August 26th, 2011

The Power of Zero on JUPITER

This voluntary contribution from Dr. Nasir, the principal investigator of a recent study published in the Lancet, is an extension of his remarks in a recent panel discussion

Background: The study showed that people with low LDL levels and high C-reactive protein levels may benefit from coronary artery calcium scans to identify the folks who are most likely to benefit from statin therapy. Data came from 950 people enrolled in the Multi-Ethnic Study of Atherosclerosis (MESA) who had met the entry criteria for the JUPITER study.

What was the aim of our Lancet study?

I believe, at least in the news media discussion, the focus has been more on the comparative role of the predictive value of CAC and CRP for cardiovascular events. Although part of the subanalysis and our findings again demonstrate the superior predictive value of CAC versus CRP, it is important to keep in mind that the results apply only to those individuals with LDL <130 mg/dL.

The real aim of the study was to assess this question: “Is everyone meeting JUPTER criteria at the same risk and will all benefit similarly with aggressive lipid-lowering pharmacotherapy, or can we selectively identify those likely to benefit more (e.g., those who have the highest risk for events). More important, can we identify a group at such a low risk that they are unlikely to derive significant benefit from aggressive pharmacotherapy on the basis of absence of CAC?”

Why is the predictive value of CAC more robust than that of any other marker? What is the evidence about the value of CAC=0?

While the presence of CAC represents actual disease, risk factors such as cholesterol levels or hsCRP represent only one of many factors that may (or may not) ultimately cause coronary heart disease. As a result in primary prevention settings, the absence of CAC features as among the most powerful “negative indicators” for development of a coronary event. This assumption is really based on our group’s extensive work showing that absence of CAC confers a very low risk for future CVD events and mortality in a large meta-analysis (JACC Cardiovasc Imaging 2009; 2:675), the findings of which were subsequently confirmed in a large retrospective study (JACC Cardiovasc Imaging 2009; 2:692) and a multiethnic prospective study (Am Heart J 2009; 158:554).

What are the implications of our study findings? Treat more or less?

The implications of our study are in response to critics who assert that CAC testing will lead to more downstream costs, whereas we have shown the opposite — that it can actually be used as a tool to withhold aggressive management as well any further testing in nearly half of the 6.5 million individuals who are now candidates for long-term statin therapy based on JUPITER criteria. Such individuals should still be counseled regarding aggressive lifestyle interventions; however, it is clear that they will be less likely to benefit from aggressive (and costly) pharmacologic interventions.

In my view, the strongest role of CAC testing is in its “power of zero,” which is a relatively novel concept and critical in this era of rising costs, polypharmacy, etc. Since hsCRP and almost all other nonspecific markers cannot really rule out disease, they can be used to raise risk estimates and, thus, are inextricably tied to more treatment and downstream cost.

Can we extrapolate from the JUPITER trial and our study that those with higher CAC benefit from treatment?

One often-raised criticism is that no RCT has shown that treating those with higher CAC will improve outcomes. I would like to point to some interesting observations, especially if we limit our discussion to the JUPITER population.

  1. JUPITER has clearly shown a benefit of aggressive lowering, even among patients with normal LDL but elevated hsCRP, to reduce the risk for future CVD events within a follow-up of nearly 2.5 years.
  2. In our study, we have observed that among those meeting JUPITER criteria, 47% had no CAC and the event was minimal over the next 5.8 years (almost twice the follow-up of the JUPITER trial).
  3. It is very clear that those with CAC=0 are highly unlikely to benefit from very aggressive therapy, even if we assume a 54% event reduction with statin therapy (much greater than reported within JUPITER), as the NNT5 still remains very high. It’s very simple: You need events to eventually intervene and reduce them.
  4. This finding naturally then leads us to ask: If those with absent CAC within the JUPITER population are not the ones likely benefiting, then which subgroup is achieving the maximum risk reduction with statin therapy, as shown in this landmark trial by Ridker et al. (N Engl J Med 2008; 359:2195)?
  5. Although it will continue be an educated guess that we cannot prove (as CAC testing was not done in the JUPITER study), I believe, in light of my above arguments, that the reduction in events seen within JUPITER was very likely in those with present and varying degrees of CAC.

Are these results enough to change practice?

At least in my practice, I would not hesitate to consider CAC testing among patients like those in the JUPITER population, to withhold treatment from those with CAC=0, and to continue to focus on lifestyle-modification strategies. I agree that good points have been raised by many sides. Although an RCT has the ability to provide the most convincing data, if funded, such a trial would take close to 10 years to complete. Furthermore, I would like to see verification of our findings in other cohorts.

However, for now, we cannot ignore the current data. As highlighted by Dr. Kaul in our previous panel discussion, the main unanswered question is whether to advocate selective therapeutic targeting or unconditional treatment of all those meeting JUPITER criteria. While neither approach is “wrong,” we must continue to study this area and advance the dialogue among all pertinent stakeholders about how best to answer this question. The status quo is definitely not an option.

One Response to “The Power of Zero on JUPITER”

  1. David Powell , MD, FACC says:

    Is it necesary to choose a subset from JUPITER type patients to treat? If so, I assume it is a cost issue. The FDA limited approval to those with another risk factor, using a posthoc subgroup analysis. This reminds me of the old ICD limitation to patients with a wide QRS. Interestingly, Dr. Ridker’s press comments suggested that he found this limitation reasonable. I find the non-zero CAC group a more compelling choice as per the above argument. But from a cost and clinical perspective, should the CAC score be repeated every 4 years? Should the CRP be repeated in 4 years (risking regression to the mean)?