Any thoughts out there on the recently announced JUPITER trial results?

http://content.nejm.org/cgi/content/full/NEJMoa0807646
If you have NEJM access.

My initial read through of the article leaves me highly impressed, although the Number needed to treat for overall benefit is in the upper 70s, the data is impressive.

It will be a few days or so before I get to dig through the article deeper.

Any thoughts out there on the recently announced JUPITER trial results?

http://content.nejm.org/cgi/content/full/NEJMoa0807646
If you have NEJM access.

I think it is groundbreaking pathophysiologically, i.e. that you can target a patient's high hs-CRP and lower their mortality risk with rosuvastatin, even though they have "normal" lipid levels. Supposedly the effects of rosuvastatin were independent of the lowering of LDL to very low levels. However how low should we go concerning LDL?

I am inclinded to believe that this study will eventually lead to other studies that will eventually have physicians ordering hs-CRP in addition to lipid levels. Some might not be willing to put patients on a statin long term, especially a patient that is otherwise healthy as statins does have some side-effects, and supposedly according to this and other studies the possibility that statins may increase diabetes rates really needs to be looked at.

Clinically, I don't think there will be a lot of big changes in the next couple years, until they integrate the study, and hopefully others, into newer guidelines. In the end exercise can reduce hs-CRP as well and a head to head study with lifestyle interventions vs. rosuvastatin need to be done. Healthcare is alreay so expensive and this medication is expensive too.

I don't think the NNT is necessarily that much of a detriment to using this approach as the study only went about 2 years and the authors extrapolated lower NNT when looked in the longterm which are much more impressive. The authors calculated that for fours on Crestor the NNT to prevent one primary end point was 31. What about treatment for 10 years, would the NNT drop to 15 or even 10???

Supposedly at 5 year extrapolation the NNT is 25 which according to Ridker is less than that used for treating hyperlipidemia in primary prevention. So it is possible that targeting hs-CRP may be more important than targeting the LDL!

As this is a primary prevention trial I think the NNT is pretty good for just less than two years. Of course this could be off-set by increased risk of diabetes, which could be chance. Some endocrinologist don't feel that statins increase diabetes incidence. There needs to be more studies done.

I actually think the relative reduction in risk, near 50%, is more important as the endpoints such as MI, stroke impact such a huge percentage of the population at some point in their life, if not the next two years, that anything as dramatic at this warrants more trials and consideration. Many patients who believe they are healthy with normal lipid profiles could benefit from knowing their hs-CRP is too high. Again, however, this may just be another number telling patients who are obese that they need to lose weight as a good share of patients in this study had metabolic syndrome and were already a setup for having an MI.

Given how expensive it is to treat an MI, the economics might favor long-term treatment with rosuvastatin or possibly another statin. Also there is the morbidity of MIs and strokes which put healthy people out of work every year, and are devastating diseases so that will also need to be factored in.

Also, consider that many of these relatively healthy patients who would receive statins are obese and would later need healthcare at a later date involving a statin so I would guess there would be benefits in getting them to eat healthier and take care of themselves better before this stage. A ounce of prevention is worth a pound of cure and I think this prevention strategy although pharmacological, will save a lot of lives and health care dollars decades from now.

If it proven that another statin can do the job as well, and if the price of statins drops even further, say to cost only 5 thousand a year or less then it might be worth it if you prevent a hospital visit and workup for an MI patient or decades of physical therapy/assistance for a stroke patient. Certainly worth it to the individual patient.

There is A LOT of skepticism and outright derogatory comments made about the authors who hold patents on these medications. I think this is jealousy.

When I was a med student I remarked during a Problem-Based learning session that infection with C. Pneumoniae may be a cause of athersclerosis and heart disease by leading to inflammation and was basically laughed at. A lot is still unknown about cardiovascular disease such as why heart attacks rates have plummetted although we are more obese as a society since the 1960s?? There is more to pathophysiology than lipids clogging arteries in obese people.

I think that physicians who didn't really understand the complex mechanism of atherosclerosis and believed lipids to be the whole story are threatened by something this relatively new concerning the role of inflammation and coronary artery disease. This study I believe will lead to clinical practice changes in the future after more trials and some borderline patients i.e. men greater than 50 years or women greater than 60 years with intermediate risk who don't qualify for statins will be tested for hs-CRP and put on statins by their cardiologists/PCP.

The significance of this study is that it is paradigm-changing, i.e. it will change how physicians look at cardiovascular disease. One day the hs-CRP could be the most watched number when looking at all types of prevention and pharmacologic therapy and when atherosclerosis is taught in medical school (and perhaps when more is known about the pathophysiology) it will be described as an inflammatoy disorder. Wow. Most cardiologists are comfortable with this, most general internists are not.

Interesting comment on C. pneumoniae. WIZARD and ACES are the largest trials involved to address that issue. My recollection was that they found no impact in trying to treat with antibiotics. Unfortunately, at least one biotech company founded on an antibiotic folded due to the trials.

My sentiments are similar regarding lipids and inflammation. Didn't think about physicians feeling threatened by the finding but that explains some of the reactions I'm seeing. A cardiologist shocked me once by telling me lipids were boring.

As far as the NNT, some say it's 120, but fail to take into count person-years were used instead of absolute numbers. I agree that a NNT of 25 is phenomenal for a prevention trial (Aspirin giving a NNT of something like 500 for CV events).

Regardless, it'll be interest to see how things pan out in the prevention world.

Interesting comment on C. pneumoniae. WIZARD and ACES are the largest trials involved to address that issue. My recollection was that they found no impact in trying to treat with antibiotics. Unfortunately, at least one biotech company founded on an antibiotic folded due to the trials.

My sentiments are similar regarding lipids and inflammation. Didn't think about physicians feeling threatened by the finding but that explains some of the reactions I'm seeing. A cardiologist shocked me once by telling me lipids were boring.

As far as the NNT, some say it's 120, but fail to take into count person-years were used instead of absolute numbers. I agree that a NNT of 25 is phenomenal for a prevention trial (Aspirin giving a NNT of something like 500 for CV events).

Regardless, it'll be interest to see how things pan out in the prevention world.

I think that the largest trials showed that azithromycin treatment for secondary prevention in people with established atherosclerosis was not effective, . . . at least in terms of using azithromycin for their treatment period. I think that this doesn't necessarily mean that C. Pneumoniae is not involved in CAD, rather in my mind that C. Pneumoniae may very well be an initiating factor in many patients with CAD and the patients treated had such advanced CAD that the presence of C. Pneumoniae was no longer required for disease progression. Sort of like getting a bug bite and then a festering wound developes and you get cellulitis . . .

I think that with the role of inflammation in CAD becoming more central there will be a push to identify more vectors associated with CAD such as viruses and bacteria. Patients with bad oral hygiene predisposes them to CAD, so there is other evidence out there.

I have hypothesized that the fluctuation, i.e. recent decrease in MI over the past decades despite increasing obesity is secondary to changes in infection with some organism or even to more widespread use of antibiotics which might be more useful in primary rather than secondary prevention, IMHO.

excellent comments by Darth.

I also agree that inflammation likely plays a quite important role in the pathophy. of cardiovascular dz. If you just look at the eleveated CAD risk in people with rheumatoid arthritis and some other rheumatologic disorders, that tells you something...or at least suggests.

I don't think this one trial is enough to massively change practice, but you can bet their will be followup studies. It remains to be seen.

As far as the comment about us having more obese patients but less cardiovascular events now, I think we have more obese patients but probably with better treatment of their comorbidities (like elevated LDL and hypertension) than we likely had in the 1970's or so. Also people are still eating too much but many know about trying to limit their total cholesterol intake (so pigging out on Healthy Choice cookies instead of bacon and eggs, perhaps?).

This study is interesting but it is has some of problems. The main one as I see it is that is was stopped early. This seriously harms the external validity.

This study is interesting but it is has some of problems. The main one as I see it is that is was stopped early. This seriously harms the external validity.

Why is this an issue when they reach their predefined statistical endpoint? and it wasn't stopped early, it reached their primary end point.
was an event-driven trial designed to continue until 520 confirmed primary end points had been documented, to provide a statistical power of 90% to detect a 25% reduction in the rate of the primary end point, with a two-sided significance level of 0.05.

Does anyone think that the findings of the study are less reliable, given the fact that AstraZeneca funded the study (maker of Rosuvastatin) and the lead author holds patents on the use of hs-CRP as a inflammatory biomarker?

Just curious... seems like both parties seem to benefit well from this study.

;

I definitely don't disagree with C. pneumoniae still being included in the equation of inflammation and cardiovascular disease. However, for antibiotics to have any clinically meaningful outcome on atherosclerosis and CAD remains to be proven. No large-scale prospective clinical trial testing this hypothesis has rejected the null. I interpret that to mean the disease is too far advanced at the point tested for antibiotics to make any significant changes in outcomes.

As far as conflict of interest go, yes, Dr. Ridker does hold a patent on hs-CRP assay. I would argue that if you discovered any phenomenal diagnostic test, you'd be a fool NOT to patent it. Researchers have got to eat, too, and a patent kills two birds by boosting the CV and providing a source of income. He also is one of the world authorities on CRP in cardiovascular disease. It would have been more surprising not to see any kind of patents, etc.

I will also point out that the JUPITER investigators approached AZP, NOT the other way around. This was an investigator-initiated multi-national project. Getting 18,000 subjects together and studying them well are no little tasks. Considering 4% of the NHANES cohort qualified (a large section of the American population, if projected), this study can potentially be a big deal if it is reproduced in subsequent trials. To question external validity means to question the generalizability of the trial. Again, this was a multi-national trial.

The flipside of the question on terminating the study early: How can you not stop the trial after you established clear evidence of harm in one of the arms (placebo)?

I can understand the mistrust of big pharma biasing everything, but one must remember they do contribute to research and fund a lot of studies that wouldn't otherwise see light of day. Where else are we going to get new drugs and devices? I strongly support evidence-based usage of them and critical evaluation of them, but I also cannot see a world where medicine and academics operate without industry helping. To discard anything done by industry is akin to tossing out the baby with the bath water.

As a correction to my earlier statement on NNT: the NNT of 120 is accurate for only the median 1.9 years of research time. The NNT projected over 5 years (<- the part everyone forgets) was 25.

P.S. Saw the subgroup analyses (metabolic syndrome, obesity, diabetes, etc.) which make the study even stronger. Metabolic syndrome subjects benefited LESS than non-metabolic syndrome subjects. The distribution of obesity was actually fairly equal. Roughly a third had normal BMI. The average American BMI is 28-29 (disgusting isn't it?), so it's fairly representative.

The flipside of the question on terminating the study early: How can you not stop the trial after you established clear evidence of harm in one of the arms (placebo)?

I fully agree with this, they reach statistical significance and found the treatment group did good, so why not stop it? It would be nice to see a little longer median follow-up time, but I guess we can live with 19 months.

they do contribute to research and fund a lot of studies that wouldn't otherwise see light of day.

They also fund a lot of research which never sees the light of day.


As a correction to my earlier statement on NNT: the NNT of 120 is accurate for only the median 1.9 years of research time. The NNT projected over 5 years (<- the part everyone forgets) was 25.

Even if you use the NNT of 25, you're looking at $174,000 to prevent 1 event. and with the way this was worded, this was the combined primary endpoint, not death or major cardiovascular event. The number needed to treat to prevent death from any cause was something like 180 over the 19 months.

Without a doubt, this seems to be a well done trial. But I don't think I'll be jumping on board with putting Crestor in the drinking water just yet.


P.S. Saw the subgroup analyses (metabolic syndrome, obesity, diabetes, etc.) which make the study even stronger. Metabolic syndrome subjects benefited LESS than non-metabolic syndrome subjects. The distribution of obesity was actually fairly equal. Roughly a third had normal BMI. The average American BMI is 28-29 (disgusting isn't it?), so it's fairly representative.

What I found just a little odd was that they seemd to try and control for DM, HTN, and other comorbid conditions and there was still 44% of patients with Metabolic syndrome in the trial. So I'd still like to see a trial which looks at out comes if you reduce metabolic syndrome (yeah, I know, an impossible trial to run).

The whole question of "physician-reported" diabetes is odd in this study. At completion, there was no significant difference between the study groups with respect to fasting blood glucose, newly diagnosed glycosuria, or HbA1c levels.

So, how exactly were these physicians diagnosing DM? With OGTT? Random blood glucose?

Further, the large number of participants with the metabolic syndrome (41%) makes it diffiult to fish out what role, if any, Rosuvastatin had in the subsequent development of diabetes.

If the majority of new cases were diagnosed by random blood glucose, then I remain unconvinced. Color me skeptical, but I question the accuracy of a "random glucose" in patients with the metabolic sydrome. The random blood glucose reflects insulin sensitivity and we already know these individuals, by definition, are resistant to the post prandial effects of insulin.

This question may be answered (or muddied even further) by the AIM-HIGH study (atherothrombosis in metabolic syndrome with low HDL/high triglyceride). This study recruited 3,300 individuals with metabolic syndrome rather than overtly diabetic individuals, and will compare a statin vs. statin + niacin with goal LDL < 80.

JUPITER certainly has garnered quite a few skeptics and rightly so . . . Suffice it to say, analyses are underway outside of Dr. Ridker's study group to address many of those concerns. The results are very surprising, but I am not allowed to release the info. I'll just say it's my opinion that I'm more convinced about applicability to the US population.

The bigger question is whether hs-CRP reduction should become a treatment goal.

As far as the NNT to prevent death . . . I acknowledge that and also recognize we tread going into a very annoying argument that misses the very big picture. One should recognize for a PRIMARY prevention trial, a NNT < 500 for anything is actually quite good. Even if the NNT were 1, someone will find something else to nitpick about the trial. Believe it or not, a certain group (refuted by the medical community at large) sincerely believes statins don't benefit in any type of prevention of MI's, simply because they are convinced it doesn't reduce death.

Has that cost of $174,000 been compared to the cost of having an event in an entire population? Standardized to per 1000 persons would make a suitable comparison. All that shows is interventions cost money. For all we know, something cheaper like niacin and exercise/diet might be equally effective.

Interesting comment on the AIM-HIGH study. I am involved at one of the sites, albeit in a very,very early trainee capacity. We serve as a site for many such trials, so that's why JUPITER is proving interesting to me. For the glucose, do not forget the ADA definition of diabetes currently relies on "random" glucose and fasting glucose. The AIM-HIGH study will tell us about niacin's effect on glucose, not statin's. It is possible Dr. Ridker will publish a sub-sub group analysis looking at incidence of diabetes adjusted for metabolic syndrome that will better answer the question. Lastly, if they did develop diabetes, they'd HAVE to be on a statin anyways based on ATP III.

The bigger question is whether hs-CRP reduction should become a treatment goal.

Agreed, it would seem that treating to this level may be more sensible than cholesterol levels alone.


As far as the NNT to prevent death . . . I acknowledge that and also recognize we tread going into a very annoying argument that misses the very big picture. One should recognize for a PRIMARY prevention trial, a NNT < 500 for anything is actually quite good.

at $90/month and just shy of $1180/year, this may nto be an acceptable NNT in primary prevention groups. Asa is a wonderful med because it's so cheap, and if I remember correctly, wasn't it's NNT 80 over a 5 year period?


Has that cost of $174,000 been compared to the cost of having an event in an entire population? Standardized to per 1000 persons would make a suitable comparison.

There was an old, limited study in annals which suggested that asa & a statin together were cost effective if their risk of an event was >10%.

From my limited observations, when I sign my charts in medical records, they include a bill sheet. Those who have the minor components of the primary endpoint's bills are typically less than $30,000 for the hospitalization. So I'm going to have a hard time seeing how $174,000 would be cost effective, but that's simply based on Crestor, not any other statin.


All that shows is interventions cost money. For all we know, something cheaper like niacin and exercise/diet might be equally effective.

I probably won't jump on the rosuvastatin boat just yet, not when we can get simvastatin for $4 for 5mg/10mg tabs, even if we have to double the number of pills/day, that will still be substancially cheaper than Crestor.

...For the glucose, do not forget the ADA definition of diabetes currently relies on "random" glucose and fasting glucose...

I did acknowledge "random" glucose and further gave my concerns regarding its use as a diagnostic modality in patient's with metabolic syndrome.

Again, these individuals have problems with post-prandial insulin resistance and this is reflected by an elevated random glucose. If these patients concurrently have normal HbA1c levels and normal fasting glucose (as in this study), then what exactly is a "random" glucose telling us? Just food for thought...

The AIM-HIGH study will tell us about niacin's effect on glucose, not statin's.

Both the control and the experimental group will be taking simvastatin +/- Niacin...my point, is that maybe something useful can be extrapolated about the incidence of diabetes in AIM-HIGH compared to previous studies, since all participants have the metabolic syndrome. Obviously, a third placebo arm would make a better comparison.

If the incidence of diabetes is similar or greater than in JUPITER (where 41% of participants had metabolic syndrome), then it may lend weight to the arguement that this is a class effect rather than inherent to Atorvastatin. Further, it would be interesting from a mechanistic standpoint to further clarify if statins simply impair post-prandial glucose metabolism, or if they do indeed hasten the development of overt diabetes. The interesting part, at least to me, is WHY?

But, as you stated, if they do develop diabetes they would have to be on a statin regardless.