Patricia F. Fitzpatrick Dimond Ph.D. Technical Editor of Clinical OMICs President of BioInsight Communications

Molecular diagnostics still have a long way to go before they can properly deliver on all their promises.

How many SNPs does it take to provide a definitive disease risk profile? Quite a few, apparently, as companies continue to pile them higher and deeper into genetic tests. Firms have bet that these tests will be widely adopted by physicians and the public to predict everything from risk of lung cancer among smokers, to prostate cancer, to Alzheimer’s disease, to baldness.

All told, about three dozen companies claim that they can provide genetic testing that predicts an individual’s risk of developing almost everything. “There is a bit of a wild wild west going on in terms of some of the DNA testing that’s out there,” said Francis S. Collins, M.D., Ph.D., the newly appointed NIH head, in an interview with CBS News in September 2008 with regard to a new offering from Smart Genetics purporting to predict susceptibility to Alzheimer’s disease (AD).

“Some of it is done by reputable companies, but there are some that are even unscrupulous who will offer you tests or DNA variations that, frankly, you’re not sure what they mean at all.” Smart Genetics stopped offering its controversial Alzheimer’s Mirror genetic test just eight months after introducing it, and the company subsequently shut its doors.

The Ones Doing the Testing

Major players include deCODE genetics, Navigenics, 23andMe, and Proactive Genomics. In January Proactive Genomics made available a $300 prostate cancer genetic test based on five SNPs, called Focus5. deCODE’s $500 PrCa test was announced less than a month later, for analyzying eight SNPs. Navigenics offers a personalized DNA profile, explained by a genetic counselor, for $2,500 but reduced its price several months ago to $999. 23andMe and deCODE also advertise their own versions of such profiles.

As recent as January, deCODEme, a division of deCODE genetics, launched two new services, one designed to detect genetic variations associated with cardiovascular diseases and a screen to detect genetic variations linked to the risk of developing various cancers. At $195 and $225, respectively, the new tests cost less than the company’s genome-wide screen, which for $985 assesses genetic risk for 34 diseases and traits ranging from diabetes to male-pattern baldness. deCODE president and founder, Kari Stefansson, M.D., said that the company wanted to give people an opportunity to buy tests that would address their specific needs.

On September 17, the company received a notice from Nasdaq stating that it was not in compliance with the Minimum Bid Price Rule. deCODE got a slight bump on September 20 after reporting that it and multinational collaborators had discovered four novel SNPs conferring increased risk of prostate cancer.

This marked the sixth set of prostate cancer risk factors the company has found. “Using our ability to put these SNPs in a population-wide context, we show that it is now possible to identify those who are at more than a 30% increased lifetime risk, independent of other standard risk factors such as age and family history,” Dr. Steffanson noted.

On September 30, however, deCODE decided to closed its Illinois facility, cutting 60 positions and estimates savings to be about $1.5 million.

The Real Story That SNPs Tell

If genetic testing companies can survive long enough, SNP testing may achieve some potential by helping to identify individuals who could benefit from further testing. Accurately screening for risk of developing prostate cancer, for example, could help avoid subjecting relatively low-risk individuals to more extensive, invasive, and expensive procedures.

SNP developers recognize that genetic tests require a lot of context to be clinically useful and that diseases that are not the result of single gene mutations result from a combination of environmental, lifestyle, and other genetic influences. Dr. Stefansson, in further commenting on his company’s prostate cancer risk assessment test, said, “By incorporating this new, personalized gauge of susceptibility into our arsenal for improving prevention and early diagnosis, we can more effectively and accurately identify those men who would benefit most from intensive screening.”

SNP-based test results provided in the context of other risk factors may also encourage individuals to change health-averse behavior. PHD Diagnostics recently introduced a test that identifies smokers and ex-smokers at the greatest risk of developing lung cancer. On its website the company recognizes that “the ability to clearly segment a group of otherwise similar people into different risk groups is one of the keys to creating a clinically useful test. While several individual SNPs have been linked scientifically to one of the many complex factors that may underpin the development of lung cancer, by themselves these SNPs do not provide a strong signal or differentiator among various people.”

PHD president Bob Walker remarked that smokers often suffer from optimistic bias, the belief that bad outcomes happen to other people, not them. Respiragene provides an easy-to-understand score and personalized information that doctors and patients can use to help individuals take the steps required to quit smoking and improve their overall health, he noted.

Tests Not Yet Ready for Prime Time

Would clinicians use SNP-based tests to screen patients and determine their need for further testing? Prostate cancer is illustrative of a fundamental conundrum; Eric A. Klein, M.D., chairman of the Glickman Urological and Kidney Institute, writing in the October issue of the Cleveland Clinic Journal of Medicine, referred to the “seemingly opposite conclusions” reached by two large studies evaluating PSA testing. Dr. Klein commented that “in spite of some recent studies, or perhaps because of them, we still are unsure about how best to screen for and prevent prostate cancer.”

And while physicians would no doubt welcome accurate, definitive tests from the SNP-testing world, Dr. Klein further commented that “some data suggests that we can use genetic tests to screen for prostate cancer, but the tests are not yet as good as we would like.”

In summary, Dr. Klein told GEN, “If we had a good genetic tool that added predictive value, we would use it. But it needs to be robust and add information to what is already used, including family history and other factors.”

A program initiated this month at Beth Israel Deaconess Hospital Medical Center seeks to familiarize pathologists in training with genetic tests marketed to consumers. It is intended to help doctors help patients derive true meaning from these tests. Mark Boguski, M.D., associate professor of pathology at Harvard Medical School and one of the program instructors, commented that the program seeks to prepare trainees “not because the technology is ready for prime time today, but because people are using it, and it’s clear it’s going to play a role in the future.”

Patricia F. Dimond, Ph.D., is a principal at BioInsight Consulting. Email: [email protected].

Previous articleGEN’s Cryptogram Challenge Solved
Next articleGates Foundation Pledges $115M to Medicines for Malaria Venture