This is definitely a ground breaking study and will change my practice. I always find it interesting to hear the comments about cost and NNT, but patients/families don't care about this when it comes to life or death. We are talking about the #1 killer for which many patients first sign of disease is death. I would guess that many who are playing devils advocate, will be the first to test themselves for CRP.

In regards to some of the other comments made here, I have no concern for the study being stopped early. It was a placebo controlled trial that would have been unethical to continue when you have 50% reduction in events. The study was powered to show a 25% reduction at 5 years and with this dramatic benefit, the monitoring board had no other choice. Regarding the NNT/absolute risk reduction, I was actually very impressed considering this was a healthy population that is not currently being treated. If I can reduce my patients risk of dying from 2% to 1%, that is nothing to turn up our noses at especially in a "relatively healthy population". In regards to generics, the data shows us that the vast majority of patients do not reach their goals when treated with generic statin medications and the risk of side effects to our patients is greatly increased when using the much higher doses necessary to try an achieve targets. Jupiter provides further evidence to the LDL hypothesis that lower is better.

Another study, since initially suggested by ASCOT-LLA, demonstrating that treating cholesterol in subjects at risk despite "normal" baseline cholesterol prevents events. The issue I have with this study is using a risk marker as weak as HS-CRP.  HS-CRP is of very marginal benefit in the primary prevention setting, MESA found it to be of no benefit in stratifying risk.  This study gives us inspiration to find a true novel marker of risk to determine who needs treatment. 

It occurs to me that EBT-CAC provides such a measure as demonstrated in MESA where CAC 1-100 was 3.6 times more powerful for all cardiovascular events than Framingham.  CAC >100 was 7.7 times more predictive than Framingham  and CAC > 300 was 9.6 times more predictive for all cardiovascular events than Framingham. 

 It is clearly time for the investigation power brokers to get past their bias against coronary calcium imaging and fund a study using CAC as the indication for treatment similar to how HS-CRP was used in this study.  

One possible clue to suggest that the observed risk reduction was indeed due to CRP reduction could be the consistent benefit observed in the "10% or lower" absolute risk subgroup. According to the authors, those patients had no metabolic syndrome. In regard to adverse effects, I believe that cancer could be an important concern in view of a recent meta-analysis published in JACC last year, where the authors found a disturbing inverse correlation between absolute levels of LDLc achieved and newly diagnosed cancer. Even though those findings can only be vieved as hypothesis - generating, it seems obvious that a 1,9-year period may be too short to reassure us about the safety of keeping such low LDL levels by pharmacologic means for several years. 

Great study:

since it is a primary prevention study,we should not expect a large absolute reduction in overall outcome!!

considering that our lifestyle "in USA"is defenitly not the healthiest,as far as diet ,exercise,etc,I think a study like this,is worth evaluating and cosidering in our practice.

I have been taking crestor 3-4 days a week,(2 brothers passed in their 50's ) yet my dad and mom lived to mid 80's.

I have occasionally screened patients with HSCRP,but I will defenitely do it more now.

Is it really a ground breaking study or is it construed to be. The ASCOT study was much more significant in the fact that we see more patients who are hypertensives and still only treated with antihypertensives.

The dose of Rosuvastatin ( 20mg ) is very significant. Why was 20mg used instead of 10mg? If 10mg is used which is logical for patients without elevated LDL, would it still benefit the patients?

In the ASCOT only 10mg of Atorvastatin was used. In the CARDS only 10mg of Atorvastatin was used for the diabetics!

It would be really interesting to know the view points on the above!

Good study, but better promo for CRP and Crestor. Nonspecific CRP results add clarity to treatment decisions for cholesterol, because of the weakness of LDL in predicting CV risk and as you pointed out Dr Topol, the prevalence of visceral adiposity and smoking in the study population. I think it would be foolish to initiate statin therapy based on LDL (or CRP ) alone. The validity of prediction rests on the statistical base with large N. In the examining room where N=1, for 95% of the encounters, there is a breakdown and LDL (and CRP?) risk is properly estimated in only 1 of 3, with the remainder under or overestimated. Given the complexity of risk factors for atherosclerosis and the fact that we measure them only in one point in time and have little clinical access to measures of oxidative stress that modify functionality of lipid particles or their metabolism, I bail out of the traditional approach. CIMT has served my patients well in the past 11 years (William B will advocate CAC). CIMT serves as the painless biopsy before starting "chemotherapy".

was the benefit due to lipid lowering or crp lowering?  how will anyone know.  there is strong evidence that LDL lowering benefits.  how does this trial prove anything beyond that?  generic simva offers similar ldl lowering.  why use this brand name when a generic will do the same.  in fact, simva 40 would have gotten most of the patients to goal in this study.

Dear Dr Topol, I feel we would certainly change our practice with Jupiter findings. Would you be available for a CME program in March2009 for Indian Drs on Statin therapy? If yes, then pls confirm the dates during which you can travel.

I feel that, whilst this trial is clearly intriguing, the trial design was sub-optimal and thus there are still some unanswered questions.

A 2x2 factorial design would have been ideal. Take patients with normal LDL cholesterol and measure hsCRP which would yield 2 groups (high CRP and normal CRP, both normal LDL): then randomise each of these to statin or placebo.

As Dr Topol points out, we don't know whether these patients improved because the CRP itself was reduced, or whether this was simply a marker of another process. 

However, I wouls say that the 40% with metabolic syndrome etc is important to note, but as long as they hjave not actually has an MI or stoke, it doesn't change the fact that this is a primary prevention study and, as pointed out above, absolute risk reductions would thus be expected to be smaller.

Worldwide I am sure this will lead to more CRP testing, but I don't think Crestor alone will reap the benefit. In cost-restricted systems (e.g. our NHS in UK!) generic statin prescriptions will see the most benefit. 

Importantly, patients will be aware of this trial and will now start turning up demanding hsCRP testing irrespective of their cholesterol!!

I just could not leave "real" comments alone.  Are you kidding me?  Where in your practice has generic simva had a consistent 50% ldl lowering at ANY dose?  WOW.  Time to hang up the shingle skippy.  Science has passed you by - LONG ago. And where in this study was the goal to "get these patients to goal"?  Have you read the study? It is available online.  Go find it. Read it. Change your practice, save some lifes.

WOW.

 

 

 

I am not sure that we have seen the whole picture, why they interrupted the follow up to just 1.9y, true if you see that an important relative reduction in all end points, but a clear tendency to increase in diabetes mellitus. I am concerned about this last issue, did they stop it because of this, what would happened if at the end of the 4y period of follow up the numbers could been worst?, would not be at the end another problem?, diabetes mellitus turns out to be a worst risk factor to coronary heart disease than mild risk factor of having hsCRP elevated?.  I think I will wait and see and continue with the actual guidelines.

I agree with B. Shah. A 2 x 2 factorial design as he proposes would have lead to more useful information.  The fact that only persons with a hs-CRP > 2.0 mg/L were included introduces a potential bias in assessing the results. A similar group with LDL <130 mg/dL & hs-CRP < 2.0 mg/L should have been used as a comparator to be able to correctly interpret the results. As has been pointed out the sample group was not truly a "healthy" cohort (40% with MS!!!). These are additional risk factors that could be acting as confounders. 

I have been testing some truly "normal" persons for hs-CRP and it is very difficult to find a really healthy person with high hs-CRP. I think a truly normal person with a high hs-CRP is an infrequent finding. The question is then: does the JUPITER population represent a real life scenario? I have my doubts and therefore I feel that before we get excited and change guideliness, better information is required. Till then, I would recommend caution in the application of the JUPITER information. "Not everything that glitters is gold" 

Too many fishy things.  They probably stopped the study because if it had continued the numbers would have gotten worse for AZ

I would really like to know more about the patient population, specifically their risk factor profile. I would assume this is available, I just haven't seen it yet. The results would be less surprising if these "healthy" patients actually had lots of major risk factors.

 

rlb

Jupiter seems to lend great support to the current theory that atherosclerosis is an inflammatory disease. Based upon the AFCAPS/TEXCAPS findings, reducing LDL-C further in patients with elevated hsCRP seems to be of great benefit. This supports the hypothesis that small dense LDL pariticles diffuse into the vessel wall, become entrapped due to their electonegativity are oxidized and induce and propogate vessel wall inflammation with subsequent unstable plaque formation. By reducing LDL-P with a statin, in this case the potent Crestor, one reduces the vessel wall inflammatory process resulting in plaque stabilization and reduced CVD events. 

The problem clinically is that hsCRP must be measured multiple times before it can be used as a marker for vascular inflammation. Perhaps as previously mentioned looking for subclinical disease with CAC scores or CIMT would be a better approach. Certainly this points out the failing of framingham risk scores.

Our recent election was decided by 52 to 47 and it seems that statin therapy will be endorsed by similar numbers. For the naysayers all I have to say is if your LDL is 135 do not treat it until after your initial coronary event. JUPITER shows that our "non elevated" cholesterol levels are too high. Clinical cardiovascular disease manifests in more than 50% of the population. In that sense universal treatment with a statin makes more sense than universal vaccination against tetanus. Questions of cost, generics etc apply but I do not expect Astra Zeneca to promote generic simvastatin. When they test the "polypill' in England we will have final proof if this approach is correct. Until then I prefer to be proactive.

Jupiter should not change anyone's practice.  It failed to prove that HS-CRP elevation demonstrated increased risk for a heart attack in the primary prevention setting as there was no control for this variable.  It failed to prove that Crestor was superior to generic statins as this was not tested.  Despite these obvious flaws, there are clinicians who are now going to order HS-CRP and place subjects on Crestor if the HS-CRP is elevated.

Jupiter is an example of marketing beating out science. Congratulations to Harvard, Ridker, and Crestor sales force.  Another example of the powerful taking advantage of the gullible.

 

If we look deep down at this study and several of Crestors previous studies...these are "Marketing" driven studies and not true "Scientifically" driven studies...please...screen 90000 patients so you can get exactly the type of patient that you can show benefit, then compare your statin to placebo...the 44% reduction in the Crestor arm was 142 patients versus 251 in the placebo arm...yes, a reduction of 44%...out of 17000 patients!!!...once again...a "marketing" driven study...which the media beautifully picked up and most people are touting as extraordinary...but why?...why?...because Crestor has to create a bigger Cholesterol market pie in order to grow more market share...why?...when Lipitor goes generic in 2011...the Cholesterol pie will take a big hit...unless you are Crestor and you have prepared some clever marketing studies over the past years to show that you are different than Lipitor and you are worth the cost beyond generics!...once again..."Marketing" driven study taken right out of the Pfizer playbook!!!

Forty percent of the patients in the study had the "metabolic" syndrome and yet the poulation was considered "healthy".  I understand that even those considered not to have the metabolic syndrome benefited. This leads me to believe that the population as a whole was not  "healthy" and that the criteria defining the metabolic syndrome are too conservative. Focus and promote weight loss, dietary restriction/modification and regular exercise.  A pill is not a panacea.

Although it is true that 40% in the trial had the metabolic syndrome (MS), the subgroup analyses reveal that Rosuvastatin still had impressive benefit in patients without MS, as well as in patients without ATP-III risk factors and those with Framingham risk scores < 10%. This suggests that even low risk patients with elevated CRP levels will benefit from statin therapy. (I agree that a 2x2 design would have been more helpful, but we have the data we have).

For this reason, I believe that this study will eventually open up routine testing of CRP in all patients over a certain age and statin treatment in patients whose levels are high. This will take time to develop as executive committee meetings need to occur. But I think the evidence here is clear.

AS, when i see a patient over 60 with CRP >4, I won't start a statin, without diagnosing the underlying inflammatory process. If it is not obesity related, it is more likely to signal connective tissue inflammation, infection or neoplasia. It would be difficult to decide what to do if CRP improves after treatment of the underlying cause of CRP elevation.

CIMT tells me if early or aggressive  statin rx is needed and if needs to be a lifetime and lifestyle project. A 3-5 year f/u of CIMT plaque characteristic and volume also tells me when I need more than statin fire power. 

Good points, Steven. You are shedding light on the fact that CRP is a nonspecific marker and that one needs to consider the larger picture first. I agree with this. There is no replacement for good clinician judgement with individual patients. It is also one of the joys of medicine and what likely got us into this in the first place.

But there is evidence that patients with diseases processes suchs as rheumatoid arthritis, HIV, etc that cause widespread inflammation and elevated CRPs are predisposed to early atherosclerotic disease. One question I have is would treating the underlying process first, ie with the standard armamentarium (DMARDS, methotrexate for RA, HAART for HIV) before prescribing statins decreases their risk of CAD? I don't know if that's been answered and I think is a relevant question.

But I think you are right to say that if you treat the underlying process first and CRP improves to normal then there is no reason to give a statin in the absence of hyperlipidemia or preexisting CAD.

We know rosuvastatin saved lives in JUPITER, but we don't know how. During the AHA press briefing, Dr. Ridker refused to tell CV death number, which is a component of the primary endpoint, saying that the trial's adjudication criteria is strict and misleading. Many statin trials failed to show survival advantages recently because we intervened early, thereby no reason to hide the number unless CV death in rosuvastatin arm is larger!

By the way, why you people ignore the fact that nearly 60% of the JUPITER participants is hypertention? Is it because of omission from the patient baseline characteristics table?

 

Impressive results with respect to relative risk reduction, but absolute risk reduction (ARR) is small.  However, the small ARR is not to be unexpected, given the relatively low risk primary prevention population evaluated.

I am not a believer in CRP as an independent risk factor that needs special attention, and do not think this trial changes that view, since we did not have a comparator group of similar clinical risk characteristics but with normal CRP.  

The conclusion that I draw from JUPITER is that CRESTOR is effective at CV risk reduction in a primary prevention setting.  This is consistent with my views on the utility of statins in primary prevention.  So CRESTOR may have a role to play here, but with the cost and small ARR, I will not be using it routinely in primary prevention.  I will tend to use statins somewhat more now in primary prevention, even though I already use them relatively aggressively in this setting, certainly for higher risk patients.  But I am not convinced that the results of JUPITER should be applied exclusively to patients with positive CRP, since this could still be an epiphenomenon.  And I do not argue with the results, and think that they can readily apply to patients with moderate atherthrombotic risk.

    

I have to admit the study is impressive.  However, it would be a big step, for me anyway, to give patients with relatively low LDL-C Crestor 20mg.  When I prescribe Crestor, I always start with 5mg and titrate to 10mg if needed.  I have one or two people, at most, on 20mg.  Yet, if we are to apply the results of this trial, we would have to use 20mg since we don't really have anything to measure.  We can't use LDL-C - that was already low.  We can't use CRP - we don't have any data on how to use it other than as a risk factor.

 For now I will continue using CRP for intermediate risk patients - especially those with strong family history.  I will continue to use the NCEP guidelines of getting LDL-c to goal and then focusing on non-HDL-c.  I fully agree with comments above that lifesyle changes need to be stressed constantly.

 Very imprressive study.

 Although the Absolute risk reduction was small... but it was a population of intemediate-to low risk in whom, probably none of us would do anything. So perhaps the study should make us to reconsider about it, measuring HS-CRP or something else...such us using a more accurate risk prediction model.

 Anyway, I think the study supports the hypothesis that hs-CRP is, no doubt, a risk marker, and probably, highly probably, a risk factor. I realize best study design to test whether CRP

I took part in JUPITER in my country, so I may be somewhat biased...

I don't think everybody should be screened for hsCRP.

The fact that about 5 patients were screened in JUPITER for one eventual participant shows the difficulties we faced in recruiting patients. We pre-sceened even more. It was not at all easy to find an elderly person without a prior CVD, no DM, no hypercholesterolemia and high hsCRP. But such patients do need hsCRP test. 1 event prevented in 120 relatively low risk patients in 2 years is not such a small advantage.

Two things still bother me - the extra DM reported, and the relatively short follow up. espaecially when it comes together.

Thanks to Dr. Avizohar for this unique perspective as one of the investigators.

I, too,  am troubled by the diabetes which has not been previously reported with a statin. And, as many of the prior comments reflect, most of the reduction in all cause mortality was related to less cancer deaths. That's not a bad thing, but not what would be expected.

It is terrific to see the varied and thoughtful processing of this data set and I appreciate all of those who take the time to write their comments.

 

I have read the many interesting comments and am in part surprised:

1. To me, JUPITER was a study testing the hypothesis that treating patients who have elevated hsCRP with aggressive statin therapy reduces Risk. High trigs, Low HDL, small Dense LDL have been shown to be some of the better markers of risk but even there, statins are the treatment option of choice. hsCRP can be now included in that category that defines the Vulnerable patient.

2.This was an event driven trial with expected low NNT's. Didn't see any vigorous disapproval of the study design all these years the Trial was in progress. Do we really need to spend a billion in multifactorial trials to test high vs low CRP, LDL, age below 30/50, short term/long term through a prospecive trial for NNT's of 100 to 200? Lets use our precious resources toward better adherence and implementation of whatever we truly believe.

3.Diabetes increase; a valid concern. However, unlike Ang-II antagonists which are know to reduce the progress to diabetes, no trial with a statin has suggested a similar impact. Hence pre-diabetics on a statin could potentially progress to diabetes. And then, would these patients not be better off on effective statin therapy? The RR of progression but the may be significant but AR!! By the way, many studies document increased progression to diabetes with all potent statins,  one only recently (Ref. Yamaka et al J  Atheroscler Thromb 2008: 15).,

4. Let's not forget the 'axis of evil'... oxLDL-AngII-LOX-1. A polypill may relieve the diabetes angst. AngII agonists are anti-inflammtory in the vasculature as well.

As a start siblings and first degree relatives need to determine their vulnerability to sudden death or MI; hsCRP has my vote. Treatment, take your pick among the most effective to get a 50% LDL lowering. Appreciate your comments on this view Dr. Topic .

Trials like Jupiter expand our knowledge in an important way but their incremental value for the well beeing of the patients is at best marginal. It has become increasingly difficult to find hypotheses that change the practise of medicine really significiantly (like thrombolysis, dPCI, betablockers in HF) so people test sort of virtual hypotheses.

Why don't we try to do better where benefits are obvious and proven rather than adding pills where they are higly unlikely to affect the course of the individual patient's life and possibly pose unknown risk? How many patients will be assigned to rosuvastatin because of randomly (i.e. falsely from this point of view) elevated CRP in the coming months/years? 

This is an interesting high cost population study that is difficult to translate into management of individual patient. Last but not least - with lower benefits , the risk/benefit ratio goes up.

 

 

 

 

 

 

 

 

 

 

 

 

 

1- I agree that the design is not the optimal one but it was done so we have to deal with. We just can't ask for another trial...

2- I will certainly not test hCRP in every patient. it would be foolish and not cost-effective. but in patients at high c-v risk I guess that hCRP can be proposed and if it is elevated why not follow jupiter? i guess that this is what I will do.

At first glance, I would also think that ASCOT would be a good study to look at when making comparisons, but since ASCOT and CARDS was performed prior to the inclusion of DM as a CAD risk equalivant they lose impact as a primary prevention trial. 

Also, when you look at the inclusion criteria of the ASCOT trial, the majority of the patients had 3 + risk factors for CAD, and enen then they were not very well controlled.  As a matter of fact, one could argue that they had one foot in the grave due to their medical histories upon entering the the trial.   Yet, even with the inclusion of Atorvastatin 10mg, there was a 30 to 35% realitive risk reduction for the most part with a 40+ for stroke.  The actual number escapes me at the moment. 

Again, these were diabetic patients, and as we now know that if they are not treated aggressively at the begining they will be lost.  Threrefore, one could suggest that eventhough ASCOT did present patients with a 30% to 35% realitive risk reduction, it could also be argued that ASCOT presented a 65% to 70% failure.  This line of thinking could also be used for prior primary prevention studies as well as secondary prevention studies.

This, the Jupiter trial, presents many questions.  But, one concept is constant within the statin class, and that is keeping LDL-C off the endothelium is a patient's best chance of preventing the first event and perhaps even surving the first event.   Therefore, in clinical practice one must decide upon what is the best choice for the given patient at that given moment during the treatment protocol.