At first glance, the idea of personalized medicine may not seem that new. After all, haven’t physicians always given personalized care? They meet with you individually, ask about your health, and take your blood to measure hormones, cholesterol, and the numbers and types of cells in your blood. If a prescribed drug doesn’t effectively treat a disease, they switch you to a different one. Isn’t that personalized medicine? Maybe—but not in the way it could be.
Today, when people refer to “Personalized Medicine,” it is generally in the context of using genomics, the science of looking at all of the information in the human genome, to tailor medical care to individuals based on their genetic makeup. The DNA in the human genome encodes the information that provides our biological machinery with the instructions for all of the molecular parts that compose the human body. Overall, there are about three billion letters in that set of instructions. When the three billion letters of one individual human are compared to those of another, unrelated person, there are something like three million differences. Those three million differences contribute significantly to what makes those individuals unique. One very tangible example of this is the physical similarities that are often seen between parents and their children, because children share much of their genetic variation with their parents. This is also why some diseases run in families.
One of the ideas behind the Human Genome Project was that it would lead to treatments that could be better tailored to the individual. By “reading” the sequence of an individual’s DNA, it is sometimes possible to predict whether or not that person will respond in the expected way to a particular drug. That information is then taken into consideration when a drug is prescribed as a treatment. For example, specific drugs such as Gleevec, Herceptin, Iressa and Tarceva, have been used over the past several years to treat leukemia, breast and lung cancer, and they’ve been shown to be the most beneficial in a subset of patients with specific genetic variations. In these cases, patients are tested prior to prescribing the drug. Patients who are unlikely to benefit are saved the expense of the drug, which can be tens of thousands of dollars, while those who will likely benefit from treatment have higher confidence of a good outcome.
Just as genetic variations can affect how patients respond to treatment, they also affect the way drugs are metabolized, or processed, by the body. For example, genetic variation is known to affect the pain-killing ability of codeine. If you have one variation, codeine is an effective painkiller, but, if you have a different variation, it is not. As we learn of more examples of genetic variation altering an individual’s response to drugs, more genetic tests will be given to determine how much of a drug will successfully treat a particular person. This is already done in cases of childhood leukemia to determine what dose of medicine a patient should be given. If a child suffering from leukemia has one genetic variation, they tend to metabolize the drug more quickly, so they require a much higher dose. If you give the high dose to a person with the genetic variation that would be appropriately treated with a lower dose, they are at risk for very severe side effects.
This approach could even allow drugs that might normally be taken off of the market because of side effects to be safely given to those who are unlikely to experience an adverse reaction. It has been estimated that there are more than two million occurrences of serious side effects from drugs each year in the United States, which may lead to more than 100,000 deaths per year. If the genetic variation responsible for the unwanted side effects of a drug is identified, a test for that variation could be given prior to treatment. This would prevent individuals at risk for side effects from receiving that treatment, allowing drugs that would otherwise pose a liability to the manufacturers—but could also benefit a large number of people—to be safely available.
This knowledge of the contribution of genetic variation to effective drug therapy is also likely to affect how drugs are developed and marketed. Currently, drugs must be tested in large and varied populations that include individuals who will respond well to the drugs, as well as some who might not. This means a drug’s potential could be underestimated because of those that did not respond well to it. However, if drug developers understood the contribution of genetic variation to the effectiveness of a drug, they could test drugs in a much smaller groups, comprised only of people likely to respond to the treatment with minimal side effects. While the drug in question might not treat everyone with a particular disease, it would very effectively treat a subset of those patients. This could make drugs that would otherwise not be seen as valuable available to those who need them, and potentially reduce the costs of drug trials.
Improved diagnosis of disease is another benefit of personalized medicine. The first example of great success was the ability to divide the disease B-cell lymphoma into two different diseases based on patterns of gene expression displayed by the cancer cells. By having the sequence of the human genome and knowing what genes are present in human DNA, researchers were able to produce a “gene chip”—a glass slide that has a test for every gene in the genome on it. The researchers could then ask which genes were actually used to produce a particular tissue—in this case, a B-cell in the blood. They compared the genetic instructions used in normal B-cells to those used in B-cells of patients with leukemia. It turns out that there are some key differences in the instructions, or genes, used in the disease. Using this approach, two distinct gene expression signatures were observed in different leukemia patients, leading to the conclusion that rather than being a single disease, the patients with different signatures had different diseases.
More recently, it has been shown that it is possible to distinguish between metastatic breast cancer and non-metastatic breast cancer by comparing the levels of expression of only seventy or so genes. Metastatic breast cancer spreads more rapidly, has a much worse prognosis, and requires much more aggressive treatment than does non-metastatic breast cancer, which tends to spread much less. The ability to distinguish them provides valuable information that can be used to tailor treatment to that particular tumor type.
Since genetic variation also has a significant impact on disease risk, an important aspect of personalized medicine will be prevention. By having knowledge about the genetic predispositions of a particular individual for a particular disease, their physician will be able to focus on prevention and screening for those conditions. Predisposition-based screening should lead to earlier intervention, if the disease does occur. In general, earlier diagnosis maximizes successful treatment and often reduces the costs associated with treatment when compared to a later diagnosis.
These advances, all based on the human genome sequence and our ability to measure natural genetic variation, will produce more effective treatments, more effective diagnoses, and may contribute to reducing health care costs. Most importantly, it will enable our physicians to provide an even more personalized approach to our healthcare—one informed by the ability to read our DNA.