Immune System and Antigen Recognition
Immunotherapy for cancer is essentially the stimulation of the immune system via a variety of reagents such as vaccines, infusion of T cells, or cytokines. These reagents act through one of several mechanisms: 1) by stimulating the antitumor response, either by increasing the number of effector cells or by producing one or more soluble mediators such as lymphokines; 2) by decreasing suppressor mechanisms; 3) by altering tumor cells to increase their immunogenicity and make them more susceptible to immunologic defenses; and 4) by improving tolerance to cytotoxic drugs or radiotherapy, such as stimulating bone marrow function with granulocyte colony-stimulating factor (G-CSF). Video Download Free RealPlayer
The basic premise of immunotherapy for cancer, then, is to stimulate the immune system in some way to treat and even prevent cancer. Historical data show that the immune system clearly plays a role in cancer progression. For example, immunosuppression is associated with cancer development. In fact, cancer is 100 times more likely to occur in people who take immunosuppressive medications (e.g., for organ transplant or rheumatic disease) than in people with normal immune function.3
Lymphoma, skin, and cervical cancer are just a few types of cancer that have been associated with immunosuppression. Patients who have undergone renal transplantation have an estimated 3 to 5 times higher overall incidence of malignancy in long-term follow-up than in the general population,4 and this is felt to be due in part to long-term immunosuppression. Conversely, heightened antitumor activity of the immune system has been suggested in the many reports of spontaneous cancer regression.5
While we have not elucidated the precise nature of the immune system's role in cancer occurrence and regression, we do know that tumors are immunogenic. Cancer is caused by a variety of genetic defects that occur in genes that encode for proteins involved in cell growth. The components of the immune system, antibodies and T cells, do not recognize or respond to defective genes, but they do recognize and respond to the abnormal proteins the cancer-causing genes encode. Thus when we talk about stimulating the immune system to attack cancer, we are thinking in terms of B-and T-lymphocytes. Now that we know the structure of some tumor antigens, we are beginning to look at global stimulation of the immune system as well. First, we will explore the roles of the individual components of the immune system in cancer.
Antibodies have been the focus of research in the past 20 years or so, as researchers discovered that antibodies recognize and respond to antigens produced by cancer cells. Antibodies are types of protein made by certain white blood cells—B cells—in response to a foreign substance. Each antibody binds to a specific antigen. The more specific the antibody, the greater the strength of the antibody-antigen bond.6 All antibodies are composed of light and heavy polypeptide chains. The number of chains may vary considerably, but at a minimum there are two of each type. Every heavy chain is paralleled by a light chain at one of its ends.6 Fc receptors on the heavy chain regions confer antibody function.
Antibodies protect the body against invading agents in one of two ways: 1) by direct attack on the invader, and 2) by activation of the complement system, which has multiple means of destroying invading cells. Direct attack is accomplished by agglutination, precipitation, neutralization, or lysis. Agglutination involves binding in a clump multiple large particles with antigens on their surfaces (e.g., bacteria or red blood cells). The antigen-antibody bonding can also create a molecular complex so large that it is rendered insoluble, and precipitates (e.g., tetanus toxin). Neutralization occurs when the antibodies simply cover the toxic sites of the antigen. Lysis involves very potent antibodies capable of directly attacking cell membranes, thus rupturing the cell.
The direct attack of antibodies on antigens is not usually sufficient to completely protect the body. The major protective efforts of antibodies take place through the amplifying effects of the "complement" system, a collection of about 20 different proteins.6 Many of these proteins are enzyme precursors that are normally inactive, but can be activated by the antigen-antibody reaction. When an antibody binds with an antigen, a specific reactive site on the antibody is activated. This site binds with a molecule of the complement system and sets into motion a cascade of reactions. A single antibody-antigen combination can activate many molecules of the complement system, which in turn activate increasing amounts of enzymes. In this manner an amplified reaction occurs. The multiple end products formed help prevent damage by the invaders.
One of the more important complement effects includes opsonization and phagocytosis. A product of the complement cascade strongly activates phagocytosis by neutrophils and macrophages. Living cells that eat proteins, the neutrophils and macrophages dock onto the Fc receptor and literally eat the bound cell. This type of antibody-mediated cell kill is called antibody-dependent cell mediated cytotoxicity (ADCC). ADCC has the advantage of catalyzing T cell activity, as the digested foreign cell proteins are presented on the major histocompatibility complex (MHC) molecules of the antigen-presenting cell (APC) as peptides.
Antibodies have also been shown to kill cells by blocking growth mechanisms, particularly in cancer cells. Growth factor proteins such as HER-2/neu can become overexpressed after the gene that encodes the protein is amplified. Antibodies specific for HER-2/neu bind to the molecule on the surface of the cell. Binding antibody may also block the growth signal.
Monoclonal antibody therapies that specifically target growth factor receptors have resulted in some significant clinical responses. We have a good understanding of the potential immunologic function of monoclonal antibodies, but their main application may be their potential to alter the biology of the proteins they target. Dr. Bishop will review this mechanism in greater detail.
Cytotoxic and helper T cells
Another focus of immunotherapy for cancer is the role of the T cells. T cells are either cytotoxic (CD8+) or helpers (CD4+). Unlike antibodies, which react to intact proteins only, the CD8+ T cells react to peptide antigens expressed on the surface of a cell. Peptide antigens are proteins that have been digested by the cell, and presented as protein fragments—or peptides—displayed in the MHC. The peptide and the MHC together attract T cells. CD8+ T cells are specific for class I MHC molecules, while CD4+ T cells are specific for class II MHC molecules.
After attaching to the MHC-peptide complex expressed on a cell, the CD8+ T cell destroys the cell by perforating its membrane with enzymes or by triggering an apoptotic or self-destructive pathway. The CD8+ T cell will then move to another cell expressing the same MHC-peptide complex, and destroy it as well. In this manner cytotoxic T cells can kill many invasive cells. Ideally, CD8+ T cells could engender a very specific and robust response against tumor cells.
One of the biggest breakthroughs in tumor immunotherapy in the past several years has been our increased understanding of the role of helper T cells. The helper T, or CD4+ T cell, is the major regulator of virtually all immune system activities.7 These cells form a series of protein mediators called lymphokines that act on other cells of the immune system and on bone marrow. Some of the most important lymphokines secreted by the helper T cells include interleukin-2, interleukin-3, interleukin-4, interleukin-5, interleukin-6, granulocyte-monocyte colony stimulating factor (GM-CSF), and interferon-gamma. Without these lymphokines, the remainder of the immune system does not function as effectively as it would with the appropriate cytokine environment.
Like the CD8+ T cells, CD4+ T cells also recognize MHC-peptide complexes in the context of class II MHC. CD4+T cells augment the immune response by secreting cytokines that stimulate either a cytotoxic T cell response (Th1 helper T cells) or an antibody response (Th2 helper T cells). These cytokines can initiate B-cells to produce antibodies, or enhance CD8+ T cell production. The function of the CD4+ T cell depends upon the type of antigen it recognizes, and the type of immune response required.
Lymphokines produced by helper T cells also regulate macrophage response. The lymphokines slow or stop the migration of macrophages after they have been engaged, allowing macrophages to accumulate at the site. These lymphokines also stimulate more efficient phagocytosis, so they can destroy rapidly increasing numbers of toxins.7
The helper T cell amplifies itself by secretion of lymphokines, particularly interleukin-2. This action enhances the helper cell response, as well as the entire immune system's response, to foreign antigens.
As we have gained understanding of how T cells interact with the immune system microenvironment, and how antigens are recognized, we have been able to better assess the immune response in cancer patients. Understanding this antigen recognition pathway and the role of helper T cells in enhancing cytotoxic T cells or antibodies has allowed us to move past simply generating an antibody response, infusing antibodies, or generating a cytotoxic T cell response. Now in tumor immunology we are exploring the feasibility of stimulating an immune response that would be therapeutically effective. In order to do this, we must be able to identify the specific arms of the immune system that must be activated to recognize those particular antigens.
Tumors are Immunogenic
Tumor immunology in the past decade has made great gains: We know now that tumors are immunogenic. We know that T cells can function at the single cell level, because they are able to leave the endothelium and migrate into tissues where they can clonally expand until the antigen is eradicated. The most exciting aspect of stimulating an endogenous immune response, however, is the potential to initiate long-term immunologic memory. Video
This represents a dramatic shift in how cancer is treated. If we can focus this type of immunologic memory, targeting immunogenic proteins involved in malignant transformations, we may be able to prevent relapse. Of course, relapse is one of the major problems in long-term survival of cancer patients. Some patients can initially respond to chemotherapy, surgery, or radiation therapy, but tumors may recur.
As we have discussed, antigens are foreign substances recognized by and targeted for destruction by the cells of the immune system. When cells become cancerous they produce new, unfamiliar antigens. The immune system may recognize these antigens as foreign, and contain or even destroy the cancer cells. However, the immune responses elicited by tumor antigens are not robust. Most tumor antigens are "self" proteins, rendering them weakly immunogenic. Our immune systems tolerate self-proteins, and tolerance is a major mechanism by which cancer can evade immune recognition.
Many tumor antigens have been defined in terms of multiple solid tumors: MAGE 1, 2, & 3, defined by immunity; MART-1/Melan-A, gp100, carcinoembryonic antigen (CEA), HER-2, mucins (i.e., MUC-1), prostate-specific antigen (PSA), and prostatic acid phosphatase (PAP) are just a short list. Viral proteins – hepatitis B (HBV), Epstein-Barr (EBV), and human papilloma (HPV) – are important in the development of hepatocellular carcinoma, lymphoma, and cervical cancer, respectively. Even proteins as ubiquitous as p53, glycosylate proteins, and carbohydrates are tumor antigens. Some immune-based therapies targeting these tumor antigens are in phase III studies assessing whether immunizing against these antigens affects overall survival.
Many of these proteins are shared between multiple tumor types, and with molecular and cellular techniques investigators have defined more than 500 tumor antigens. These antigens have been elucidated by virtue of the fact that they elicit an immune response in patients who have cancer, but not in volunteer blood donors or people who do not have cancer. For instance, the blood of patients with cancer of the colon, breast, pancreas, bladder, ovary, or cervix may have high levels of CEA, and actually is used to help detect the presence of cancer. PSA levels may be high in men with benign prostate enlargement, but are typically much higher in men with prostate cancer.
The question now is why, if tumors are immunogenic, do cancers grow? We know an immune response is elicited, yet what prevents the immune response from obliterating the cancer?
One explanation is that in patients with cancer, the immune response is simply not robust enough. Ward et al.8 evaluated the endogenous HER-2/neu specific antibody response in patients with colorectal cancer. This protein is overexpressed in approximately 20% of human adenocarcinomas, and is a defined tumor antigen in breast cancer. HER-2/neu antibodies (titer > or = 1:100) were detected in 14% (8/57) of patients with colorectal cancer compared to none of the control population (0/200).8 Detection of HER-2/neu specific antibodies in the patient population was significantly associated with HER-2/neu protein overexpression in the patients' tumor (p < 0.01). Nearly half (46%) of the patients with HER-2/neu overexpressing tumors (6/13) and 5% of HER-2/neu negative tumors (2/44) had detectableHER-2/neu specific antibodies.
Immunity and tumor growth
The antibody responses to HER-2/neu generated by tumor overexpression is logs lower than what would be expected from an infectious disease vaccine. One might postulate that this difference is due to cancer-mediated immune suppression. Yet their response to tetanus vaccines was similar to that of the control population. While the cancer patients could respond quite actively to a foreign antigen vaccine, they did not respond so well to an endogenous vaccination of HER-2/neu overexpression.
This model suggests an insufficient response to the oncogenic protein. The cytokine environment does not allow amplification of helper T cells to occur.9-11 Several studies have examined the phenotype of these cells found at the tumor site, and found they are not functional. They are not generating a Th1 or Th2 response, and secreting low levels of cytokines.12-15 Antigen-presenting cells, critical to stimulating T cell activity, are also not functional.16, 17 Either immune receptor molecules have been downregulated or the most potent antigen-presenting cells are absent from the tumor site.
As tumors grow, they secrete immunosuppressant factors. This immunomodulatory effect occurs directly, by viral proteins binding to immune receptor molecules and thus preventing their expression on the surface of the virally infected cell, or by tumors secreting factors that downregulate immune activation.18, 19 We have sophisticated molecular techniques that allow gene array, following the immune phenotype of a malignancy through its progression. Many of the genes that encode chemokines and cytokines that would normally allow the immune system to function are lost in those cells as they become anaplastic tumors.
As previously stated, one of the essential things to understand about human tumor antigens is that they are self antigens. Unlike viral proteins, they are derived from otherwise normal cells whose biologic function has been altered in such a way that they no longer respond to the body's normal mechanisms for controlling cell growth and reproduction. They may be qualitatively distinct in that they are overexpressed, but they are simply cell cycle regulatory proteins gone awry. The immune system is designed to protect "self," and we now recognize this is a major mechanism by which tumors escape immunization. Many of the newer immunotherapeutic strategies focus on getting the immune system to recognize tumors as dangerous, mount a full-fledged attack, and eradicate the cancerous cells.
Immunotherapy is generally thought of as conferring either passive or active immunity. Passive immunity supplies the immune response – antibodies, cytotoxic T cells – rather than activating the immune system directly. These approaches have met with some success, albeit short-lived. Any element infused this way has a half-life, so for the effect to continue the infusion must be repeated. Since the immune system is not engaged, the attack may not be full-fledged. For example, when we infuse cytotoxic T cells we do not see the expansion in vivo with helper T cells. Video
Active immunity may be the ideal of immunotherapy. What we try to achieve with active immunity is an endogenous immune response, where the immune system is primed to recognize the tumor as foreign. This approach has not been successful in patients with widespread disease, as their immune systems are unable to mount a sufficient response. In the past several years, efforts have focused on using active immune therapies in patients with minimal disease. However, we have seen that the immune system can be quite functional despite advanced stage cancer when the patient has been treated to maximal response. That is, the cancer patient can be vaccinated.
Schiffman et al.20 demonstrated that patients with stage 3 and 4 breast cancer (n=35) could be immunized. We compared the immune response to tetanus, a recall antigen, and keyhole limpet hemocyanin (KLH), a neoantigen. These patients had all been treated with chemotherapy to maximal response. A tetanus booster was administered, and their immune response measured 2 months later. There was no difference in response between the cancer patients and the volunteer donors.
We saw a similar response with KLH. None of these patients had been exposed to this neoantigen. While antibody and T cell responses were negative prior to immunization, they were not different from immune responses in volunteer blood. After receiving a vaccine against KLH, most of these patients were able to generate significant, specific anti-KLH immunity.
The future of immunotherapy holds two specific challenges. First we need to discover antigenic formulations that target multiple antigens. Monovalent vaccines, which target a single protein, may not supply the long-term immunity necessary to prevent relapse. Single-antigen vaccines may have a role in advanced disease, if the targeted antigen is critical to tumor growth. Clinically useful tumor vaccines will have to immunize against multiple immunogenic proteins, targeting the important proteins involved in malignant transformation.
We also need to determine a therapeutically effective range for antitumor protection. We are beginning to think in terms of T cell precursor frequencies specific to different types of disease status. Precursor frequency defines the level at which immunity is conferred. For example, protective immunity against hepatitis B may occur at 1 in 10,000 T cells. In infections such as influenza precursor frequency during the active flu infection can get to 1 in 50 or 100 T cells specific for influenza. At this point, we do not know what precursor frequency would be necessary to have an effect on cancer. What levels confer immunity in patients with clinically undetectable disease – those who have been cured by chemotherapy but are at high risk for relapse? Where should we set the bar for patients with localized disease, or uncontrolled advanced stage disease?
Knutson et al.21 vaccinated 19 patients with a simple peptide mixed with GM-CSF. Precursor frequencies ranged from 1 in 60,000 at the lowest, which would not be considered effective immunization, to 1 in 3,000. Median precursor frequency in patients with previously undetectable levels was about 1 in 18,000. In this study, immune responses were long-lived and detectable for more than 1 year after final vaccination in certain patients.
Immunotherapy for cancer is in transition. It is clear that different strategies will benefit different patient populations. Because the struggle is between immune control and tumor escape, the best strategies to combat cancer will need to attack on multiple fronts. Most of our efforts now are focused on making self more immunogenic using immune system activators, supplying antigen-presenting cells, or actually predigesting some of these tumor antigen proteins into immunogenic peptides to thwart all tumor-evading mechanisms. Ultimately, an achievable goal may be a durable anti-tumor immune response that can be maintained throughout the patient's lifespan.