Health In The Headlines: Personalized Medicine

by Lorien E. Menhennett

“Wait and see.”

How often has your doctor spoken those words to you after handing you a prescription? Too often, it seems to me. And this is not where we should be in the year 2010, says Dr. Bryan Schneider, an oncologist who was interviewed in June on Indiana University’s Sound Medicine podcast.

Instead of “wait and see,” Schneider thinks doctors should be able to predict a patient’s reaction — both in terms of positive effects of the medication and its negative side effects — saving both lives and money.

This idea — called “personalized medicine” — is not really new. But it’s gaining traction and speed, with recent support from the Food and Drug Administration and the National Institutes of Health. And while the government is behind on regulating this new field (which has led to a number of problems, including reports of women unnecessarily having their ovaries removed), there are promises to catch up on that front.

So why isn’t all of our medicine personalized? Good question. It would be nice. But we are not there in terms of technology.

Personalized medicine is really about genetics. People react differently to medications because of variations in their DNA. The Mayo Clinic’s Web site explains it well: For example, one genetic variation might cause a drug to remain in your system longer than normal, leading to dangerous side effects. Another variation might make your body less sensitive to a medication, meaning it won’t work as well, and you might need a higher dose than other people.

If doctors could test for different genetic variations, and how those variations affected your body’s response to a medication, they could prescribe more efficiently. Doctors could choose both more effective medications, and more effective dosages, for individual people. This would save not only in terms of a “grief” factor, but it could save lives — according to the Mayo Clinic, some 100,000 Americans die each year from reactions to medications, and more than 2 million wind up in the hospital. If a doctor knew someone would react poorly, he or she could avoid prescribing a medication entirely. This philosophy would also save money (always a selling point). Many treatments — especially those for long-term illnesses such as cancer — cost tens, if not hundreds, of thousands of dollars per year. Prescribing selectively and effectively eliminates guessing, grief, and waste.

The problem is that there could literally be millions of genetic variations (no one’s sure how many). And sometimes a response to a medication may not be governed by one gene, but by many genes interacting with each other. We simply do not know enough about the interactions between genetics and medications.

There are, however, a few genetic tests available today that allow doctors to personalize a patient’s medical care, most of them related to cancer treatment.

One example a test for HER2, a gene involved in breast cancer. If a woman’s cancer is HER2 positive, the results indicate use of Herceptin, a drug that can shrink the tumor, get rid of HER2 cells that have spread beyond the original tumor, and possibly prevent recurrence of the HER2 positive cancer. HER2 negative tumors do not respond to Herceptin, so before Herceptin is used, patients are tested with either an immunohistochemistry test or a gene-amplification test.

Another example involves metastatic colorectal cancer. If the cancer expresses the protein epidermal growth factor receptor (EGFR), then a medication called Erbitux may be used in combination with other treatments. Erbitux inhibits EGFR (which helps the cancer grow), so the medication may slow or stop the spread of the cancer. But Erbitux only works if the cancer expresses EGFR, so a genetic test called an EGFR immunohistochemistry test is required first.

So where do we go from here?

Scientists are making great strides in discovering those genetic “variations” that interplay between disease and medications, according to a recent article in The New England Journal of Medicine (published online on June 15). The problem, according to this article, is that “there is insufficient evidence of a downstream market to entice the private sector to explore most of them [the genetic variations].”

Basically, drug companies don’t see the dollar signs. So they don’t invest anything.

The purpose of that New England Journal of Medicine article (called “The Path to Personalized Medicine”) is to lay out a plan to show how the Food and Drug Administration and the National Institutes of Health plan to “fill that void.”

The FDA and NIH plan to do this in several ways. The first step is an “integrated pathway” between academic research and approving treatments for clinical use. The FDA and NIH promise public-private partnerships, as well as a program called Therapeutics for Rare and Neglected Diseases (TRND), aimed at accelerating the development of drugs, developing “targeted” drugs, and rescuing “failed” drugs by finding certain groups of patients who might benefit from them.

A second step is collecting information about genetics and treatment outcomes. For this, the NIH is relying on its various research centers and hospitals.

And last but not least, the FDA has established the “Critical Path Initiative” to regulate this up-and-coming industry and science. It is, in my opinion, about time they did this. As the article authors admit, there have already been instances of regulatory failure: “Increasingly, however, the use of therapeutic innovations for a specific patient is contingent on or guided by the results from a diagnostic test that has not been independently reviewed for accuracy and reliability by the FDA.” In other words, doctors may be using tests that the FDA has not approved to guide their treatment decisions. Not good.

For example, the FDA approved the drug Rituxan for certain cancer treatments in 2006. According to the NEJM article, a particular lab company has since then been marketing a test that claims it can tease out the 20 percent of people who will not repsond to the drug from those who will respond to the drug. Says the article: “The FDA has not reviewed the scientific justification for this claim, but health care providers may use the test results to guide therapy.”


In addition, the article’s authors point to other issues with laboratory tests that lacked FDA oversight, including situations where women were given “false readings that reportedly led to the unnecessary removal of women’s ovaries.”

Apparently, these glitches will now be taken care of. I sure hope so, because personalized medicine might just be the future of medicine.