Our apologies for the length of this post, but JUPITER and its impact on approaches to cardiovascular risk evaluation are complex and important to us all. We have tried to be clear and brief as possible.
Background: Heart attacks and strokes are the major killers of people in mid-life and later. We know that high cholesterol (low-density lipoprotein cholesterol, or LDL, the “bad” cholesterol), diabetes, high blood pressure, and cigarette smoking make such cardiovascular events much more likely. For several decades we have been treating high cholesterol with statin medications (such as simvastatin, Lipitor and, more recently, Crestor) in addition to persistent attempts to improve patients’ diets and increase their exercise.Such efforts, especially the intensive therapy with statins to lower the LDL, have been highly successful in reducing the incidence of new heart attacks and strokes.
However, a large proportion of heart attacks and strokes (estimates range up to half) occur in people whose LDL cholesterol is below the current healthy target of 130 mg/dL, and even below 100; in other words, clearly normal.Extensive research has been directed to figuring out what mechanism produces these heart attacks and strokes in people with normal cholesterol. One important line of investigation focuses on chronic low-grade inflammation in the body that adversely impacts the lining of the arteries, called the endothelium. The damage to the endothelium then sets the stage for the heart attack or stroke. High cholesterol may increase that damage, or possibly directly generate some of it, but high cholesterol may not always be an essential component.
About a decade ago, Dr. Paul M. Ridker at the Brigham and Women’s Hospital began studying an old blood test for inflammation, called the C-reactive protein. This test goes up in a wide variety of inflammatory conditions, such as arthritis, colitis, and appendicitis. He demonstrated that a high-sensitivity version of that test (“high-sensitivity CRP” or “HS-CRP”) could distinguish different levels of background inflammation within the normal range of the regular C-reactive protein. Moreover, the higher background levels of C-reactive protein appeared to correlate with more heart disease, independent of the LDL-cholesterol level.It was quickly noted that statin medications lowered HS-CRP as well as LDL-cholesterol and it is well established that statin treatment for high LDL-cholesterol substantially reduces heart attacks and strokes and overall cardiovascular mortality. What was unknown was whether the use of statins to treat apparently healthy patients with normal LDL-cholesterol but elevated HS-CRP would also result in reduced cardiovascular morbidity and mortality. The JUPITER study was designed to answer that question. The answer appears to be “yes.” Details follow
The JUPITER Study: The formal title of the study, “Justification for the Use of Statins in Prevention: An Intervention Trial Evaluating Rosuvastatin,” became the acronym “JUPITER.” The results were just published in the New England Journal of Medicine Vol. 359, No. 21, on November 20, 2008, under lead author Paul M. Ridker.Patients enrolled included men over age 50 and women over age 60, apparently healthy, with LDL-cholesterols lower than 130. Diabetics were excluded, as were people with poorly controlled hypertension, but not persons with prediabetic conditions. Chronic inflammatory conditions (such as arthritis or colitis) were excluded. Smoking was allowed. All patients had to have at least moderately elevated levels of HS-CRP (over 2 mg/L) to qualify (the range is from 0.1 to 5.0 or so for this test).
About 90,000 people were screened by these criteria in sites around the world, with 80% excluded predominantly because the LDL-cholesterol was too high (over 130) or the HS-CRP too low (under 2.0). Women were about 38% of the total, smokers about 16%. Average starting LDL was 108. Average initial HS-CRP was 4.2. Blood sugars averaged 94 and A1c 5.7% (both normal values). Blood pressures were 134/80 on average, slightly high. Median age was 66, with half the subjects between 60 and 71.
After randomization, half the patients were treated with Crestor (rosuvastatin) 20 mg, a somewhat newer statin that is a bit stronger than Lipitor (atorvastatin), the others with placebo. Blood tests were followed. The outcomes were defined as various major cardiovascular events or death. The study was supposed to follow the patients for 5 years, but it was terminated after a median of 1.9 years because the results were said to be so strikingly beneficial.
Key Jupiter Results: There were 142 first major cardiovascular events in the Crestor group and 251 in the placebo group, or about 7.7 per 1000 v. 13.6 per 1000 person-years of follow-up, meaning a “hazard ratio” of 0.56 for the treatment group. This is a substantial 44% reduction. However, given that everyone was pretty healthy, the number needed to treat (NNT) to save one event was estimated in various fashions (always subject to argument) and ranged from 95 persons down to 25-30 with different assumptions. At close to $2,000 annually for Crestor and expenses of monitoring, the cost of preventing one event would be fairly pricey for widespread use.
The treated group’s LDL’s averaged 55 (with half between 44 to 72) mg/dL, while the HS-CRP reduced to a median 2.2 mg/L from the starting value of 4.3. Interestingly, the placebo group’s HS-CRP also reduced, to 3.3 mg/dL, which was not commented on. Nor was the slightly higher incidence of miscellaneous adverse effects (1377 v. 1352) in the placebo group.
Most subgroups were said to show the same reduced risk, more or less, independent of age, smoking and so forth. However, those with a strong family history of heart disease appeared to have benefitted slightly more from Crestor than everyone else, while more cases of diabetes were reported in the treatment group compared to the placebo group.
JUPITER Recap: JUPITER demonstrates that, in a relatively low-risk group of older men and women, treatment of elevated HS-CRP values with a powerful statin (Crestor), in the absence of elevated LDL-cholesterol, substantially lowers both the CRP and the LDL and approximately halves the incidence of cardiovascular events and deaths.
How Does JUPITER Relate to OHC Members?: How does this study impact our approach to cardiovascular risk assessment for our patients?
Most Orchard Health Care members over age 50 have been already been tested for HS-CRP in the past several years, in addition to routine cholesterol and diabetic testing. For the most part, the HS-CRP gave results similar to the cholesterol value. If the latter was high, the CRP wasn’t needed to decide on statin treatment. When the cholesterol was low, most of you also had low HS-CRP values, below the treatment threshold of 2 mg/L used in JUPITER.
But a few of you probably had elevated HS-CRP with low LDL-cholesterols, as in the study group. Before JUPITER, we did not think there was sufficient data to push you to take a statin medicine without any other indication. Now we have to reconsider that decision.
We also have to consider measures other than HS-CRP and LDL cholesterol in your cardiovascular risk evaluation. For example, the JUPITER study offered no data on another promising measure of vascular risk, the cardiac calcium score. This is the non-contrast CT exam that detects and quantifies calcification in the coronary arteries themselves, a marker of the results of chronic inflammatory and atherosclerotic changes in the arteries that appears as well to predict coronary events. Many of you have had this test upon our recommendation. Its predictive value in low-risk patients (comparable to the JUPITER trial) has been demonstrated; one such study was posted on this website early this year (Cardiac Calcium Score Helps Define Women’s Cardiac Risk, Feb 28, 2008). We have much to sort out to give you the best available counsel.
What Will We Do Differently?: Going forward, we will plan to add a HS-CRP to periodic blood testing for each of you over 50 or so or at a younger age if you have an important family risk of cardiovascular disease. We will also, at your office visits, review the data we have on each of you for recent years (about five) since the HS-CRP has been reasonably reliable as a test. And we will certainly consider the HS-CRP result in our overall evaluation of what your cardiovascular risk is and whether any intervention is needed, statins or otherwise. Ultimately, we may suggest you take a statin medication if only your HS-CRP is elevated, which would not have been the case before JUPITER.
More Uncertainties for Us to Ponder: We must remember, above all, that the body is complex and that we are not always as smart as we think we are. Long-term outcomes (much longer than JUPITER’s 1.9 year median follow-up) are required to be secure about benefits and safety. We simply don’t have 20-year data following patients whose LDL-cholesterols were driven down to 50 mg/dL, the level in this study. Could it be harmful? Cholesterol is the building block of the cell wall structure as well as all of the steroid and sex hormones in our bodies. Are exceedingly low levels safe? The study was done with Crestor, a newer and expensive statin. Would generic simvastatin, well studied and much cheaper, work as well? How would we know? And what about the 25% rise in reported cases of diabetes in the treatment group, despite no obvious change in the diabetic blood tests? Do we approach treating a person of 70 years of age the same as one who is 51? Facing uncertainties and unanswerables, caution and careful thought and discussion are truly required. This must be an interactive process between physician and patient.
Last, no one should think that the HS-CRP test is the last word in detection or prevention or that it truly gives us complete knowledge about the processes of atherogenesis that ruin our arteries to produce heart attacks and strokes. Almost undoubtedly there are other mechanisms of blood vessel disease to be deciphered and there will be other tests developed.
To illustrate, let me cite the case of one of our members, who is over a bit over 50, healthy, with no risk factors, with a low LDL-cholesterol (90) and an extremely low HS-CRP. Yet a cardiac calcium score test done out of curiosity was extraordinarily high, and follow-up testing demonstrated widespread coronary artery disease (atherosclerosis). So all our current testing failed to detect the lurking but widespread vascular disease.
The Moral? We have much more to learn about how the human body really works. Certainly, in our decision-making and recommendations to you, we attempt to integrate all reliable new information, but we had better retain a healthy skepticism about how much we really know so that we proceed with caution and intellectual humility.