Despite science’s ability to develop sophisticated and targeted new drugs, predicting the effect of a drug in an individual is still maddeningly difficult. Not every drug works for everyone that takes it. Similarly, the very same drug can be well tolerated in some, but can cause intolerable side effects in others. So-called “targeted therapies” were supposed to improve our accuracy, by focusing on specific targets on cells. That’s been good – but not sufficient to make drug treatments more consistently effective. Pharmacogenomics is the relationship between your DNA and how your body responds to drugs: how they’re absorbed, how they work, and how they’re eliminated from the body. It has been heralded for some time as the white knight of drug therapy. The genome revolution was supposed to remove (or dramatically reduce) the uncertainty in medicine, telling us which drugs will work more effectively, and which we might want to avoid. And to some extent, the genome-based treatment era is already here. There are over 100 drugs approved by the Food and Drug Administration (FDA) now that include genomic information in their prescribing information. For a small number of drugs, genomic testing is warranted. Increasingly, genomic testing is more accessible, moving from the research bench directly into retail pharmacies for sale when you pick up your prescription. Given pharmacies have a less-than-stellar record of selling laboratory testing that isn’t validated or even useful, I was immediately skeptical when I saw a new story on pharmacy-based genomic testing. Titled “Your pharmacist’s secret weapon: How your DNA can help perfect your medication,” it appeared a recent Globe and Mail column:
Catherine McLeod suffered one of her worst drug reactions – a minor stroke – while behind the wheel. She drove through a red light, and an oncoming vehicle totalled her car.
No one was seriously injured. But after the accident in 2005, she experienced two more minor strokes and kidney damage because of the litany of medications she was taking for bipolar disorder, diabetes and other conditions, she said. “I was overmedicated.” McLeod, now 62, is working with a different psychiatrist to adjust her medications, which add up to at least 10 pills a day. In the past year, she has endured side effects ranging from blood-pressure fluctuations to Parkinson’s-like symptoms. Finding the right medications in dosages she can tolerate is “still hit-and-miss,” she said.
McLeod is hoping to find answers in her DNA. A team of researchers at the University of British Columbia are testing patients’ saliva for genetic variations that affect how individuals respond to medications. She said she “jumped at the chance” to participate. The study is the largest in North America to look at whether community pharmacists can gain the skills required to collect DNA samples from patients, walk them through the consent process and explain how specific drugs may interact with the patient’s genetic makeup. So far, 200 patients in 33 pharmacies throughout British Columbia have participated in the study. The next phase will involve 1,000 patients in pharmacies across Canada this year. The goal, researchers say, is to bring personalized medicine to a pharmacy near you.
The column goes on to describe the clinical trial, being conducted by the British Columbia Pharmacists Association, Genome British Columbia, and the University of British Columbia. Community (retail) pharmacies recruited patients for the study, where patients undergo genomic testing by spitting into a vial. After analysis, participants receive a report that looks at genomic markers that are though to predict drug effects. The author is understandably skeptical. The trial actually seemed more focused on demonstrating that pharmacies can actually collect and administer testing, and that the University could sequence it, rather than focus on providing any objectively useful data for research subjects. According to releases, there will be a retrospective analysis of how a patient’s drug effects may have been uniquely affected by variations in their DNA. But at this point, there’s no evidence (nor are the researchers claiming) that this study will improve the health of participants. The study is paving the way for what’s to come: the elimination of the physician as testing gatekeeper, with the pharmacist positioned to give consumers access to data on their own DNA. But will that information help us? Proponents say yes:
“One of the most immediate opportunities for genomics in health care is to guide treatment decisions and reduce the risk of adverse drug reactions. This project is tackling just that by enabling pharmacists the insights needed to match the right medication, at the right dose, to the right patient,” said Dr. Catalina Lopez-Correa, Vice President, Sectors and Chief Scientific Officer, Genome BC. “This work reflects Genome BC’s ambition to translate the value of genomics to end-users in BC and beyond.”
And maybe that will come with time. But there’s no evidence yet that demonstrates that routine genomic testing (unrelated to specific drugs) is improving patient care and outcomes.
A business case to sell testing is not the same as a medical reason to offer testing
Direct-to-consumer marketing and sales of DNA and genomic testing has been referred to as “controversial” for a reason. There is a significant disconnect between the ability to offer genomic testing and the usefulness of that information. 23andMe offers extensive, but problematic, access to your DNA – going beyond drug effects, to looking at ancestry markers, and even inherited disease risks. What one should do with this information remains unclear. Genomic testing offered by pharmacies is more restricted, and focuses on markers that affect medicines, but it’s no less controversial. While testing is increasingly being offered by pharmacies, it’s not clear that the information is useful, or actually improves health outcomes. A “normal” result with a gene variant doesn’t mean that a drug is effective, or will be free of side effects – it doesn’t tell the whole story. But does it make drug therapy more precise? The story of warfarin is cautionary, it was supposed to be the first drug for which genomic testing would be routine. Science and evidence found otherwise.
Genomic testing with Coumadin therapy offers no benefit over not testing
Warfarin (Coumadin) is an anticoagulant (“blood thinner”) used routinely to prevent blood clots (thus preventing strokes and heart attacks) in numerous medical conditions. Warfarin is one of the more challenging drugs to dose. It counteracts the action of vitamin K, meaning that the amount of vitamin K in your diet will influence its effectiveness. Too little warfarin in your blood, and the risk of blood clots goes up. Too much warfarin, and you are at higher risk of bleeding. Dose changes can be frequent and are guided by the “INR” test. Standardized dosing is usually used when you start therapy, with close monitoring that decreases over time as long your INR stabilizes. The effect of a dose of warfarin is affected by your genome which can be tested with saliva samples: The gene CYP2C9 affects how warfarin is deactivated in the body. The gene VKORC1 activates vitamin K in the body, which reduces the effect of warfarin. Testing for these genes can lead to a more-customized dosing plan, which was expected to result in improved dosing of warfarin, and in the long run, fewer side effects of treatment.
As is completely appropriate, this genome-informed approach to warfarin dosing was tested in over 1,000 new warfarin patients. What the researchers found was surprising: Genomic testing or not, there was no difference between the groups. Genomic testing didn’t improve the dosing accuracy of warfarin. Most of the variation in response was actually due to other factors. And the trial gave caution to the use of genomic information without completely understanding the implications of the test: African-American patients actually did worse with the genomic testing, although there was no difference in overall adverse effects by the end of the study. Genomic testing for warfarin gave information, but it wasn’t useful and it didn’t improve health outcomes. While that hasn’t stopped the research, it also hasn’t stopped the marketing.
There are population-level cautionary cases, too. A genomic screening program in Hong Kong to identify patients at risk of severe skin reactions with one epilepsy drug (carbamazepine) led to an increase in in the use of another epilepsy drug (phenytoin) which also could cause skin reactions which were not predicable with genomic testing. The result was no change in the overall incidence of drug-induced skin reactions. What’s clear is that it’s not just the test we need to evaluate – we need to understand decision-making that emerges from using that data.
Dubious origins and sketchy sales practices
Laboratory testing is an area of health care that’s been called the “Wild West” because of a mix of entrepreneurship, unclear standards, and sometimes blatant hucksterism that prioritizes performing (and selling) a test over providing a medical service that is actually useful. While pharmacogenomic testing may not provide meaningful health information, the testing is a business opportunity for hundreds (or perhaps thousands) of private laboratories. It’s a competitive marketplace, and in trying to establish preferential relationship with health professionals, questions of conflict of interest are being raised, along with concerns of fraud. Private and public insurers are skeptical of the benefits and the costs, and insurance coverage is not yet widespread. In the United States, the FDA only started to assert its regulatory role over lab-developed tests last year.
Are pharmacies promising more than they can deliver?
Chemmart claims “myDNA is a genetic test that identifies which medications will work best for you”. This overstates the role and value of this test. It has limited applicability only to certain drugs in particular situations.
We do not believe the test is “particularly relevant” to those who “take multiple medications, have children on prescribed medication”, or “are pregnant or planning pregnancy” because of the extremely limited applicability of the test to these patient groups. We also have problems with the claim that “the myDNA test covers 50% of the most commonly prescribed medications”. This is not in accord with data from the United States, which shows that just 7% of approved medications and 18% of outpatient prescriptions are affected by actionable pharmacogenes (genes you can test for and alter medication around). Nor is it in accord with Australian data on the top ten drugs prescribed.
We have submitted our concerns to the Therapeutic Goods Advertising Complaint Resolution Panel for an independent determination. More research (and GP training) will be required to determine if PGX tests improve patient care and are cost-effective. In the meantime, marketing claims should reflect the current uncertain clinical role of these tests.
It’s not just the test – it’s the interpretation of the results
Given the modest effects that individual DNA variations seem to have on the majority of drugs, understanding test results in the proper context is essential. Yet the science may not yet be clear enough to give definitive answers. Should you change your drug therapies if it’s working for you, but your test suggests they may not be idea? Probably not.As the Globe article above notes,
But adjusting medications based on genetic testing is not as easy as it sounds, according to Dr. Richard Kim, a clinical pharmacologist at the University of Western Ontario. If it were, he said, “every hospital in North America would be doing it.
What’s clear is that patients and health professionals alike need good evidence to guide decision-making: Not one manufacturer’s opinion, or one health professional’s anecdotal evidence. There are sites like PharmGkb that are emerging as unbiased clearing houses for information – which should, as more evidence emerges, help us sort out the science from the marketing.
There’s no question that genomic testing, along with most other forms of laboratory testing, will become more and more widely available to consumers. What are important are the claims made about the tests, and whether or not they can be substantiated with good evidence. While testing could emerge (in time) as an evidence-based guide that informs treatment decisions, improves care and decreases costs, it could also become the laboratory version of snake oil, promising benefits that can’t be backed up. Regulatory schemes to restrict access won’t work, or be accepted by consumers. But every consumer should welcome measures to hold manufacturers accountable for the claims that they make, and to ensure that testing is safe, accurate, consistent, and provides meaningful information.
Does genomic testing represent good value for the cost?
In comparison to what drug treatments can cost these days, genomic testing looks like an attractive investment, if it can actually do what we want: predict which treatments are effective, avoid harmful effects, and improve overall health outcomes. It should be very attractive for insurers, who don’t want to fund drugs and treatments that don’t work. But that argument is a theoretical one, as the benefits have not been shown. Whether there will be an economic case for routine genomic testing remains to be seen. At this point, there’s no clear economic value for testing. Perhaps that will change in the future – if and when testing is shown to offer truly useful information.
In the long-run, genomic testing is probably inevitable. Done once at birth, it will be available to inform decision-making for a lifetime. We will soon be moving past the “test or not” debate, to focus on just how informative this information actually is. As more evidence emerges, the relative benefits will become more clear.
Not yet a revolution in medicine
Genomic testing isn’t routinely offered by your physician for a reason: At this time, its usefulness is not clear. Until better evidence emerges to demonstrate that genomic testing actually helps improve health, buying it from your pharmacy may do more harm than good. Genomic testing looks promising. But until its effectiveness has been validated, it’s probably best to seek out clinical trials, where you can help build the evidence base. There may as yet be an important role for pharmacists in the community to offer this testing. They are well positioned to guide advice to consumers (and other health professionals) on therapy decisions. But testing and advice needs to be offered when it’s useful to patient care – and not just when it supports a pharmacy’s bottom line.
Comments open over at Science-Based Medicine.