Trained as a cardiologist, Dr. Eric Topol, director of the Scripps Translational Science Institute, hasn’t shied away from pushing his peers to do better — from being among the first physicians to champion the recall of Vioxx a decade ago, to his embrace of genetics and mobile technologies in his 2012 book, The Creative Destruction of Medicine. Throughout, Dr. Topol has advocated for patients as individuals and called for recruiting every technology to improve their odds in the numbers game of medicine.
GoodRx’s Thomas Goetz spoke with Dr. Topol about the current guidelines for prescribing statins and how doctors could be using technology more effectively to get the right drugs to people who need them.
– – –
GOETZ: Given all the challenges in healthcare today, how well do you think medications are being used?
TOPOL: Drugs now represent around $350 billion a year of spending just in the US, and at least a third of that is total waste. In fact, if we were more intelligent about how we use medications, we could probably lower that number considerably. The challenge is that we are not giving patients all the information they deserve. There should be a systematic rebooting of how we use medications, from the initial communication with the patient to getting them the right drug at the right dose.
GOETZ: What does that reboot look like?
TOPOL: To start, at some point in the future, we won’t talk about “prescriptions” because that term doesn’t designate a real partnership. Once there’s engagement from the patient, using a drug might be helpful and a good trade-off with respect to side effects (which is pervasive with all drugs). Then, the next question will be: what about DNA-specific drug interactions? Of course, today, we hardly do anything with that. A lot of information is already well‑established, but not being implemented in daily care when prescriptions are written or filled.
GOETZ: Right. Of course, the FDA already publishes a list of drugs with known pharmacogenomic effects. That’s a start, no?
TOPOL: Yes, there are 130 or 140 drugs on that FDA list — but the information is hardly used. Physicians will rarely recommend that a genotype be done for a drug. And that’s only 100 plus drugs. There are thousands of drugs out there, plus the ones that are still in development. So we’re not using the information we do have well, and we have a truly inadequate amount of information on other drugs.
GOETZ: Let’s talk about the current statin guidelines. The intention with these was to make sure that effective drugs get those who need them. They don’t take genetic information into account, but an effort was still made to better identify patients who might benefit from statins. What’s your impression of how effective these new guidelines are?
TOPOL: They’re highly controversial. Almost 40 million Americans — one in four adults over age 45 — are prescribed statins. The new risk calculator could double that to 80 million. It’s basically the “statinization” of our population. This, to me, is crazy. The benefits are tiny and the risks are not trivial. The guidelines don’t even factor in family history, which is at least a crude genetic assessment. As a cardiologist, one of the most important things to know is whether a patient has a family history of heart disease at a young age.
GOETZ: A lot of people talk about how statins are so effective that we should just put them in the water.
GOETZ: But your point is that they are effective for just some people.
TOPOL: They’re effective in almost everyone for lowering LDL cholesterol. But in terms of actual prevention in people who haven’t had heart disease, statins are only effective a tiny percent of the time — two, three, four percent. Which means we’d be giving statins to some 95 out of 100 of individuals just to achieve better lab results. It’s a lab benefit versus a real medical benefit. That’s the crux of this issue: how do we increase the real medical benefit?
GOETZ: So how would we improve the delivery of statins, but not in a way that simply spreads them far and wide?
TOPOL: We would find out who really needs it, instead of relying on this very crude risk calculator that just uses age and blood pressure. Why don’t we have much more granular information on who truly does benefit at the genomic level — at the sensor level, at the imaging level, at every level? For instance, statins have a well-known side effect of muscle inflammation. Occurrences are pretty darn easy to predict if you look for a specific variant of the gene, SCLO1B1, that codes for the protein that transports statins in the liver. People who carry two copies of this variant have a roughly 30-fold risk of developing serious muscle inflammation. Even people with one copy have a significant risk. As if we’re idiots, we don’t use this information. We just give out prescriptions, and then, 10% of patients have serious muscle aches and have to stop taking their medications. We could essentially prevent this from the get go with a simple genotype that costs pennies to perform.
GOETZ: How do we start changing that? Is it a matter of generating demand among patients or educating physicians?
TOPOL: That’s a start, but it’s a slow go there. We envision activating pharmacists and drug stores with smartphone or handheld device technology, which can perform genotypes for drugs (like statins) quickly and inexpensively. Right now, we’re still sending it off to central laboratories and waiting a week. The status quo is very expensive and isn’t working.
GOETZ: And what about conveying this information to patients?
TOPOL: Since the average length of a physician-patient visit is around 10 minutes, there isn’t a lot of time to review a drug’s effectiveness and potential side effects. I’ve usually done it in a very truncated way. Sometimes it’s done more thoroughly by pharmacists or with printed handouts. I spend a fair amount of time with patients explaining why we’re trying a drug and what to expect, but we could make that into a video on a mobile device. That would be great.
GOETZ: And what about once the patient is taking the drug? How can we better monitor potential adverse events?
TOPOL: Even if we study adverse events in clinical trials, they are very contrived experiments. When you go to the real world, you don’t have the opportunity to cherry pick patients, and drugs can be dangerous when they go to scale in unstudied individuals. That’s when you need to have electronic capture of every individual who gets a new medicine, before it becomes a monster. But we don’t capture that data today. We don’t even really prepare for something going south. That’s eminently doable with electronic health records.
GOETZ: So it sounds like we need to embrace that life is complicated and try to bring in some logic and tools to improve it. Because, in reality, how people take drugs and how they respond to them can get very messy. How do you see us bridging clinical measurement with reality?
TOPOL: One thing that I don’t think most people realize is that finding genomic interactions with drugs is not hard. As human beings evolved over millions of years, it’s only been in the last century that there’s been much exposure to these drugs, so we haven’t had any chance to adapt them. This is a fantastic opportunity to really jump forward. Just look at how we figured out that genotyping can perfectly predict responses to lithium in patients with bipolar disorder. The other thing we need to consider is the different approaches we’ve touched on — implementing electronic surveillance when drugs are given or tapping into tools we have on hand to educate patients at the time of their prescription. We should also look into better engaging allied health professionals — certainly pharmacists and people who work in drug stores — in this whole process. There are opportunities galore to take this to new heights, and I look forward to that. It’s likely to happen over time.
We've sent a link to download the GoodRx mobile app to your phone.
We were unable to send a link to your phone.