One example of the dilemmas posed by the rise of personalized medicine is Alzheimer's Disease: the ApoE genetic test can predict early onset of it, but there's still no effective therapy to stop dementia.
Payers, patients and physicians will all be facing these questions more and more, says Joanne Armstrong, MD, senior medical director at Aetna, who leads the firm's women's health and genetic programs. The market for genetics, genomics and molecular diagnostics is expected to reach about $6 billion this year, growing at a pace 20 percent to 50 percent annually.
In 1998, Aetna developed a program for the BRCA mutations associated with inherited breast and ovarian cancer, and subsequently was one of the first insurers to create policies for covering genetic medicine and for the use of genetic information. Armstrong, who is also on the board of the Personalized Medicine Coalition, talks about how she's seen the field grow, how Aetna covers genetic diagnostics and treatments, and the information gap.
HPN: The completion of the Human Genome Project in the early 2000s seemed to be one important foundation for personalized-medicine, but you've watched genetics develop in medicine for a while?
Armstrong: It started way before the 1990s. The Human Genome Project sort of engaged the country, but a lot of genetic technologies preceded that by decades in many cases. Looking at chromosomes in pregnancy to diagnose fetal chromosomal problems was around for 20 years or so before that.
Our knowledge of genetics didn't start in the nineties, but it has accelerated very dramatically. When you look at the Gene Test, a project of the University of Washington, they publish a few times a year an update on the number of diseases for which genetic tests are known to be available, and when you look at the curve, now it's even rising exponentially. In terms of the number of genetic tests that are out there, it's just phenomenal.
When we look at the use of the tests, a big challenge is: Of the tests that are available, how many are really both clinically valid and clinically useful? We see a big increase in the spending on genetic tests, and it's still relatively modest. For Aetna it's less than one percent of total medical spending. That doesn't seem dramatic, but when you look at the trend, it rises by about 10 to 12 percent per year, so the amount of spending on genetic tests outpaces inflation.
The other thing we see in the market is the rise of the availability of drugs that are informed by a genetic test. That's exciting, because it changes how we practice -- from an older paradigm that says everybody with hypertension gets this drug, everybody with depression gets this drug, everybody with breast cancer gets this drug -- to one that says, we know something about the genetics of this disease and this drug only works if you have a certain biomarker or a certain genetic feature of this disease. And now if you look at the number of drugs that are on the market that have what are called companion diagnostics, that list has grown really a lot in just about three years or so.
HPN: How are you approaching coverage of genetic and personalized medical technologies?
Armstrong: In terms of the framework we use to think about all this new genetic information, we believe we're prepared for it; it's the same framework we use to evaluate everything. The principles of coverage policy: services need to be related to the prevention, diagnosis or treatment of an illness. Getting information that's curious to have but is not related to prevention, diagnosis or treatment of an illness is not covered under a medical plan.
The information that you get from those tests need to affect the course of treatment, the care or the treatment needs to be likely to improve the outcome and should be attainable outside of investigational settings. Those are the general principles.
So something like genetic screening for Alzheimer's Disease is done through a gene called ApoE. You keep running this test on DNA sequences and if in fact you find it's same sequence again and again, it's analytically valid. It's clinically valid, in that we know that this gene is associated with early onset of Alzheimer's Disease, and if you have this gene it does predict early onset. It has no clinical utility today, in that there is no treatment for early Alzheimer's Disease, and so that ApoE is not covered under the benefit plan. As soon as there is a drug that is effective in individuals who have this genetic marker, the test will certainly be covered. It's just right now, what you do with the test is still in a research environment. Now there are many, many, many things that are in the pathway of analytical validity, clinical validity, clinical utility and those things are covered.
The other thing we've done -- in some ways it's traditional health plan practice -- is we've looked at selected drugs and selected therapeutics to say, Is there an opportunity for Aetna to put any of these technologies on a pre-certification list? For the drug Erbitux for colorectal cancer, the likelihood of its success is very strongly related to whether the tumor has a marker called KRAS. To ensure that KRAS testing is done, we put the drug on the precertification list, where we basically say, 'If you want this drug, we need confirmation that a KRAS test was done. In the case of interferon treatment for Hepatitis C, the duration of treatment is a function of what the genetic marker for what that virus is. For Aetna, we actually ask: Was genetic testing of the virus done, what is the genotype of that virus?
HPN: Considering the growth of genetic medicine and molecular-based diagnostics in general, how do you think physicians, patients and payers should approach their new availability, which could become overwhelming?
Armstrong: This is a huge challenge for medical educators, not just physicians, but nurse practitioners, nurses and the general public as well. There's plenty of literature that says practicing physicians rate their own knowledge of genetics as inadequate. Consumer-patient surveys show the same thing. And the challenge is when you look at the number of technologies that are coming through the pipeline, it's coming so fast that it is very difficult for anyone to understand how to use all these tests. There is an increasing need for support, and that support traditionally is in the form of genetic counselors. I actually believe that much of this knowledge is actually going to overwhelm genetic counselors, especially if they're sort of generalists. And so what's now being developed -- and we're doing some of this work as well -- is looking at ways to help provide decision-support tools to practicing physicians and patients about what these tests mean, what they don't mean and perhaps where they're used in clinical pathways.
People who have expression of disease that maybe is due to a genetic cause, that is not recognized to the family or the physician: that may be the case of patterns of breast cancer, ovarian cancer and various patterns of colorectal cancer -- multiple generations have early onset of disease. So we at Aetna have something called care considerations, where we mine our claims database and we look for markers of disease in our members that have the possibility of having a genetic underpinning. In the case of inherited susceptibility to breast and ovarian cancer, women who have ovarian cancer it's a marker of potentially BRCA, a gene mutation; women who have early breast cancer before age 45, that's a marker of perhaps a genetic basis of breast cancer. Those women we actually message to, to say, Talk to your physician or perhaps a genetic counselor about the possibility of a genetic underpinning of that disease. It's something to sort of trigger a discussion.
We also developed a network of telephone-based genetic counselors to address the information gap that's happening in the country. We did some early research with an academic partner to understand what are the determinants of whether a patient actually sees a genetic counselor or not. If much of this new genetic information is going to overload primary care physicians, are genetic counselors are going to fill that need and are patients actually getting to them? What we found is it's the physical distance between the physician and the genetic counselor that determines whether a patient actually gets counseling. Because of that we developed this telephonic model; initially it was genetic cancer counseling and now we've expanded it to all areas of genetics.
I think there is a very important health system question that we all need to engage in more deeply. We've got in this country a crisis in the cost of medical care. So every technology that comes on the market needs to have these questions asked of it: What is the value to patients, what is the impact of this on the overall health system? As we bring new drugs, new diagnostics on the market -- some of them with eye-popping price tags: $100,000, $300,000 a drug treatment course; some of the diagnostics are coming on the market at $5,000 a test -- we have to ask ourselves: Are we getting measurable clinical improvement? What is this doing to the cost of healthcare? And how do we balance those two things?
That's one kind of thing that does kind of keep me up at night. The other thing, in terms of technologies that are really exciting, I think the next thing will be next generation sequencing, providing large amounts of data on individuals. That offers the possibility -- not the reality, but the possibility -- that you'll get a lot of information upfront, that in some cases you can shorten the amount of time it takes to get from the question about the underlying cause of disease and the possibility of treatment. It has lots of applications potentially in cancer, where many people go through first-line treatment, second-line treatment, third-line treatment, although there's a lot of technology problems to work out there.
And then of course when you think about what it means to provide all of this information to a patient or a physician, you get back to this question of what is preparedness of the workforce. What would happen to a primary care physician for whom a patient walks in with three million data points? Clearly there needs to be specialized physicians and centers that are interpreting that information, making it actionable and this is the challenge of big data. Those things will be here in the next, I don't know how many number of years. Five? Ten?