Gooz...I think the way this'll shake out is that overweight/obesity mediates inflammation and increases CV risk even with normal cholesterol levels (perhaps through mild insulin resistance or activation of the SIRT 1 pathway...see my post on this later today). If so this means we can limit CRP screening to this subset, leaving out the other apparently healthy individuals. Also at least one generic statin, simvastatin, which costs pennies a day, has been shown to reduce CRP as well as cholesterol, just like Crestor...so AstraZeneca shouldn't clean up that much...

I am in absolute agreement about the NNT and cost analysis. My first reaction to the trial results, and in particular, the warm embrace by the media was, "did anyone look at the costs and upside/downsides? Why are they touting this as penicillin reincarnated?

As far as offering counseling and ethics, I am more underwhelmed. Front line folks, despite best efforts, know where that leads. Especially given subset of total population was BMI challenged, and knowing benefits applied to all comers, not sure if you can invalidate positive findings?

DM association with statin make biologic sense? Odd.

Also, your back of the envelope calculation needs to take into account disability from MI and CVA. Obviously, I get your point, precision was not your intent, but it is possible that statin (Crestor?) use in this group of enrollees will be cost effective, despite conflicts.

Overall, media dropped the ball, I 100% agree. The dollars and cents question must be put front and center.

I just want to add a couple of points. I believe the main problem with treating people with statins for primary prevention is our inability to predict which people are destined to have heart attacks and strokes. If measuring CRP or other inflammatory markers can provide greater precision in estimating individual risk (this could mean some people's estimated risk goes down), that seems worth considering in certain cases, especially for people who are on the fence about whether to take a statin. Blood tests are relatively low cost and safe, especially compared to certain imaging tests, some of which involve exposure to radiation.

I agree that absolute risk reduction is more important than relative risk reduction in making decisions about treatment. However, each person's risk is different and people have different opinions about what degree of risk reduction is worthwhile to them. Some people would put a high value on reducing their risk of having a heart attack by a few percentage points over 10 years. Others would not.

I agree that statins that are still on patent are quite expensive, but generic statins are also available. In a few years, atorvastatin will be available in generic form. That said, there is no question that Astra-Zeneca's purpose in funding the study is to sell more of their drug.

You are right that the paper does not provide a complete breakdown of serious adverse events. It does provide *some* information in Table 4. Of note, risk of hemorrhagic stroke and cancer were not higher in the rosuvastatin group. The increased incidence of diabetes is worrisome, although a recent meta-analysis of statin trials did not find that effect. See:

http://www.ingentaconnect.com/content/libra/cmro/2008/00000024/00000005/...

Two other studies relating to CRP, also published yesterday, may be of interest:

http://circoutcomes.ahajournals.org:80/cgi/content/abstract/CIRCOUTCOMES...

http://circ.ahajournals.org/cgi/content/abstract/CIRCULATIONAHA.108.8142...

I agree that with Brad that the costs of treating cardiovascular disease need to be figured into the analysis. I addition, one must consider disability and death that is avoided.

Marilyn Mann

The issue isn't cost-effectiveness, which looks at the cost of a medical intervention per quality-adjusted life year achieved (even though I suspect including the costs of treating cardiovascular disease and avoided disability and deaths still wouldn't bring the level below $50,000 per QALY, the generally accepted cutoff point for cost-effectiveness). Rather, the issue for me is how to spend our marginal health care dollar in a resource-constrained environment. If the front end costs is, say $300,000 per foregone serious incident using statins for people with elevated CRP, how many foregone incidents would that same $300,000 achieve if spent on a program to reduce more serious risk factors like smoking, obesity or hypertension? As long as it was more than one (and there are drugs and counseling programs for each of those conditions that are significantly higher than that), then the marginal dollars would be better invested in those programs. In the coming era, if we're going to be serious about holding down health care costs and getting better results, we're going to have to engage in serious comparative cost-effectiveness analysis before deploying health care system resources in barely useful technologies like statins for elevated CRP.

The other way to look at it is from a moral perspective. Do we really want to live in a world where well-insured individuals can say, "hey, it lowers my risk a little, doesn't really hurt, and I'm well-insured so it doesn't cost me much beyond a co-pay" when so many people at serious risk go without any kind of preventive care?

"Did [rosuvastatin] reduce CRP levels, and did that reduce heart attacks, strokes and, most importantly, sudden death from cardiovascular disease?"

I'm not sure you intended this meaning, but this sentence seems to imply that lowering CRP was causative in reducing these events. It is quite possible, however, that CRP is just a marker of inflammation and not causative. In other words, statins lower inflammation which in turn lowers risk of events -- and lowers CRP. (I hope that made sense.) Part of the evidence for this is a recent study that showed that people with higher CRP for genetic reasons did not have higher risk of cardiovascular disease:

http://content.nejm.org/cgi/content/abstract/359/18/1897

A couple of more points. Death from all causes was also reduced, by 20%. That seems important, because I'm only aware of one other primary prevention trial (WOSCOPS) where that was the case. The absolute benefit will of course vary according to baseline risk.

Finally, women and minorities showed a clear benefit in this trial.

Marilyn

In the interest of full disclosure, my husband is on atorvastatin for primary prevention and his cardiologist is Steve Nissen, one of the people quoted in most or all of the articles about JUPITER. However, he is considered high risk for several reasons not related to lifestyle and I think even Merrill would not object to him being on a statin. Merrill, if you do object, I don't want to hear about it!!

Marilyn

Aside from possibly being oversold, the thing that leaped out at me from the story was the fact that the DSMB stopped the trial after two years.  I think that's an incredible waste of an rare scientific opportunity to learn more about myopathy and other safety issues, in part because the question about cost-benefit analysis should involve long-run safety as well.  If as I suspect millions more Americans are placed on Crestor now, virtually all of them will be taking Crestor for much longer than the trial duration.

I understand the ethical arguments against continuing the placebo arm in a trial like that, but there are ways of modifying the trial design to get around this problem.  The bigger problem is that with the closure of the study, the patients are not likely to be followed in a scientific manner.  (The article says that a few were followed up to 4 years, but this was not systematic.) Even if you put all of the placebo/control arm folks on Crestor now, you could still compare the treatment and control groups three years down the road, b/c the treatment group would have had two years' more exposure to the statin (could use an 'intent to treat' analysis).

So an underappreciated problem as I see it is this: the result of this study will be vastly expanded prescription of rosuvastatin, and the safety costs of that expansion are largely unknown, but could have been addressed with a continued study.

This is one of those cases that cries out for a stronger, truly robust Phase IV system, much more than what we got with FDAAA.

Dan C.

Thanks for giving us some perspective on this over reported study. I am very alarmed that no one in these studies or the media ever report on the side effects of long term statin use.

The medical profession desperately needs to re-develop some skepticism regarding anything that comes from the pharmaceutical industry!

I see there are a couple of familiar names here and they are aware of what health hazards statins truly are as they have been written about elsewhere.
For goodness sake it is about time we all stopped the pretence that statins are a safe option for treating cardiovascular disease and to let people know what really can and does happen. We all know how they work, the pathway they interrupt and the vital co-enzymes etc that they interfere with.

I find it really hard to believe that so many postings that I have read in various newspaper comments have actually been genuine and not bought and paid for by someone with a steak in a positive outcome for statins.

Nobody is that stupid as to keep the "Wonder Drug" myth alive and stop anyone else from seeing the reality and harm that is being caused unless they are insane or being paid to do so.

I think it is time for everyone to take a reality check and stop trying to make statins something that they are not. These things cause serious health problems for so many people yet we end up with some idiot/s claiming they can do no harm and the facts are quickly brushed under the carpet or just played down.

The name of this trial is JUPITER, however I believe it should have been called PLUTO because there has to be some ties to Mickey Mouse with all these claims and realistically why would they end the trial unless there were signs of something going wrong for the outcome?

Thanks for this, Gooz. It is a great service.

Re: the DM increased relative risk, I believe the authors at one point suggest it may be 'the result of chance' (although they do not deny stat significance). However all this turns out, there is much that is eerily familiar.

Merrill--

First, I agree that we should look at the cost in terms of what else might have been done what that money to prevent premature death. As you point out, how many lives could be saved if tha
money were spent on smoking cessation clinics, etc?

The more I read and learn about health care, the more persuaded I am that much of the money
spent on medical services to prevent disease would
do more good if spent on public health programs.

Finally I, too, am curious about that figure suggesting that 20% fewer of the people on Crestor died of all causes. Is there any evidence that this was because they were on Crestor?

As you point out this was a group of older-elderly adults, many obese. One could expect many of them to die over a two-year period for many reasons. That 20% figure feels like a red herring--could it be a blip on the screen?

While writing this, I just scrolled up and saw\
the comment from Justice MI.

He suggests that 20% reduction might have been "just chance"

I had seen the headline and suspected very much of the 'results' were massaged. Gooz did a good job of exposing its reality.

PBS interviewed Drs. Harlan Krumholz and Mark Hlatky on the JUPITER trial. Here's a link to the transcript:

http://www.pbs.org/newshour/bb/health/july-dec08/statins_11-10.html

Maggie, the decrease in total mortality was statistically significant and unlikely to be due to chance. When you decrease death from cardiovascular disease enough, as in JUPITER, and do not have any increase in death from non-cardiovascular causes, you are likely to get a modest reduction in total mortality, in this case 20%. This has been shown over and over in statin secondary prevention trials, and was shown in one primary prevention trial that I know of (WOSCOPS).

Marilyn

"I . . . am curious about that figure suggesting that 20% fewer of the people on Crestor died of all causes. Is there any evidence that this was because they were on Crestor?"

Maggie,

This was a huge randomized controlled trial. If you look at the two groups in Table 1, you will see that they are extremely similar in age, percentage of women, BMI, blood pressure, cholesterol levels, and so forth. There is no reason to think that the reduction in total mortality or any of the other results are due to anything other than treatment with rosuvastatin.

Marilyn