Tuesday, October 16, 2007

How personal will Personalized Medicine become?

Personalized Medicine is a great idea in theory...the idea that the drugs that your doctor prescribes to you will be matched to your body's unique biological characteristics. But will it become reality? The "made for me" drug is unlikely, but the "best match for me" drug is both feasible and a significant improvement to the way most drugs are prescribed today.

The financing needed for drug development is substantially high. In order for a drug to be worth the cost of discovery, testing, and marketing (estimated upwards of 1 billion dollars per drug), it must have either 1) a large target patient population who can take the drug, or 2) have very high prices in order to recoup the sunken cost of making the drug. On an economics argument alone, the idea that a drug could be made specifically to match my personal unique biology is unreasonable. At best, we can hope to have some level of prescreening patients to match them with the drugs that are most likely to help and not harm them.

Currently, when most patients see a doctor for an ailment, they are given a treatment or drug in the hopes that it will help them based on the findings that it has previously worked on other people. This seems reasonable. However, this can sometimes be simply a hope. There have been studies that much (some studies even say most) of the prescribed drugs do not help and may even hurt the patients that receive them. Drugs often have the fine balance between the helpful effects (efficacy) and the harmful effects (adverse events). The drugs must go through clinical trials to prove their efficacy and to find the right amount to use (therapeutic dose) while avoiding as much of the adverse side effects as possible. One problem of course is that no two people are identical. Even "identical" twins have biological differences between them. Therefore, drug effects seen in people in the clinical trials will not capture the diversity of the general population. Clinical trials involve thousands of patients, but even these numbers may not be sufficient to catch rare side effects that only occur once in a million people. It is only once the drugs are approved for marketing and are actually prescribed and taken by millions of people do some of these rare adverse events become recognizable.

So often a delimma develops. What do we do with a drug that has been shown to help thousands or even millions of people, but some people not only fail to benefit from the drug but suffer adverse events? Often if the adverse effects are severe or lethal, the risk to the few outweighs the benefit to the many and the drug is pulled from the market. This makes sense because without knowing who will suffer the side effects, no matter how rare they may be, few doctors would be willing to take the chance nor would many patients want to take that chance depending on how severe the adverse event was compared to their ongoing illness. Personalized medicine has promised to help solve this issue. If a test can be made that would distinguish the patients who would benefit from those who would not, the test could then be used to prescreen patients before they take the drug. The drug would then be given only to those who would benefit. Thus, a helpful drug could be kept on the market rather than be removed.

However, lately personalized medicine claims have gone beyond this simple scenario. We are now in the era of genomics. It is possible to read the DNA sequences of individuals and gain insight into their genetic makeup. A large number of scientific publications have come out in the past two years linking DNA sequences that are commonly present in the population to various common diseases such as heart disease, obesity, certain forms of cancer, crohn's disease and others. These diseases are known as complex diseases because no single marker, gene, or environmental exposure appears to cause them, but rather they are due to the complex interactions and relationships of multiple DNA sequences (genes) and environmental exposures and events.

These recent findings (and many more to come) are very helpful in understanding the causes of complex diseases. However, while identifying DNA sequences that are linked to these highly prevalent diseases is very helpful, how that information is used and interpreted can be troublesome if not carefully managed. We are now in an era where the technology of obtaining DNA sequences far outpaces our ability to understand how the DNA sequences are biologically linked to disease. In other words, we are reading the text of the book much faster than our ability to comprehend what it means. What makes the situation even more complicated is that the DNA sequences being discovered are most often not direct causes of the disease but rather are susceptibility indicators. They are identified as being linked to the disease under study because they are found more often in people with the disease than in people without the disease at a statistically significant difference.....and here is where my main concern lies with the public's use of this information. It all comes down to statistics and probability. If a biological test was performed on you to see if you had a DNA sequence that was linked to heart disease, would you know what to do with that information?

First off, you may be thinking you don't want the test if it's going to hurt, but don't worry, collecting DNA for these tests is easy and painless. Your body cells can be obtained by a small blood sample such as a finger prick, or by swabbing the inside of your cheek with a Q-tip, or even by having you rinse your mouth and spit into a small tube. Your DNA can be obtained from all of these simple methods with no harm to you.

Ok, so now I have your DNA. I run the test on you and find that you have a DNA sequence that has been shown to be linked to heart disease. What do you do? Well, it is important that you understand probability and the fact that this DNA sequence does not guarantee that you will have heart disease. You should find out how strong the statistical correlation is between the DNA marker and the likelihood you'll get heart disease. The correlation may be strong, hence you are likely to suffer the illness, or could be weak, in which case you are at less risk.

You may say that heart disease is such a life threatening illness that you will make a lifestyle change no matter what the odds may be. Maybe you had relatives with heart disease giving you more reason to believe your chances of having the disease are high. You can vow to change your diet to have less fat and cholesterol, you can exercise more, your doctor may prescribe cholesterol lowering drugs, etc.

Ok, great. Eating better and exercising more is fantastic and strongly recommended. The prescription of drugs can be benefitial, but what if you're taking medications for an event that won't happen? Let's change the scenario. What if the test wasn't for heart disease, but rather for restless leg syndrome or pattern baldness? Diet and exercise are unlikely to help much (although I would still recommend them anyways). Are the risks of getting these ailments of such great concern to you that you might start taking medications to prevent restless leg syndrome or baldness? Remember, the test showed you were more likely than not to have them, but not a guarantee you will. What if the medications for these have side effects? How do you balance the side effects with the fact that you may not need these medications at all if you aren't going to have the problem?

What if the test was for alzheimer's disease for which their are no well documented ways of preventing it or curing it? How would knowing that it may be in your future effect you? These types of questions are being addressed now. In recent cases of a genetic test for breast cancer (BRCA genes), women who have mutations in these genes are more likely to develop breast and ovarian cancer than women who don't have the mutations. There is a test available for this and some women who test positive for the mutation have opted to have both of their breasts and their ovaries removed to try to prevent the cancer from occuring.

This issue is likely to quickly become even larger not only because more and more DNA markers are being linked to common diseases, but your DNA information may become readily accessible to you as a consumer. Several companies have now been formed where you can send your DNA to them and they will test you for not just one or a few DNA sequences, but rather 500,000 or more at once. Imagine that you can send in your cheek swab or tube of saliva to a company via the mail, and within a few weeks, receive your DNA profile on what markers you have that are linked to heart disease, breast cancer, restless leg syndrome, psoriasis, Crohn's disease, macular degeneration, Alzheimer's disease, attention deficit disorder, schizophrenia, major depression, arthritis, diabetes, manic-depression (bipolar disorder), etc, etc..

How do you process all this information and who will explain it to you? How will you feel when you log onto the internet to retreive and search through your genetic profile and find targeted advertisements for diabetes medications and hair loss treatments matching your genetic risk profile page? Will you be asked to join social networking sites and blog groups for people who share your disease risk markers? Not only should you consider how you will respond to it, but consider how others will respond to it once it becomes available. Your spouse, family, employer, insurance company?

We're far from the movie Gattaca where people were socially and economically stratified at birth based on their genetic predispositions, but we are moving towards a time when the general public will soon be just like the scientists are today, i.e. they will have the text of their genetic book long before they have the ability to understand what it means. My advice to those out there ready to mail in their saliva and see what their DNA blueprint shows......make sure you take a statistics class on probability theory!


Steve Murphy MD said...

Great post. I loved your examples.

George Melcher said...

Hi Tim!

I'm just curious if personalized medicines are now available in the market or in the hospitals? You know, having these kinds of more innovative products can make a big difference in treating certain diseases. I also wonder if companies with these kind of medicine uses FT NIR analyzer? But I think they are using tools like this for accurate measurements of medicines that they are about to sell.

Well, thanks for sharing this post!

Tim Grammer said...

Thanks for the comment, George.
There are some diagnostic products on the market that match a patients disease condition with their likelihood of successful drug treatment. Probably the most well known of these are the BRCA tests I mention and the HER test for breast cancer, which is used to determine likelihood of successful response to HER-targeting drugs such as Avastin. However, the field is still in its infancy.

Follow me on Twitter!

    follow me on Twitter