens or signaling receptors that are expressed at some level on normal cells of particular lineages. One approach to overcome this problem is to identify pairs of antigens that are expressed together at high levels only on tumor cells and to transduce T cells with two different CARs, each specific for one of the antigens and each with different signaling domains, both of which have to be engaged to activate the T cells. It is hoped that antigens can be identified such that the likelihood of tumor cells expressing both antigens is greater than that of any normal cells expressing both.
Studies are also ongoing to express CARs in cells other than CD8+ T cells, such as NK cells and γδ T cells, with the hope that these will be effective in killing tumors but will cause less toxicity than do CAR T cells.
Immune Checkpoint Blockade: Targeting T-Cell Inhibitory Pathways
Blockade of T-cell inhibitory molecules has emerged as a promising method for effectively enhancing patients’immune responses to their tumors. This approach is based on the idea that tumor cells exploit various normal pathways of immune regulation or tolerance to evade the host immune response, as discussed earlier. Because these inhibitory mechanisms establish checkpoints in immune responses, the approach of stimulating immune responses by a drug that blocks the inhibitors is called immune checkpoint blockade (Fig. 5). The first drug developed in this class is a monoclonal antibody specific for CTLA-4, the receptor on T cells that binds B7-1 and B7-2 and reduces T-cell costimulation, presumably both during T-cell priming in secondary lymphoid organs, and in the activation of memory T cells in tumor infiltrates. Anti-CTLA-4 was first approved as a therapy for advanced melanoma, and it is effective in stop ping or slowing tumor progression in many, but not a majority, of treated patients. This antibody may work by blocking the action of CTLA-4 expressed on Tregs and on activated T cells. As discussed earlier, T-cell responses against tumors also may be inhibited by the PD-L1/PD-1 pathway, which works by activating a phosphatase that blocks the activation of effector and memory T cells, and is required for the development of T-cell exhaustion. Antibody blockade of PD-1 or its ligand PD-L1 appears to be even more effective than anti-CTLA-4 in enhancing T-cell killing of tumors and halting the progression of otherwise lethal advanced cancers. Anti-PD-1 and anti-PD-L1 antibodies also cause less severe adverse effects (described later) than does anti-CTLA-4, and these antibodies are now approved for the treatment of many types of cancers, including melanoma, lung carcinomas, renal carcinomas, bladder carcinomas, colon carcinomas, Hodgkin lymphoma, and others. In fact, anti-PD-1 is approved for all recurrent or metastatic tumors with mismatch repair defects, which result in high levels of mutations and thus production of abundant neoantigens in every histologic type of tumor. This is the first cancer therapy that has been approved based on a genetic sig nature of the tumor regardless of the tissue or cell of origin of the tumor. It is likely that the antitumor T cells that respond to this type of therapy in each patient are CD8+ T cells that recognize neoantigen-derived peptides presented by MHC-I. Combined blockade of both PD-1 and CTLA-4 appears to be more effective against certain cancers than either alone and is approved for several cancers. LAG-3 is a third immune checkpoint molecule that is targeted together with PD-1 in an approved therapy for melanoma.
Fig5. Checkpoint blockade. Tumor patients often mount ineffective T-cell responses to their tumors because of repetitive tumor antigen activation of T cells, causing upregulation of inhibitory receptors such as cytotoxic T-lymphocyte antigen 4 (CTLA-4) and programmed cell death protein-1 (PD-1) leading to T-cell exhaustion. Blocking anti-CTLA-4 antibodies or anti-PD-1 or anti-PD ligand 1 (PD-L1) antibodies are highly effective in treating several types of cancers, likely working by blocking PD-1 and CTLA-4 on memory T-cell precursors of exhausted T cells, thereby promoting their differentiation toward functional antitumor effector T cells. TCF1 and TOX are transcription factors expressed by precursors of exhausted T cells, and terminally differentiated exhausted T cells, respectively.
Common adverse effects of checkpoint blockade treatment of cancers are autoimmune and inflammatory reactions, collectively called immune-related adverse events, which is predictable in light of the known roles of CTLA-4 and PD-1 in maintaining self-tolerance and regulating T-cell responses. The most frequent adverse events involve inflammation of the skin, colon, lung, liver, and various endocrine organs, although many other organs and tissues, including muscles, the heart, and central nervous system, can be affected. Some of these inflammatory dis orders are unusual in the absence of checkpoint blockade, such as autoimmune destruction of the anterior pituitary gland, or acute-onset and rapid destruction of pancreatic islets, leading to insulin-dependent diabetes. Most of these adverse events can be successfully treated with antiinflammatory medications such as corticosteroids or corrected with hormone replacement therapy, but many require cessation of the checkpoint blockade therapy and some, such as myocarditis, have a high mortality rate.
Overall, only about 15% to 20% of patients with cancer treated with anti–CTLA-4 and/or anti–PD-1/PDL-1 have objective improvement in their disease, and, among these, some tumors recur after an initial response. The response rate varies in different tumor types. Some, such as melanomas, respond well, but others, such as gliomas and pancreatic cancers, do not. There are several possible reasons for these therapeutic failures.
• Checkpoint blockade therapy is unlikely to work in patients with tumors that have relatively few somatic mutations encoding neoantigens because there will be few clones of tumor-specific T cells that can respond.
• Many tumors do not take advantage of the CTLA-4 or PD-1/ PD-L1 pathways as a strategy to evade antitumor immunity, but rather employ other immune evasion mechanisms. Consistent with this concept, low levels of expression of PD-L1 on some tumor types, detected by immunohistochemistry, predict a poor response to anti-PD-1 therapy.
• Tumors that initially evade immunity by engaging the CTLA-4 or PD-1/PD-L1 pathways may become resistant in the presence of the strong immune response, by selective outgrowth of tumor subclones that can evade T-cell killing by other mechanisms.
An important goal of cancer immunologists and oncologists is to identify histologic or genetic features of tumors or circulating biomarkers that may predict which patients will respond best to which checkpoint blockade therapy. The nature of the cellular infiltrate around the tumor has some predictive value of the response to checkpoint blockade. In general, abundant effector T cells, even if they have the phenotype of dysfunctional (or exhausted) cells, predict a good response, whereas sparse cellular infiltrates or an abundance of Tregs predict poor responses. In the future, assays for T cells expressing antigen receptors (TCRs) specific for neoantigens may be combined with analysis of neoantigen abundance to provide greater predictive value. The presence of B cells in the tumor infiltrate also correlates with good responses to checkpoint blockade. It is not known if this is because B cells and antibodies contribute to antitumor immunity or if the presence of B cells indicates a strong immune response, including T cells against the tumor.
To increase the percentage of patients who respond to check point blockade, oncologists are testing the efficacy of blocking other T-cell inhibitory molecules, including TIGIT and TIM3, alone or in combination with anti-PD-1 or anti-CTLA-4. Other approaches being tried include combining checkpoint blockade with more conventional radiation or chemotherapy protocols, or combining checkpoint blockade with tumor vaccines, kinase inhibitors that block oncogenic pathways in the tumors, or stimulating (agonistic) antibodies specific for activating receptors on T cells.
Vaccination With Tumor Antigens
Vaccination of tumor-bearing individuals with tumor anti gens may result in enhanced immune responses against the tumor. The earliest attempts to boost antitumor immunity relied on nonspecific immune stimulation. More recently, vaccines composed of killed tumor cells, recombinant tumor antigens, or DCs incubated with tumor antigens have been tested in animal models and in clinical trials with patients with cancer.
The identification of peptides recognized by tumor-specific CTLs and the identification of mutant genes that encode tumor specific antigens recognized by CTLs have provided many candidate antigens to include in tumor vaccines. New DNA sequencing technologies are now widely used to rapidly deter mine all the mutations in the protein-coding DNA sequences (exomes) of cancer cell genomes. Prediction algorithms are applied to these data to identify mutant peptides that are most likely to bind to the MHC alleles of each patient. These technical advances now allow for the precise identification of tumor-specific neoantigens in individual tumors, and this has stimulated efforts for the development of personalized tumor vaccines (Fig. 6).
Fig6. Detecting tumor neoantigens that elicit T-cell responses. Tumor DNA can be purified (1), and exome sequencing can detect random mutations in the genome of cancer cells (2). A computer algorithm then can be used to determine which mutations occur in amino acid sequences that encode peptides that would bind to the major histocompatibility complex (MHC) alleles in that patient (3). The validity of the putative neoantigenic peptides can be tested by assays of the patient T-cell response to these peptides in vitro or by testing if MHC-peptide multimeric complexes can bind to the T cells (4). This approach is being used to create personalized tumor vaccines. gen, Generation; HLA, human leukocyte antigen.
Tumor vaccination strategies employ a variety of adjuvants and delivery methods.
• Proinflammatory molecules are used to enhance the numbers of activated DCs at the vaccination site. These adjuvants include toll-like receptor ligands, such as CpG DNA and mimics of double-stranded RNA (dsRNA), and cytokines such as granulocyte-macrophage colony-stimulating factor (GM-CSF) and IL-12.
• Tumor antigens are delivered in the form of DC vaccines. In this approach, DCs are purified from patients, incubated with tumor antigens, and then injected back into the patients. A DC-based vaccine is now approved to treat advanced prostate cancer, but it has not proved to be effective in most patients. Technical challenges with DC vaccines are that the cells have to be harvested from each patient and they require expansion in cell culture, which is difficult to standardize.
• DNA and mRNA vaccines and viral vectors encoding tumor antigens are being tested in clinical trials. These may be the best ways to induce CTL responses because the encoded antigens are synthesized in the cytosol of cells, such as DCs, and efficiently enter the MHC-I pathway of antigen presentation. The rapid advances in mRNA vaccine technology that occurred in response to the COVID-19 pandemic are being applied to tumor vaccine development; small clinical trials of an RNA vaccine encoding tumor neoantigens have shown some promising results.
Different types of personalized tumor vaccines are still being developed and over the next decade we may have a better understanding of their clinical potential. Unlike many standard vaccines against microbes, which prophylactically prevent infections, tumor vaccines are used as therapies that need to be effective in stopping the progression of tumors that have already developed. Because evasion of host immunity is a hallmark of developing tumors, therapeutic vaccines have to overcome these regulatory mechanisms to stimulate effective antitumor immune responses. Many tumor vaccines tried to date have targeted antigens that are commonly shared by the same tumor type in different patients, and these antigens are usually differentiation antigens also expressed on cells of the normal tissue from which the cancer arose. Vaccines using such antigens have generally not been successful, likely because the antigens in normal cells induce tolerance that also has to be overcome for the induction of effective antitumor immunity.
The development of virus-induced tumors can be reduced by preventive vaccination with viral antigens or attenuated live viruses. As mentioned earlier, HPV vaccines have been effective in decreasing the incidence of HPV-induced premalignant lesions and cancers in the cervix. This approach has also been extremely successful in reducing the incidence of feline leukemia virus-induced hematologic cancers in cats and in preventing Marek’s disease, a herpes virus-induced lymphoma, in chickens.
Other Approaches for Stimulating Antitumor Immunity
Several additional approaches have been used to enhance host immunity against tumors, with variable success.
Cytokine Therapy
Cancer patients can be treated with cytokines that stimulate the proliferation and differentiation of T lymphocytes and NK cells. These cytokines can enhance the activation of DCs and tumor-specific T cells, particularly CD8+ CTLs. Many cytokines also have the potential to induce nonspecific inflammatory responses, which by themselves may have antitumor activity. The largest clinical experience is with high-dose IL-2 given intravenously, which has been effective in inducing measurable tumor regression in about 10% of patients with advanced melanoma and renal cell carcinoma and is currently an approved therapy for these cancers. The use of high-dose IL-2 is, however, limited because it stimulates the production of toxic amounts of proinflammatory cytokines, such as TNF and IFN-γ, which act on vascular endothelial and other cells and lead to a serious vascular leak syndrome. Many attempts are ongoing to intro duce mutations in IL-2 that enhance its stimulatory capacity and reduce toxicity.
IFN-α is approved for the treatment of several cancers, including melanoma, certain lymphomas and leukemias, and AIDS-related Kaposi sarcoma. The mechanisms of the antineoplastic effects of IFN-α probably include inhibition of tumor cell proliferation, increased cytotoxic activity of NK cells, and increased MHC-I expression on tumor cells, which makes them more susceptible to killing by CTLs.
Other cytokines, such as TNF and IFN-γ, are effective antitumor agents in animal models, but their use in patients is limited by their toxic side effects. Hematopoietic growth factors, including GM-CSF and G-CSF, are used in cancer treatment protocols to shorten the periods of neutropenia and thrombocytopenia after chemotherapy or hematopoietic stem cell transplantation.
Oncolytic Viruses
Oncolytic viruses are genetically modified viruses that replicate in and cause lytic death of cancer cells and, in doing so, release tumor antigens that stimulate CTL responses. The first oncolytic virus approved for clinical use is TVEC (talimogene laherparepvec) to treat metastatic melanoma. TVEC is a herpes simplex virus in which genes have been deleted to enhance viral-induced cell lysis and reduce chronic infection of nerves. Furthermore, a gene that encodes a TAP inhibitor has been deleted, thus enhancing the MHC-I antigen presentation path way in infected cells, and a gene encoding GM-CSF has been added, which increases DC accumulation in the tumor micro environment. TVEC is injected directly into solid tumor sites but enhances antitumor CTL responses even at distant locations from the injection site, which is evidence that the oncolytic virus promotes systemic antitumor immunity. Clinical trials are being conducted to test the efficacy of TVEC on other tumor types, and other oncolytic viruses are in development.
Nonspecific Inflammatory Stimuli
Immune responses to tumors may be stimulated by the local administration of inflammatory substances or by systemic treatment with agents that function as polyclonal activators of lymphocytes. One of the oldest examples of tumor immunotherapy was practiced by 19th-century physician William Coley, who treated cancer patients with extracts of dead bacteria, so called Coley’s toxin. This approach may have been intermittently successful due to the induction of strong innate responses leading to the production of TNFs and other cytokines that caused acute inflammation that killed tumor cells. Nonspecific immune stimulation of patients with tumors by injection of inflammatory substances such as killed bacillus Calmette Guérin (BCG) at the sites of tumor growth has been used for many years. The BCG mycobacteria activate macrophages and thereby promote macrophage-mediated killing of the tumor cells. In addition, the bacteria function as adjuvants and may stimulate T-cell responses to tumor antigens. Intravesicular BCG is currently used to treat bladder cancer. Cytokine therapies, discussed earlier, represent another method of enhancing immune responses in a nonspecific manner.
Graft-Versus-Leukemia Effect
In leukemia patients treated by allogeneic hematopoietic stem cell (HSC) transplantation, the presence of T cells and NK cells in the HSC inoculum can contribute to the eradication of the tumor. The T-cell–mediated graft-versus-leukemia effect is directed at molecules present on the recipient’s hematopoietic cells, including the leukemia cells, which are recognized as foreign by the administered T cells. Donor NK cells respond to the tumor cells because tumors may express low levels of MHC-I molecules or they express MHC-I alleles not recognized by the donor NK cells. Recall that recognition of self MHC-I normally inhibits the activation of NK cells (see Chapter 4). The challenge in use of this treatment to improve clinical outcome is to minimize the dangerous graft-versus-host disease that may be mediated by the same donor T cells.
The remarkable recent advances in cancer immunotherapy promise to dramatically change the care of patients with these dreaded diseases. The success of checkpoint blockade for many solid tumors and of CAR T-cell infusion for hematologic malignancies has revitalized the field of tumor immunology. Although limitations and problems remain, the enormous effort being invested in this field makes it likely that further advances will happen rapidly.
