HOW DOES GENETIC INFORMATION HELP US TREAT CANCER?
Genetic information may be useful in tailoring cancer treatment to the specific molecular characteristics of a tumor. Although cancers have traditionally been described by their tissue of origin (e.g., breast cancer, lung cancer, or prostate cancer), in reality cancers of the same tissue may look and behave very differently depending on which mutations are present and which genes are expressed. This is known as the cancer’s molecular “signature.” For example, breast cancer may be classified into various types based upon which proteins are expressed on the surface of the tumor cells. Breast tumors that express human epidermal growth factor 2 (HER2), estrogen receptor (ER), and progesterone receptor (PR), or are triple negative (do not express HER2, ER, or PR) behave differently and have different prognoses. Tumors that are HER2 positive are treated with medications that bind to HER2 (trastuzumab, lapatinib) and inhibit its activity. ER and PR are hormone receptors, and ER/PR positive tumors are treated with antihormonal therapies. Triple negative tumors have the poorest prognosis and are unlikely to respond to HER2-targeted therapies or antihormonal therapies. Such cancers are usually treated very aggressively with chemotherapy.
As more has been learned about the molecular signature of various cancer subtypes, therapies that are specifically targeted to those signatures have been developed. Conventional chemotherapy acts on all rapidly dividing cells and does not distinguish between cancer cells and normal cells. Chemotherapy may cause substantial side effects, because normal, rapidly dividing cells (e.g., those lining the stomach) are killed along with cancer cells. Radiation therapy is another general approach to eliminating rapidly dividing cells and has the advantage that it can be directed specifically at the tumor site; however, surrounding normal cells may still be affected. The normal cells that survive chemotherapy and radiation therapy may sometimes acquire harmful mutations that may cause them to become cancerous in the future. The new, targeted therapies attack tumor cells with greater precision and specificity than conventional chemotherapy approaches, and, therefore, can have enhanced antitumor effects and reduced associated side effects.
One notable example of the power of a targeted cancer therapy is imatinib, which is very effective for treating Chronic Myelogenous Leukemia (CML). Imatinib specifically inhibits the BCR-ABL fusion protein that is constitutively activated in CML. Another targeted therapy, erlotinib, inhibits a receptor involved in cell growth control, the epidermal growth factor receptor (EGFR). EGFR is often mutated in many types of cancer, and erlotinib actually binds tighter and, therefore, is more inhibitory against those mutated forms of EGFR than against normal EGFR. Erlotinib is efficacious in the treatment of tumors that carry those mutated forms of EGFR.
In the above examples, a sample of the tumor (e.g., biopsy) is tested to determine the molecular signature. Testing may be by genetic sequence tests (e.g., for BCR-ABL, mutated EGFR, or HER2 gene amplification) or tissue protein stains (e.g., for the presence of ER/PR receptors or HER2 protein overexpression). The results of the testing will guide the choice of treatment —it will be personalized for the individual.
Tailored treatment based upon results of molecular characterization of tumor cells is fairly common in oncology. The molecular characterization, however, is usually confined to those factors (aberrant proteins, mutated genes) that are typically associated with a given tumor type. Also, molecular characterization is usually limited to known prognostic factors and a handful of high-frequency “druggable targets,” that is, factors common in a given tumor type and for which targeted therapies exist. Rare but potentially “druggable” changes might be present but overlooked. Also, because the molecular characterization of the tumor is limited to already known, well- characterized factors, the information generated has limited value for advancing our overall understanding of cancer biology.