CRP or not CRP? That Is the Question

Abstract

C-reactive protein (CRP) and cardiovascular disease is a very hot topic at present. Excitement and interest have spilled over dramatically from the scientific literature into the media and popular press leading to much speculative comment. However, rigorously controlled and reproducible studies are now laying the basis for a more realistic consensus. The article in this issue of Arteriosclerosis, Thrombosis, and Vascular Biology from Carmen van den Berg’s laboratory makes a notable contribution to this process. 1 See page 1225 Shortly after its discovery in 1929 as “the acute phase protein,” increased production of CRP was recognized as a characteristic feature of the response to acute myocardial infarction, and until the 1960’s detection of CRP was widely used for routine monitoring of acute rheumatic fever. In 1961 Irving Kushner, in one of the first applications of the then new technique of immunofluorescence, demonstrated the deposition of rabbit CRP in experimental myocardial infarction lesions. The association between CRP and cardiovascular disease thus has a very long history. In 1979 Kushner reported the kinetics of the acute phase CRP response to human acute myocardial infarction, and soon afterward we first investigated critically the behavior of CRP in clinical coronary artery disease and myocardial infarction.2 We observed that persistently elevated circulating CRP concentrations after an infarct were associated with a poor prognosis, although at that time we were focusing mainly on CRP as a sensitive nonspecific marker of all the various complications of coronary occlusion and their treatment. The current phase of interest in CRP and cardiovascular disease started in the early 1990’s with observations of increased CRP concentrations in some patients with “active coronary syndromes” and some individuals with acute myocardial infarction tested very soon after onset of pain, before the acute phase response to infarction could have started. These findings were consistent with the growing recognition at that time of the importance of inflammation in atherogenesis and the inflammatory nature of unstable atherosclerotic lesions. At the same time the large European prospective study (ECAT) of coagulation factors as possible prognostic markers in outpatients with stable and unstable angina unexpectedly revealed that the baseline value of CRP, which had only been included as a control for the acute phase behavior of some clotting proteins, significantly predicted future coronary events.3 The assay used was insufficiently sensitive to detect CRP in many of the samples in the ECAT study, and we therefore developed the first automated high-sensitivity method to re-assay their 3000 samples. Separately we also used this method to study, in collaboration with the Maseri group, carefully characterized patients with severe unstable angina who had not yet experienced any myocardial necrosis. We selected a cut point of 3 mg/L as the upper limit of normal, based on our original 1981 study of 468 healthy volunteer blood donors in whom this was the 90th centile of the CRP distribution. 4 In both the ECAT and the unstable angina studies, increased baseline CRP values were associated with significantly increased risk of future coronary events. 5,6 Subsequently several groups showed independently that baseline CRP measurements are associated with future coronary events in general populations, without known preexisting coronary artery disease. 7–10 In 2000 we reported with Danesh and colleagues the results from the British Regional Heart Study comprising 506 coronary events, 2 as many as in any other such study up to that time, together with a meta-analysis of all previous studies.11 This showed that the relative risk of having a coronary event among individuals with a baseline CRP value in the upper compared with the lower tertile of the CRP distribution was 2.0. In 2004 the Reykjavik Study, comprising 2459 coronary events, showed this relative risk to be 1.45 (95% CI, 1.25 to 168), and meta-analysis of all previously published general populations studies, comprising 7068 patients with coronary events, gave a similar result. 12 The results from this now quite rigorously