Chapter 67: Tumor Immunity

One of the key functions of the immune system is to patrol the body for cancer cells in order to kill these cells before they become malignant tumors. In this chapter we will discuss (1) how the immune system is able to recognize cancer cells; (2) how cancer cells evolve to hide from and suppress the immune system; and (3) how recent advances in immunotherapy overcome these strategies to re-invigorate the immune system’s killing mechanisms.

Malignant cancer cells are genetically similar to healthy cells. They generally display the same major (HLA) and minor histocompatibility proteins as other nucleated “self” cells, and lymphocytes with antigen receptors that could recognize these proteins are largely removed during negative selection (see Chapter 66). Nevertheless, animals carrying a malignant tumor can develop an immune response to that tumor and cause its regression. In the course of neoplastic transformation, new antigens (neoantigens), called tumor-associated antigens (TAAs), develop at the cell surface, and the host recognizes such cells as “nonself.” An immune response then kills the offending cells, preventing the formation of a malignant cancer.

TAAs can either be highly specific (i.e., cells of one tumor will have different TAAs from the cells of another tumor) or be shared by different tumors, even tumors that occur in different hosts. For example, virus-induced tumors (such as those due to human papillomavirus) tend to have TAAs that cross-react with one another if induced by the same virus strain. This feature has spurred interest in developing antitumor vaccines that use the shared, virus-induced TAAs found in all individuals who have that virus-induced tumor.

Some human tumors contain antigens that normally occur in fetal but not in adult human cells.

1. Carcinoembryonic antigen circulates at elevated levels in the serum of many patients with carcinoma of the colon, pancreas, breast, or liver. It is found in fetal gut, liver, and pancreas and in very small amounts in normal sera. Detection of this antigen can be helpful in the diagnosis of such tumors, and if the level declines after surgery, it suggests that the tumor is not spreading. Conversely, a rise in the level of carcinoembryonic antigen in patients with resected carcinoma of the colon suggests recurrence or spread of the tumor.

2. Alpha-fetoprotein is present at elevated levels in the sera of patients with hepatocellular cancer and is used as a marker for this disease. It is produced by fetal liver and is found in small amounts in some normal sera. However, it is nonspecific; it occurs in several other malignant and nonmalignant diseases.

   Monoclonal antibodies directed against new surface antigens on malignant cells (e.g., B-cell lymphomas) can be useful in diagnosis. Monoclonal antibodies coupled to toxins, such as diphtheria toxin or ricin, a product of the Ricinus plant, can kill tumor cells in vitro and someday may be useful for cancer therapy.

The immune response that attacks nonself tumor cells is directed by T cells. Such immune responses probably act as a surveillance system to detect and eliminate newly arising clones of neoplastic cells. The cells that infiltrate tumors include natural killer (NK) cells, which can kill cells directly or can react to cells bound by antibody (antibody-dependent cellular cytotoxicity); CD8-positive cytotoxic T cells; and activated macrophages, which rely on antigen-specific Th-1 cells and cytokines for activation.

In general, the immune response against one or a few tumor cells is effective. However, because tumor cells both proliferate and mutate rapidly, selection pressure favors those cells that can escape immune surveillance by evading or suppressing the antitumor immune response. Tumor cells do this through a number of mechanisms, including (1) decreased expression of major histocompatibility complex (MHC) class I–complexed TAAs; (2) release of soluble factors, such as the enzyme indoleamine 2,3-dioxygenase (IDO), that promote the activity of immunosuppressive leukocytes, including T regulatory (Treg) cells; and (3) expression of cell surface molecules, such as programmed cell death-ligand 1 (PD-L1) and cytotoxic T lymphocyte-associated antigen-4 (CTLA-4; see Chapter 60), that inhibit the function of cytotoxic T cells and NK cells.

Tumor antigens can stimulate the development of specific antibodies as well. Some of these antibodies are cytotoxic, but others, called blocking antibodies, can actually enhance tumor growth. This might be due to blocking recognition of tumor antigens by the host or due to binding of the Fc regions to inhibitory Fc receptors.

Therapies that boost the antitumor immune response, known as immunotherapies, are clinically effective, and the field is evolving rapidly. One of the longest used cancer immunotherapies is the BCG vaccine (bacillus Calmette-Guérin, a bovine mycobacterium). It was observed that injecting BCG into the lesions of skin melanoma and bladder cancer activates and recruits the immune cell into the tumor, leading to tumor regression. Treatment with interleukin-2 was shown to improve survival rates in certain melanomas and renal cell cancers, although interleukin-2 treatment is limited by the toxicity of a systemic inflammatory response.

Another approach to cancer immunotherapy involves the use of tumor-infiltrating lymphocytes (TILs). The basis for this approach is the observation that some cancers are infiltrated by lymphocytes (NK cells and cytotoxic T cells) that appear to be destroying the cancer cells. These lymphocytes are recovered from the surgically removed cancer, grown in cell culture until a large number of cells are obtained, activated with interleukin-2, and returned to the patient in the expectation that the TILs will “home in” specifically on the cancer cells and kill them.

Monoclonal antibodies directed against CTLA-4 and PD-1 (see Chapter 60) are effective in enhancing the immune response against cancer cells. CTLA-4 and PD-1 on T cells inhibit the co-stimulatory signal, and antibody against these proteins blocks this inhibitory effect. This removal of an inhibition, or checkpoint, enhances the immune response against the tumor, and therefore, this strategy is called checkpoint blockade immunotherapy. Not surprisingly, checkpoint blockade therapies are limited by the fact that some patients develop autoimmune diseases as a consequence of off-target immune activation.

Finally, another successful approach has produced sustained remission in patients with certain leukemias and lymphomas. These remissions are induced by infusion of chimeric antigen-receptor modified T (CAR-T) cells that target proteins on the surface of the leukemic B cells. When the CAR-T cells recognize these proteins, e.g. CD19 that is found on malignant B cells, they release cytokines, perforin, and granzymes, killing the cells much as a cytotoxic T cell recognizes and kills virus-infected cells (see Chapter 60).

These “cell-based” therapies are extremely expensive because they require manipulating each individual patient’s cells and then reinfusing them as an autologous transplant. In addition, CAR-T cells can cause systemic inflammatory responses due to excessive cytokine release, and because healthy B cells also have CD19, patients must take pooled intravenous immunoglobulin (IVIG) replacement to counteract the persistent B-cell deficiency that occurs. The first commercial CAR-T cell therapy was approved by the US Food and Drug Administration for B-cell leukemia in 2017.