NOVEMBER 10, 2008

JUPITER IS NOT LIKELY TO TURN ON APHRODITE

Mrs. Bremner [DOCTOR BREMNER to YOU guys!] is attending the Annual Meeting of the American Heart Association this week in New Orleans and texted me some news about the release today of results from the JUPITeR study ("Justification for the Use of statins in Prevention: an Intervention Trial evaluating Rosuvastatin (JUPITeR)") published in the New England Journal of Medicine

.

Jupiter

(gasp), Oh really. I think this one would definitely not have been a turn on for Aphrodite.

Ha, ha.

Based on the results of this study, that a biomarker of heart disease, C-reactive Protein (CRP), predicts response to the drug Crestor (rosuvastatin) (and not just total cholesterol or LDL cholesterol), the authors of this study would apparently like us to come to the conclusion that, basically, all of us should be on a statin medication for the prevention of heart disease. In fact, I can already hear the shills and the media saying "revolutionary" and "likely to change practice"! In fact there is even a cheesy site on the New England's website asking us how it will change our practices (gasp, shame).

We've been through this drill before, when our country's "leading" cardiologists informed us that statins were so great that they should be put in the drinking water.

But now, let us turn our attention to the results of the JUPITeR trial.

17,802 healthy men and women with LDL cholesterol of less than 130 mg per deciliter and CRPs of greater than 2.0 mg per liter were given either rosuvastatin 20 mg per day or placebo. They were then followed for the occurrence of the "combined primary endpoint" of heart attack, stroked, arterial revascularization, hospitalization for unstable angina, or death from cardiovascular causes.

First of all, whenever doctors start combining real disease outcomes (like dying of a heart attack) with what are primarily... er... inconveniences, like going to the hospital, you have to start wondering about the study.

OK. So what about the results, you say. For their primary outcome out of 100 people treated each year there were 0.77 versus 1.36 who had an event like heart attack, hospitalization, etc. That is a difference of 0.59%. But I am going to ignore their cheesy primary endpoint and look at what to me matters, which is heart attacks. For each year of treatment out of 100 people 0.17 in the Crestor group versus 0.37 in the placebo group had a heart attack, or a difference of 0.2%, which would have amounted to a 1% difference over five years if the study had continued for that long, but the study committee cheerfully "stopped" the study after 1.9 years since they thought things were going so well that it would be "unethical" to continue the study (Hooray!).

I guess that 1% of responders outweighs the 3% or so that will develop liver toxicity or muscle damage.

Study results that will change clinical practice, indeed.

BTW, look at the end of the article for the long list of consulting agreements to the pharmaceutical industry engaged in by the authors of the study. And funding by AstraZeneca.

COMMENTS

Viola Vaccarino wrote on November 10, 2008

The study excluded subjects with diabetes or hypertension and there was a four week "placebo" run in phase and subjects who were not compliant during that phase were dropped. 80% of the patients were excluded and many more were excluded during the placebo run in phase. What this means for the regular patient is unknown.

Marilyn Mann wrote on November 10, 2008

Just read this. Wondering why you did not mention 20% reduction in total mortality. Also interesting that women benefited to the same extent as men.

See studies published ahead-of-print yesterday in Circulation and Circulation Cardiovascular Quality and Outcomes for more info on CRP.

C-Reactive Protein and Parental History Improve Global Cardiovascular Risk Prediction. The Reynolds Risk Score for Men Paul M Ridker, Nina P. Paynter, Nader Rifai, J. Michael Gaziano, and Nancy R. Cook Circulation published 9 November 2008, 10.1161/CIRCULATIONAHA.108.814251

C-Reactive Protein and Reclassification of Cardiovascular Risk in the Framingham Heart Study Peter W.F. Wilson, Michael Pencina, Paul Jacques, Jacob Selhub, Ralph D'Agostino, and Christopher J. O'Donnell Circ Cardiovasc Qual Outcomes published 9 November 2008, 10.1161/CIRCOUTCOMES.108.831198

Doug Bremner wrote on November 10, 2008

The reduction in total mortality was 1.0 per 100 patient years versus 1.25 in the placebo group, a difference of 25% in relative risk, but of 0.25% in absolute risk (i.e. not that great). I think one thing this study highlights is that maybe LDL reduction is not the "cause" of risk reduction, or at least exclusively.

The other issue is what the clinical implications of this study are. Are doctors going to check CRP and LDL in everyone and treat whoever fits either profile? That means everyone will be on this stuff. Crazy.

David Diamond wrote on November 10, 2008

Doug – great commentary on the Jupiter trial and kudos to you for your muckraking blog

the whole statin revolution seems driven by big pharma greed with little real benefit to the population, followed by sheep-like docs who can’t see the flaws in the research

can you imagine turning the CHD word upside down by showing better CHD outcomes with a diet low in sugar and high in chocolate vs a statin?

Odysseas wrote on November 10, 2008

Again, I think people get so enamored by the relative risk that they dont even consider the difference in absolute risk or the NNT... (The calculation may be useful for some readers) As the fanfare dies down, perhaps a thoughtful discussion will ensue..

Doug Bremner wrote on November 10, 2008

Yes, the relative risk v absolute risk distinction is critical. The story I use is for primary prevention in men with risk factors. Out of 100 men, 4 men on placebo will have a heart attack in five years v 3 on a statin. If you told someone that a drug would reduce his risk of having a heart attack by 25% (and told him none of the risks, he would take it! If you went into a room of 100 men and said I want you all to take a drug and it will prevent only one of you from having a heart attack, but three of you will have some liver damage or muscle pain, and maybe it causes cancer but we are not sure, they wouldn't take it. Both describe the same situation (going from 3 to 4 is a 25% change (relative risk) but 4%-3%=1%). Crucial distinction. Most doctors don't understand, but the pharma detail people do. And they always use the one that suits them best. Also, NNT is the number needed to treat. Means that (in this case) 31 patients need to be treated for four years to save one from a cardiovascular event or procedure (their primary outcome). Which isn't all that great. The authors pointed out that this is similar to prior studies of those with high LDL, but all that says is that LDL is not the be all and end all (the cardiology research community has been moving toward the inflammatory pathways area for several years now; elevated CRP is probably associated with the metabolic syndrome, etc).

Thanks for making these important points. They deserve a press story.

Mathew? Caleb? Anyone?

David wrote on November 11, 2008 on his blog dcscience

Here is a link to a discussion of should we all take statins and relative versus absolute risk

[David is currently discussing creationism; recommended blog for fans of science and skepticism]

Therapy Patient wrote on November 12

I groaned when I heard that report on NPR today and was disappointed none of the interviewed doctors mentioned muscle and liver problems. My Dad was one with muscle pain and weakness and a deteriorating liver but his cardiologist insisted he stay on the drug. Dad stayed on the drug and Dad died. Well he was 90 and had a host of problems, so who is to say what killed him? The autopsy listed a long list of problems. I take a statin, that's for sure!

With statins as a class of medications: A.E.s are often under-reported with high doses of potent statins in particular. Additionally, there is no reduction in mortality or increase in the lifespan with those on statin therapy.

Dan Abshear wrote on November 12

Several risk factors should determine if one is placed on statin therapy, and not just one. Statins do decrease CV events and risks significantly.

The meds. increase endothelial function, stabilize coronary plaque build up, and decrease thrombus formation. Maximum reduction in LDL is evaluated after about a month of therapy from any particular statin. There is evidence to suggest that statins have other benefits besides lowering LDL, such as reducing inflammation (CRP), those with dementia or Parkinson's disease, and some forms of Cancer and cataracts. It appears those statins produced by fermentation, such as Zocor and Pravachol, have less myopathy than the other synthetic statins, possibly due to being more hydrophyllic. Yet overall, the existing cholesterol lowering recommendations should be re-evaluated, as they may be over-exaggerated. With children, diet needs to be controlled with them so they do not develop arteries of one who is middle age and a possible candidate for statin therapy,

Dan Abshear

To comment on this blog please write to jamesdouglasbremner@yahoo.com and indicate whether comments are public or private. If you click on the comment link your comments will get truncated after a couple of sentences (sorry).

Read More and read comments (5 comments) --

permanent link --

Make a Comment
View Comments