Blog | Wednesday, April 24, 2013

Personal genomics--maybe an excellent idea?


My son, who is 19, wanted a nice jacket and his genome for Christmas this year. He had found out that a company called 23andMe would tell him about his genetic susceptibility for diseases and his ancestry for only $99. 23andMe is a private company that started doing direct to consumer genetic testing in 2007 and has progressively lowered their prices and increased the amount and sophistication of information they give clients since then.

My sister bought my son the genetic kit and the padded envelope was under the tree for him on December 25th. He opened up the little plastic tube inside, spit in it and sent it back. In a couple of months he had all sorts of interesting information. He was most interested to see what parts of the world his ancestors sprang from, but also got information about scads of other things including his sensitivity to caffeine, ability to recognize bitter tastes, whether his muscles were predominantly "fast twitch" suggesting he would be a good sprinter, whether he would be sensitive to certain medications for which we have that genetic test, and whether he carried genes for some common conditions such as cystic fibrosis or hemochromatosis. With each piece of information there was a link to information about how good the science was that backed up the association, what the disease was and how common it was in the population.

23andMe got its name from the fact that there are 23 pairs of chromosomes in the human genome. Evaluating genes from a spit sample in this way was named invention of the year in 2008. It uses microarray analysis to look at single nucleotide polymorphisms (SNPs) via the company Illumina. (Illumina also provides its own service of full genome sequencing which is far more expensive than the $99 test that 23andMe performs.) After submitting the sample, the client can fill out tons of surveys and questionnaires which the company can use to make even more predictions about characteristics based on SNPs.

It is possible to detect increased or decreased risk of certain diseases such as Alzheimer's disease, melanoma and breast cancer. There are certain SNPs that are associated with an increased risk of heart attacks when one drinks coffee. There are SNPs associated with higher risk of an adverse reaction to warfarin, statins or certain drugs for HIV or hepatitis C.

When I, as a primary care physician or hospital physician, see a patient, it would be incredibly useful to know what the person's individual propensities are. Is this the patient who should be watched closely after being started on a certain drug, or perhaps not started on it at all? Should further testing be done for cystic fibrosis, or is this very unlikely? Is this one of those people I should tell to drink coffee, because it is good for them, or are their needs different? I spend many hundreds of dollars of the patient's and their insurance company's money finding answers to these questions.

Hemochromatosis is a genetic disease in which iron builds up in the tissues, causing arthritis, fatigue, diabetes, heart failure, skin changes, liver cirrhosis and cancer. It is pretty common, about 1/200 Caucasians have it, and it usually doesn't present until the fourth decade, often with some life threatening disease. By reducing iron intake and therapeutic removal of blood, iron stores can be kept at safe levels, completely preventing actual disease, but none of the most standard blood tests detect it when it is still early. A 23andMe test detects the most common genetic abnormalities that cause it. Using the tests that I order to detect it would cost over $600, the last time I looked into it.

Lately there have been many articles looking at the health benefits of drinking coffee and eating dark chocolate. We have a large database of nurses who drank astounding amounts of coffee and also participated in a longitudinal study of health outcomes. We don't know anything about these peoples' genetics, but we do know that, in general, the moderate to heavy coffee drinkers did better on many measures. Dark chocolate appears to have positive effects on blood pressure and other cardiovascular outcomes. One of the tests in the 23andMe profile is fast vs. slow caffeine metabolism. It turns out that fast caffeine metabolizers may have reduced risk of heart attack when they drink coffee and slow ones have an increased risk. There are many reasons why a person might drink coffee or eat dark chocolate other than improving their health, but knowing about my caffeine metabolism would definitely have some effect on my consumption.

For a patient, there are so many terrible diseases out there vying for one's attention. Which should we focus on? The site has some very good generic advise for patients on what they might do to reduce risk for disease, since only a portion of that risk is genetic.

There have been many articles discussing the pros and cons of personal genomics. It is possible that detecting an increased risk of an expensive disease could affect insurance costs or willingness of a company to hire a person. Insurance discrimination based on this information is definitely illegal, and 23andMe information is kept secure, but anything anywhere on the internet is potentially common knowledge. Some people have been concerned that since the information is given directly to the consumer rather than involving a physician to interpret the results, misunderstandings could occur with unknown consequences. Genomic information is used as part of medical evaluations, including full genome sequencing for rare and mysterious diseases and characterization of specific mutations in malignant tumors to find out which of the many very expensive cancer drugs will work for them. This information definitely requires a doctor to order and interpret it correctly.

The information a person gets from the 23andMe site is useful at some level for an educated consumer, but would be even more useful if shared with a physician. That process would take time and it might be challenging to get a primary care doctor to engage in it. If physicians became familiar with the format, it would be possible to review the information at a designated appointment for that purpose, as doctor and patient looked at the results together on a computer. Because genetic data is difficult to interpret and explain, physicians often refer patients to genetic counselors, but these visits are usually expensive and time consuming and reserved for people with definite medical risks or recognized diseases.

Why, one might wonder, does it cost only $99 to get all of this awesomely useful information? Ordering any little part of this information, without the website support or genealogy stuff, costs far more than if I, as a physician, order it for my patients. The low cost is at least partly because this venture bypassed the usual regulatory steps and isn't paid for by medical insurance. It is also not yet profitable, and was initially bankrolled by Google cofounder Sergey Brin, Genentech and some other investors. Because it gets more powerful information and becomes more efficient as it gets more customers, it is becoming progressively more profitable at the same time it becomes more affordable. There are other companies that do personal genomics, but I have been unable to find any that use this model and have anywhere near the level of success that this one does, so I don't think competition is doing much to bring costs down.

Today I got online and pressed the "buy now" button for a 23andMe kit. Soon my padded envelope will arrive and I will spit in a tube and wait for my results. Ninety nine dollars is a very small price to pay for a chance to know what my particular medical issues might be. So much of the evidence in evidence based medicine is based on the idea that the responses of a single person will be the same as those of an unselected population. Population based research leads us to believe that exercise, olive oil, coffee, dark chocolate, red wine, statins and fish are good for everyone, but are they, in fact, good for me? 23andme can't answer most of those questions yet, but if enough people are analyzed there eventually will be a better understanding of individual patient's needs.

As I re-read this post, after getting the initial comment (see below), I realize that it reads like an advertisement for this particular company. I do not intend that, but I really can't find other services that are comparable. There are lots of companies that are rolling out genetic evaluation products and I expect that in a few years there will be many viable competitors. Using genetic testing to individualize treatment is all over the doctor news this week. Eric Topol, a cardiologist and geneticist, spoke about it here. The practice of using genomics is moving very fast and it is impossible to know what impact it will have. It has the potential to increase health costs astronomically, or improve and streamline care. I am excited to see what a direct-to-consumer model can do, unconstrained by many of the perverse incentives that make medical progress slow and expensive.

Janice Boughton, MD, ACP Member, practiced in the Seattle area for four years and in rural Idaho for 17 years before deciding to take a few years off to see more places, learn more about medicine and increase her knowledge base and perspective by practicing hospital and primary care medicine as a locum tenens physician. She lives in Idaho when not traveling. Disturbed by various aspects of the practice of medicine that make no sense and concerned about the cost of providing health care to every American, she blogs at Why is American Health care so expensive?, where this post originally appeared.