Take more pills?

The big medical breakthrough news of the day is the so-called JUPITER trial, which you can read all about here thanks to NEJM's increasing open access practices. It's all the rage these days to give clinical trials snappy acronyms -- in this case it's Justification for the Use of Statins in Primary Prevention: an Intervention Trial Evaluating Rosuvastatin. The reason for the clever branding is that this trial was paid for entirely by AstraZeneca, the manufacturer of and patent holder on rosuvastatin, which they would prefer we call by the brand name Crestor.

Okay, here's the background, the study result, and what I think it means. As usual, we are not dispensing medical advice here. Inform yourself, consult your doctor, make a decision you are comfortable with.

Statins were originally developed and tested, and subsequently prescribed, based on the observation that high levels of Low Density Lipoprotein, aka bad cholesterol, are associated with vascular disease leading to heart attacks and strokes. They block the action of an enzyme involved in the synthesis of LDL, hence lowering LDL levels. Ergo, they have been prescribed for people with elevated levels of LDL.

For quite some time, they were prescribed even though there was little or no evidence that they actually prevent heart attack or stroke in people who haven't already had one. Some people objected to this practice on the grounds of monetary cost and the risk of side effects. The latter concern was amplified when one statin was found to pose a substantial risk of a very serious side effect called rabdomyolisis. That particular drug was withdrawn from the market, and others have proven to be much less likely to cause that particular problem.

Evidence has been growing that statins can indeed reduce the risk of cardiovascular and cerebrovascular disease and serious events in otherwise healthy people with high LDL levels, and so they are increasingly prescribed. However, it turns out there is another marker for risk of vascular disease, called C Reactive Protein, CRP, which is a sign of inflammation. It so happens that statins also have an anti-inflammatory effect and lower CRP.

So, AstraZeneca paid the patent holder of the CRP test to do a large scale trial in which people who did not have elevated LDL, but did have elevated CRP, were given rosuvastatin. The trial was stopped early, because treatment reduced the rate of hospitalization, angioplasty or bypass surgery, heart attack and stroke by about half, and the death rate from all causes by 20%.

Slam dunk, right? Maybe, but these questions often look different when you look at absolute risk, rather than relative risk. Another way of stating these results is that 25 people would have to take the drug for 5 years to prevent a single adverse incident. In other words, if you are in the low LDL-high CRP group, there's a 4% chance you would actually get a noticeable benefit by taking this stuff for 5 years. It's a big benefit, to be sure -- who wants to have a heart attack? Still, if you put it that way, you're more likely to think about the cost and possible risks. Side effects were not a big problem in this trial -- the only notable concern was an increase in the diagnosis of diabetes, although blood tests for the effects of diabetes didn't show a meaningful difference.

Still, the trial was halted after less than two years, and while there is some longer term follow-up available, we've had lots of experiences with adverse effects of medications emerging only after long experience with large populations. In particular, people who start out with normal LDL levels end up with abnormally low levels when they take statins. That does seem at least intuitively suggestive of a concern.

As for cost, there are generic statins available at 1/3 or less the cost of rosuvastatin. They all work the same way, so I personally see no reason for anybody, anywhere, to be prescribed a brand name statin, including Crestor. That's purely a waste of money. Take a generic. Of course, AstraZeneca would not have paid for a trial of a generic chemical.

Second, it will require considerably more information and analysis to understand the cost effectiveness of such an intervention at the population level. There is an opportunity cost -- there are other ways to spend the health care dollar. For one thing, a very good way to get elevated CRP is to have gum disease. Maybe universal dental care would do more than prescribing statins, and have other benefits as well. (Yep, it's true -- having your teeth cleaned prevents heart attacks.)

CRP is also raised by spending time near highways, breathing ultrafine particles from motor vehicle exhaust, and other exposures to air pollution. It so happens that a disproportionately popular place to put public housing is near highways, because the land is cheap and people who can afford not to don't want to live there. Maybe we could achieve a population health benefit, not to mention justice, by choosing to build affordable housing in safer locations, and reducing other sources of air pollution.

So think about it. Do we really want to view pills as the answer to every problem? Or is it worth considering other options?