Over the past eight years I’ve written more than 100 stories criticizing a drug called ezetimibe, which is sold by itself under the brand name Zetia and in combination with an older drug under the brand Vytorin.
When I started off, in 2006, it was just “the Vytorin question:” did these drugs prevent heart attacks as well as older medicines like Lipitor? Their makers, Merck and Schering-Plough, had delayed analyzing a study that hinted the answer was no. My criticisms of the companies’ marketing became more and more heated as sales of the drugs dropped and top doctors disparaged the drugs. There was a “backlash,” a “cancer question” an “ethical morass” and a “secret panel.” I suggested that the CEO of Schering should give back his bonus. The controversy hurt the stock prices of both companies, and helped lead to Merck’s $41 billion purchase of Schering.
Then, last November, came the results of an 11-year study funded by Merck. In at least some heart patients – those who have had a heart attack and have chest pain – adding ezetimibe to an older cholesterol drug, Merck’s Zocor, reduced the risk of heart attacks and strokes. It’s a tiny benefit: with ezetimibe, 33 patients out of 100 would have a heart attack or stroke; without, there would be 35, a difference of one heart attack and one stroke for every hundred patients over the course of seven years. Though small, that difference is likely real. Which means I need to ask myself: was I wrong this entire time?
I think the answer to that question goes deep, straight to the public’s damaged relationship with the pharmaceutical industry and to a reality that impacts everything from anti-vaccine ideas that are endangering kids to our distrust for medicine in general. Here’s a hard reality to face: the drugs don’t care whether the people who invented them were saints or sinners. It’s true that drug companies frequently try to mislead us when it comes to the medicines they make, but our anger at Big Pharma can mislead us, too. And as for ezetimibe: recent data on the drug and on its biological workings make me wonder if I should consider taking it myself.
"This starts smelling..."
Merck and Schering behaved terribly when developing ezetimibe: burying data, heavily marketing the drug with television ads, and paying some doctors to talk it up. From the start, the companies dragged their heels on starting a study to test whether or not adding ezetimibe to a statin prevented heart attacks and strokes. Such a study didn’t begin until 2005, a year after the drug was approved.
Critics saw the whole ezetimibe franchise – both Zetia and Vytorin – as a cheap trick to extend the life of Merck’s $7-billion-a-year Zocor, which would lose patent protection in 2006. The ezetimibe pills would have another decade of patent protection. When Vytorin, an ezetimibe-Zocor combination, was approved in 2005, Steven Nissen of the Cleveland called it “a step backward.” He added: “There is not enough evidence to justify putting millions of Americans on a statin-sparing medicine. Ezetimibe and Vytorin are just marketing tools, a way to evergreen the patent on Zocor."
What came next shocked the world of heart drugs. In 2002, Merck and Schering had started a study looking whether adding ezetimibe to Zocor would better prevent arteries from thickening compared to Zocor alone. They had hired a top-flight academic, a dashing ultrasound expert named John J. Kastelein from the University of Amsterdam, to conduct the study on a group of patients with a rare genetic disease that causes super-high cholesterol levels.
Yet Kastelein’s contract didn’t give him the right to analyze his own data without Schering’s permission. When the study completed in 2006, Schering-Plough would not let him finish it.
“Is it correct that [Schering-Plough] has decided not to present at [the American Heart Association], but to await the two other, completely unvalidated, endpoints, which analysis is going to take us straight into 2008??!!??,” Kastelein wrote in an email unearthed by Congressional investigators. “If this is true, SP must have taken this decision without even the semblance of decency to consult me as [primary investigator] of the study. I can tell you that if this is the case, our collaboration is over. This starts smelling like extending the publication for no other [than] political reasons and I cannot live with that.”
When a Merck executive, Soren Bo Christensen, protested the lack of communication with Kastelein, Enrico Veltri, the head of cardiovascular research at Schering, replied: “Soren is a prick. How’s that for staying calm. Tell him to f— off.”
It was only after journalists – first me, followed by Alex Berenson at the New York Times – wrote about the delay and Congressional investigations began that the data were finally analyzed. Other imaging experts said that, at the end of the day, there were no problems with Kastelein’s data. But the conclusion of the study was clear: adding ezetimibe to Zocor did nothing. Merck and Schering’s blockbuster drugs might be nothing more than placebos.
Backlash
The backlash was swift and long-lasting. Shares of Schering dropped 26% the week the imaging data were presented at a big cardiology meeting. Merck shares fell 15%. Over the next five years, the number of prescriptions written for Zetia and Vytorin would fall more than 50%, with sales figures buoyed by price increases on the pills.
I found Merck and Schering’s response to the controversy infuriating. As one top cardiologist after another advised that patients shouldn’t get ezetimibe unless they couldn’t take a higher dose of a statin because of muscle aches or other side effects, the companies just seemed to keep insisting their own data was meaningless. I remember Merck PR people pushing me hard to interview Antonio Gotto, the dean of Cornell-Weill Medical School, whose views on ezetimibe seemed far more positive than anyone else I spoke to. According to investigators for Senator Charles Grassley, Merck paid him $27,146 for lectures and consulting during that five-month period.
But once this negative narrative was set, there was a tendency for ezetimibe to look even worse than it was. It became the poster-drug for not gathering enough information about a medicine before marketing it. When more patients taking Vytorin got cancer than those taking placebo, it was easy to find critics who were worried about the connection. Scientists working with Merck misplayed their hand by trying to insist that there was absolutely no risk, creating a scary-sounding debate about whether the cancer link was unlikely or completely impossible.
“The questions were legitimate, but I think some of the grandstanding that happened was probably not ideal,” says Robert Califf, who helped run the study that proved ezetimibe does prevent heart attacks. (Califf is now an official at the FDA, but spoke to me last fall.) Adds Harvard cardiologist Eugene Braunwald, who co-ran that study: “I think that it was unhelpful. There was too much noise made about studies that were inadequately sized even though one of them was published in the New England Journal of Medicine.”
The news coverage, Braunwald says, made patients afraid to stay in clinical trials that would actually provide answers about Zetia and Vytorin. But were patients otherwise harmed by aggressive news coverage?
“You did what you had to do based on what you knew at that time,” says Fred Hassan, who was Schering-Plough’s chief executive during the Vytorin controversy. “I don't think you guys should feel bad about it. I think it's unfortunate that there was so much misunderstanding that occurred at that time and also some emotion, but I think that happens with many products.”
It sure does – often, as with Schering, because a drug company did something that, to people outside the industry, seems nasty and inexplicable. In 2003, new data emerged that Paxil, Prozac, and other related antidepressants might actually increase the risk of suicide in depressed teenagers when they started taking the pills. In 2004, the New York Attorney General brought a lawsuit against
Bad drugmakers, good drugs
Not being clear about the risks of medicines given to children is a terrible thing for a company to do. But what happened as a result of the antidepressant controversy? A study published in the British Medical Journal last May found that the warnings in 2003 decreased adolescent use of antidepressants by 31%. At the same time, the number of deliberate overdoses on psychotropic drugs, which correlate with suicide attempts, increased 22%. Did the warnings scare kids away from antidepressants that would have helped them?
Or take the case of Tamiflu.
And the worst example? Distrust of pharmaceutical companies has been given as a given as a reason not to vaccinate children – even though vaccines are among the most studied medical products. That leads to outbreaks, like the measles outbreak that hit Disneyland earlier this year.
Big bad drug companies make it easy to fall into a kind of conspiracy-thinking mentality. If Roche and Glaxo are going to behave like lying, data-burying scoundrels, their drugs must be useless, right? But, because medicines have to prove themselves in both a regulatory gauntlet and the market, that’s not always so. And we can be misled by that kind of thinking. I’ve seen it with Vytorin.
BusinessWeek used Schering’s failed ultrasound study as a hook for an award-winning cover story titled “Do Cholesterol Drugs Do Any Good?” It argued “many researchers harbor doubts about the need to drive down cholesterol levels in the first place.” Aside from the ezetimibe results, the article hinged its argument on how low the rate of heart attacks was in some statin trials. For instance, one Lipitor study decreased rate of heart attacks over a 3-1/3-year period from 3% to 2%. That is a difference of only one heart attack for every hundred people treated for three years – in medical parlance, “a number needed to treat” or NNT of 100. Why, the article asked, would anybody take a medicine with such a miniscule benefit?
Lowering cholesterol saves lives
In 1995 a study funded by Merck proved that statins (specifically Zocor) reduce the risk of death by a third in patients who’d already had heart attacks. The BusinessWeek story meant to question where exactly to draw the line on cholesterol treatment for people who have yet to have a heart attack, called primary prevention. How low should we push cholesterol levels in people who are not yet sick?
But remember that heart disease is the leading killer in the U.S., and many heart attacks and strokes occur in patients at relatively low risk. A 2014 analysis by the independent Cochrane Collaboration found that in studies of primary prevention, statins reduced the risk of death by any cause by 14%, cut the risk of a heart attack, stroke, or cardiovascular death by 25%, and that they are cost-effective and improve patients’ quality of life.
Thinking about the benefit of cholesterol drugs in terms of what can be shown in a three- or five- or even seven-year study is missing the point. These experiments are the best we can do to be sure that these medicines work as advertised. But the real benefits of cholesterol medicines are likely to accrue over many years, because cholesterol builds up in arteries like interest on a loan. It’s only after years, or decades, that you see the difference.
This fact is better understood now than it was when Vytorin was approved because we now have studies that show the benefits of living your entire life with genes that lower cholesterol. The first of these was described in the New England Journal in 2006. A genetic variant in a gene called PCSK9 that reduced low-density lipoprotein, or LDL, the “bad cholesterol, by 15% reduced lifetime cardiovascular risk by 47%. A more potent PCSK9 variant cut LDL 28% and heart attack risk by 88%.
Drugs based on those genes have now progressed through clinical trials, and the first, from
Should I take Zetia?
The questions this begs: should I think about taking ezetimibe myself? I eat broccoli and take the stairs. But my cholesterol has sometimes tended to be high-- though I’ve lately kept my LDL around 100 with a combination of exercise and a healthy diet rich in nuts and olive oil. If I started taking Zetia now, lowering my cholesterol further, would it decrease my risk of a heart attack or stroke over decades? We don't know but the prospect is tantalizing.
For years, many cardiologists have been arguing that we might be able to eliminate heart disease if we could get our cholesterol down to the levels seen in hunter-gatherers. Even very healthy diets do not do this. A daily ezetimibe might -- but we don't have the data to prove it would. What would doctors say to patients who suggested it?
“The FDA has not approved it for anything like what you’re suggesting,” says Braunwald. “But you know in medical practice you sometimes have to go by the best evidence that’s around. Just recognizing the fact that the FDA hasn’t approved it for these purposes, I would certainly strongly consider it.”
There’s a big counter-argument. The study of ezetimibe from last fall – the only real test of the drug that Merck deigned to do – was in heart patients, not healthy people. Nissen, the longtime Vytorin critic, proposed a different solution to me if I insisted on being a one-person cholesterol-drug experiment: take a half-dose of the well-proven statin Pravachol, which has a great safety record and has been much more thoroughly studied. Most likely I’ll do what I always do when there’s conflicting medical evidence and I’m in no immediate peril: nothing.
But the saddest part is that we almost had an answer.
Califf says that when he initially approached Schering-Plough, he proposed a very different study. The side effect profile of ezetimibe was like placebo, he says. So why not test it in people like me? “I wanted to take 50,000 people, age 40, randomized to Zetia or placebo because adherence should be pretty high. But they didn’t want to do that trial. It would have been expensive,” Califf says. “It sure would be nice to know that.”
A missed opportunity
At the end of the day, the failure to do the right studies to test ezetimibe is as bad when we know the drug is effective as it would have been if it had turned out to be a placebo. Merck and Schering-Plough failed at their job as a drug companies, not only because more, better-designed studies would have protected them from criticism, but because if they had done them the market for Zetia could have been astronomically big. Imagine the sales of a medicine with incredibly mild side effects that could reduce heart attacks and strokes dramatically in anyone starting at the age of 40. It would have been a big gamble, but it was one with pretty good odds.
The lesson from the ezetimibe mess is still that we need big studies. One of the good things about the controversy is that it has forced companies developing new cholesterol drugs to test their effect on heart attacks and strokes early – the results for Regeneron’s drug, for instance, will be available in 2018, and possibly sooner. Without testing it fully, nobody is really smart enough to predict what a drug will do. Niacin, a B vitamin that was recommended by many doctors as an ezetimibe alternative, actually turned out to be ineffective and possibly harmful when larger studies were done.
These studies increase the cost of developing drugs dramatically. Califf hopes using electronic medical records and simple designs could bring the cost of medical research down. In the meantime, we’re left having to remember not to glorify drugs that lack good supporting data – but not to vilify them either. The single most trenchant thing anyone said to me about ezetimibe in more than a decade of writing about it was this, from Charles Hennekens of Florida Atlantic University back in 2009:
“Based on the evidence we had I never could understand why so many people were put on it. Now in what I still consider to be an absence of data, it’s good to use it less but it’s not good to brand it as ineffective or hazardous. I’d just really rather say there’s never been a lot of evidence that it should be widely used.”
The hardest thing about medicine – and about writing about it – is admitting that we don’t know what we don’t know. At some point, doctors and patients have to assemble all the evidence they can, make their best guesses as to why to do, and take leaps of faith. All we can do is collect more data and make the leaps a bit smaller.
This story has been updated slightly from its original version.
Also on Forbes:
