January 28th, 2011
Does CRP Level Modify the Benefit of Statins? Paul Ridker Reacts to New Data
CardioExchange welcomes Paul M. Ridker, a leading researcher on the value of C-reactive protein (CRP) concentration as a prognostic marker of cardiovascular risk, to respond to the latest data on CRP and statin therapy from the Heart Protection Study (HPS).
The New HPS Findings on CRP: The HPS investigators sought to determine whether statin therapy might be especially beneficial to patients with high CRP concentrations. They took the more than 20,000 HPS participants who had been randomized to receive simvastatin (40 mg/day) or placebo for a mean of 5 years and stratified them into 6 groups according to baseline CRP level, ranging from <1.25 mg/L to ≥8 mg/L.
The incidence of major vascular events — coronary death, MI, stroke, or revascularization — was significantly lower among simvastatin recipients than among placebo recipients (19.8% vs. 25.2%). The investigators found no evidence that incidence of that endpoint or its components varied according to baseline CRP concentration. Nor did relative risk reduction vary across four subgroups that were defined by high or low baseline levels of LDL cholesterol and CRP. Even patients with low baseline levels of both LDL-c and CRP benefited significantly from simvastatin.
Dr. Ridker’s Response:
First, although the HPS investigators report that baseline CRP does not modify the vascular benefits of statin therapy, it is important to remember that baseline LDL-c also did not modify the benefit of statin therapy in their study.
Second, the data presented clearly show that CRP levels track directly with absolute risk in the HPS. For example, within the placebo group, the baseline CRP level corresponded with the incidence of vascular events: 19% incidence among patients with low CRP, climbing to 31% among those with high CRP. Thus, although these data are not discussed at all in the article, they nonetheless confirm the ability of CRP to predict high risk.
Third, the HPS investigators elected not to address the core hypothesis that on-treatment levels of CRP correlate with the risk reduction attributable to statins, as has been shown in many prior statin trials. The authors make the claim that such an analysis should not be done, as it would be nonrandomized. Yet they go on to conclude that the benefits of statin therapy accrue solely to on-treatment LDL, an analysis that suffers equally from the same nonrandomized limitations.
Fourth, the well-conducted HPS was performed almost entirely among patients with a prior history of MI, stroke, peripheral artery disease, or diabetes. Therefore, the HPS cannot be interpreted as a primary-prevention trial. Primary prevention is the setting in which (1) CRP measurement has been shown to provide as much incremental information on vascular risk as does total or HDL cholesterol and (2) CRP testing is recommended to identify patient groups who benefit from statin therapy but otherwise would not qualify for treatment on the basis of LDL-c levels.
In the end, the only way to test the inflammatory hypothesis of atheroscerlosis is to directly randomize patients to targeted anti-inflammatory therapies. My research team and I hope to launch two such trials this year.