Clinical Chemistry 47:3

403– 411 (2001) Review

High-Sensitivity C-Reactive Protein: A Novel and

Promising Marker of Coronary Heart Disease
Nader Rifai1,2,4* and Paul M. Ridker2,3,5

Background: Coronary heart disease remains the lead- dent predictor of future coronary events in apparently
ing cause of morbidity and mortality in the industri- healthy subjects.
alized world. Clinical and laboratory studies have © 2001 American Association for Clinical Chemistry
shown that inflammation plays a major role in the
initiation, progression, and destabilization of athero- Coronary heart disease (CHD)6 is the major cause of death
mas. C-Reactive protein (CRP), an acute phase reac- in the developed world. Atherosclerosis, the underlying
tant that reflects low-grade systemic inflammation, cause of most CHD, is a process that starts early in life and
has been studied in a variety of cardiovascular dis- progresses slowly and silently for decades. The clinical
eases. manifestation usually occurs in the form of myocardial
Approach: Findings from prospective clinical trials infarction (MI), stroke, angina, or sudden death between
were examined to determine the prognostic utility of ages 50 and 60 years in men and between 60 and 70 years
CRP in acute coronary syndromes, and observations in women. Cholesterol screening has been used as a tool
from epidemiological studies were reviewed to deter- to identify individuals who are at increased risk of devel-
mine the ability of CRP to predict future first coronary oping future coronary events. Although this approach has
events. The analytical considerations of CRP measure- been useful, it fails to identify almost one-half of the 1.3
ment in these clinical applications were also examined. million individuals who develop MI in the US each year
Content: In patients with established coronary disease, who have either normal or only moderately increased
CRP has been shown to predict adverse clinical serum cholesterol concentrations.
events. In addition, prospective studies have consis- Laboratory and clinical evidence has demonstrated
tently shown that CRP is a strong predictor of future that atherosclerosis is not simply a disease of lipid depos-
coronary events in apparently healthy men and its. Rather, systemic inflammation also plays a pivotal role
women. The relative risk associated with CRP is in atherothrombotic inception and progression (1–3 ).
independent of other cardiovascular disease risk fac- Mononuclear cells, macrophages, and T lymphocytes are
tors. High-sensitivity CRP (hs-CRP) assays are needed prominent in atheromatous plaques in the arterial wall
for risk assessment of cardiovascular disease. Such (4 –7 ). Furthermore, the shoulder region of a plaque, the
assays are currently available but may require further most vulnerable site for rupture in acute coronary syn-
standardization because patients’ results will be in- dromes, is heavily infiltrated with inflammatory cells
terpreted using population-based cutpoints. Preven- (8 –10 ). Cytokines, which cause the de novo hepatic
tive therapies to attenuate coronary risk in individu- production of acute phase reactants such as C-reactive
als with increased hs-CRP concentrations include protein (CRP) (11 ), have been shown to increase in acute
aspirin and statin-type drugs. coronary syndromes even in the absence of myocardial
Summary: hs-CRP has prognostic utility in patients necrosis (12 ). Therefore, CRP has been examined as a
with acute coronary syndromes and is a strong indepen- surrogate marker of other inflammatory mediators such
as interleukin-6 and tumor necrosis factor-! to better
understand the inflammatory component of atherosclero-
1
Department of Laboratory Medicine, Children’s Hospital, 2 Center for
Cardiovascular Disease Prevention, Divisions of Preventive Medicine, and
3 6
Cardiovascular Diseases, Brigham and Women’s Hospital, and Departments Nonstandard abbreviations: CHD, coronary heart disease; MI, myocar-
of 4 Pathology and 5 Medicine, Harvard Medical School, Boston, MA 02115. dial infraction; hs-CRP, high-sensitivity C-reactive protein; RR, relative risk;
*Address correspondence to this author at: Children’s Hospital, Depart- 95% CI, 95% confidence interval; CARE, Cholesterol and Recurrent Events;
ment of Laboratory Medicine, 300 Longwood Ave., Boston, MA 02115. Fax MRFIT, Multiple Risk Factors Intervention Trial; PHS, Physicians’ Health
617-713-4347; e-mail [email protected]. Study; PVD, peripheral vascular disease; WHS, Women’s Health Study; TC,
Received September 14, 2000; accepted December 15, 2000. total cholesterol; HDL-C and LDL-C, HDL- and LDL-cholesterol.

403
404 Rifai and Ridker: hs-CRP and CHD

sis (13, 14 ). Current knowledge, however, suggests that follow-up report by the same group using serum amy-
the CRP concentration might reflect the vulnerability of loid A, another acute phase reactant, instead of hs-CRP
the atheromatous lesion and the likelihood of a plaque to (31 ). A recent report by de Winter et al. (17 ) showed
rupture (2, 3, 15 ). This acute phase reactant has been that hs-CRP concentrations !5 mg/L at admission in
studied over the last several years in a wide variety of 150 patients with non-ST-elevation acute coronary syn-
atherosclerotic diseases (12, 16 –20 ). Its prognostic utility dromes were associated with an increased incidence of
in acute coronary syndromes (12, 16 –20 ) and its ability to major cardiac events within 6 months, regardless of
predict future coronary events in apparently healthy men cardiac troponin I values.
and women (21–30 ) have been demonstrated. The devel-
opment of high-sensitivity CRP (hs-CRP) assays has been hs-CRP as a Predictor of Future Coronary Events
instrumental in exploration of the role of this acute phase Over the last 6 years, several prospective studies have
reactant in predicting first cardiovascular events. Prospec- demonstrated that hs-CRP is a predictor of future cardio-
tive studies have consistently demonstrated a positive vascular morbidity and mortality among individuals with
association between hs-CRP and future coronary events. known cardiovascular disease. Data from the European
For hs-CRP to make the transition from clinical research Concerted Action on Thrombosis and Disabilities (ECAT)
to the routine clinical setting, however, several important Angina Pectoris Study Group, a study of 2121 men and
issues must be satisfactorily addressed: (a) the availability women with stable and unstable angina, demonstrated
of population-based cutpoints for interpretation and risk that each standard deviation increase in hs-CRP was
assessment; (b) the existence of potential therapeutic mo- associated with a 45% increase in the relative risk (RR) of
dalities; and (c) the reliability of the analytical systems nonfatal MI or sudden cardiac death [95% confidence
used for measurement. interval (95% CI), 1.15–1.83] (20 ). Similarly, in the Choles-
terol and Recurrent Events (CARE) trial, hs-CRP was a
hs-CRP as a Prognostic Indicator in predictor of recurrent coronary events in men and women
Acute Coronary Syndromes who had already suffered a MI (32 ). Those with hs-CRP
Several studies have demonstrated that hs-CRP, mea- concentrations in the highest quintile had an 80% higher
sured at either presentation or discharge, may have prog- chance of developing another coronary event within the
nostic value in patients with acute coronary syndromes. 5-year study period (RR # 1.77; 95% CI, 1.1–2.9). There-
Some reports have also examined the risk stratification of fore, hs-CRP has the potential to be used in the stratifica-
patients by hs-CRP alone or in combination with cardiac tion of patients into high- and low-risk groups.
troponins. Perhaps of greater clinical importance is the demon-
Liuzzo et al. (12 ) showed that in 31 patients with stration that hs-CRP concentrations predict first MI and
severe unstable angina and no evidence of myocardial stroke. To date, 10 prospective studies, 6 in the US and 4
necrosis, as documented by the absence of increased in Europe, have consistently shown that hs-CRP is a
cardiac troponin T, hs-CRP concentrations !3 mg/L at powerful predictor of future first coronary event in ap-
admission were associated with an increased incidence of parently healthy men and women (Fig. 1). Findings from
recurrent angina, coronary revascularization, MI, and the Multiple Risk Factors Intervention Trial (MRFIT)
cardiovascular death. The same group later demonstrated demonstrated a direct positive association between hs-
that hs-CRP !3 mg/L at discharge in 53 unstable angina CRP and CHD mortality in men followed over a 17-year
patients was associated with increased readmission for period (RR # 2.8; 95% CI, 1.4 –5.4) (22 ). This relationship,
recurrent instability and MI (16 ). In a similar study of however, was evident only among smokers. A similar
unstable angina, Ferreiros et al. (18 ) concluded that the association between hs-CRP and future coronary events
prognostic value of hs-CRP measured at discharge was was noted in the Cardiovascular Health Study and Rural
better than that determined at admission in predicting Health Promotion Project, which included men and
adverse outcome at 90 days. Furthermore, hs-CRP was the women over 65 years of age with subclinical cardiovas-
strongest independent predictor of adverse events in cular disease (26 ). The Physicians’ Health Study (PHS)
multivariate analysis. Data from the Thrombolysis In demonstrated similar positive association between hs-
Myocardial Infarction 11A (TIMI 11A), a study of unstable CRP and future coronary events in apparently healthy
angina and non-Q-wave MI, showed that markedly in- men (23 ). Unlike the observation in MRFIT, however, this
creased hs-CRP (15.5 mg/L) at presentation in 437 pa- association was evident in both smokers and nonsmokers.
tients was a good predictor of 14-day mortality in that This study showed that those in the highest quartile of
population (19 ). Furthermore, hs-CRP helped to identify hs-CRP had a twofold higher risk of future stroke (RR #
those patients with negative cardiac troponin T (qualita- 1.9; 95% CI, 1.1–3.3), threefold higher risk of future MI
tive rapid bedside method with cutoff of "0.2 "g/L) who (RR # 2.9; 95% CI, 1.8 – 4.6), and fourfold higher risk of
were at increased risk of mortality (19 ). Morrow et al. (19 ) future peripheral vascular disease (PVD; RR # 4.1; 95%
concluded from that study that a strategy for risk strati- CI, 1.2– 6.0) (23, 28 ). The RRs were stable over a long
fication using both cardiac troponin T and hs-CRP should period of time (#6 years) and independent of other CHD
be considered. Similar conclusions were reported in a risk factors. The European MONICA (Monitoring Trends
 Clinical Chemistry 47, No. 3, 2001 405

Fig. 1. Prospective studies of hs-CRP as a risk factor for future cardiovascular disease in populations of apparently healthy men and women.
RR estimates (f) and 95% CIs (lines) are computed for those in the top compared with the bottom quartile. Data from Refs. (21–30 ).

and Determinants in Cardiovascular Disease) Augsburg Predictive Value of hs-CRP and Other Biochemical Markers
study showed that an increase of one standard deviation for CHD Risk
in the log-transformed value of hs-CRP was associated The RR estimates derived from most of the above-men-
with a 50% increase in coronary risk and that subjects tioned prospective studies were independent of other
with hs-CRP concentrations in the highest quintile had a recognized cardiovascular risk factors. Data from both the
2.6-fold higher risk of developing future coronary events PHS (35 ) and WHS (24 ) showed that the predictive value
(21 ). A recent report from the Helsinki Heart Study of hs-CRP was significantly higher than that associated
confirmed these observations and demonstrated that with traditional biochemical CHD risk markers [total
those in the highest quartile of hs-CRP had a more than cholesterol (TC), HDL-cholesterol (HDL-C), and LDL-
threefold higher risk of future MI or cardiac death (RR # cholesterol (LDL-C)] or novel markers [lipoprotein(a),
3.56; 95% CI, 1.93– 6.57) (27 ). homocysteine, apolipoproteins AI and B]. In women, for
Two reports from the Women’s Health Study (WHS) example, the univariate RR of future cardiovascular
showed that hs-CRP is a strong predictor of future car- events presented in Fig. 2 demonstrate that hs-CRP was
diovascular events in women (RR # 4.4; 95% CI, 2.2– 8.9) the single strongest predictor of risk (RR # 4.4; 95% CI,
(24, 25 ). In stratified analyses, hs-CRP continued to be a 2.2– 8.9). In comparison, LDL-C, a well-established
strong predictor of future cardiovascular events even marker of CHD, was a lesser predictor of future risk
among subgroups of women with no history of hyperlip- (RR # 2.4; 95% CI, 1.3– 4.6). Furthermore, in a multivariate
idemia, hypertension, smoking, diabetes, or family his- analysis that accounted for other CHD risk factors (obe-
tory of CHD (25 ). The hs-CRP concentrations seen in sity, hypertension, diabetes, family history), only hs-CRP
these postmenopausal women were somewhat higher and the ratio of TC to HDL-C had independent predictive
than those reported previously in men. Although no value. In the same study of postmenopausal women (24 ),
difference in hs-CRP values was noted between premeno- hs-CRP was shown to predict risk among those with
pausal women and age-matched males, recent reports LDL-C values "1300 mg/L, a concentration deemed
showed that hormone replacement therapy (estrogen “desirable” by the current National Cholesterol Education
alone or estrogen and progestin) is associated with in- Program guidelines for primary prevention. In this sub-
creased hs-CRP concentrations (33, 34 ). These findings group (mean LDL-C, 1040 mg/L), the RRs of future MI,
suggest that the increased hs-CRP seen in the WHS stroke, and coronary revascularization in the lowest to the
subjects may reflect the influence of hormone replacement highest quartiles of hs-CRP were 1.0, 2.4, 2.9, and 4.1,
therapy rather than the effect of gender. respectively (95% CI for the 4th vs 1st quartile, 1.7–11.3).
406 Rifai and Ridker: hs-CRP and CHD

Fig. 2. RRs for future cardiovascular events among

apparently healthy women in the WHS according to
baseline values of several biochemical markers.
For consistency, risk estimates (f) and 95% CIs (lines) are
computed for those in the top compared with the bottom
quartile for each marker. Lp(a), lipoprotein(a); tHCY, total
homocysteine; IL-6, interleukin-6; sICAM-1, soluble intercel-
lular adhesion molecule-1; LDLC, LDL-C; SAA, serum amy-
loid A; Apo B, apolipoprotein B; HDLC, HDL-C. Adapted from
Ridker et al. (24 ).

After adjustment for other CHD risk factors and concen- that incorporated lipids were significantly better (P
tration of HDL-C, the RR associated with hs-CRP re- "0.001) than those based on hs-CRP alone (24 ).
mained highly significant (RR # 3.1; 95% CI, 1.1– 8.3) and
increased $39% with each increasing quartile of hs-CRP. Interpretation of hs-CRP Values
This study thus demonstrated that hs-CRP can identify For the purpose of assessing risk of future first coronary
individuals at increased risk of developing future coro- events, hs-CRP concentration should be interpreted using
nary events who otherwise would be missed if only lipid cut points established by prospective clinical studies. Each
measurements were used. Other examined markers of patient will be classified into a quintile of risk, depending
inflammation, e.g., serum amyloid A, interleukin-6, and on the hs-CRP concentration. Therefore, the reporting of
soluble intercellular adhesion molecule-1, showed consis- hs-CRP results focuses on the quintile of risk and not on
tent association but a slightly weaker RR of future coro- the actual mass concentration.
nary events. The within-person biologic variability of hs-CRP is low
The predictive value of hs-CRP in men and women over a long period of time (36 ). Laboratory measurements
increased considerably when evaluated in models that on paired samples obtained from 236 subjects at baseline
included lipid values. Data from the PHS demonstrated and 5 years later showed that an individual’s log-normal-
that, compared with those with TC and hs-CRP below the ized hs-CRP concentrations are highly correlated (r #
75th percentile, those with increased TC alone had a 0.60). Somewhat comparable correlation coefficients were
2.3-fold increase in risk (95% CI, 1.5–3.7), whereas those noted for TC (r # 0.37), LDL-C (r # 0.32), HDL-C (r #
with increased hs-CRP alone had a 1.5-fold increase in 0.74), and triglycerides (r # 0.49) over the 5-year fol-
risk (95% CI, 0.9 –2.4) (35 ). In contrast, the risk of devel- low-up period. This finding lends further support to the
oping coronary events increased 5-fold (95% CI, 2.5–9.8) fact that hs-CRP is a good and biologically stable predic-
among those with high concentrations of both TC and tor of future MI despite the fact that it is an acute phase
hs-CRP. Therefore, the joint effects of both risk factors are reactant, providing that the patient is not suffering from
greater than the product of the individual effects of each an active infection or using a drug that affects hs-CRP
risk factor considered alone. Furthermore, when the study concentration. hs-CRP values !15 mg/L ($99th percen-
participants were stratified according to quintile of hs- tile of the general population) indicate an active inflam-
CRP and quintile of TC:HDL-C ratio, the RR of first mation; patients should be advised to have a repeat
coronary event in those in the highest quintiles of both measurement in 2–3 weeks or after the infection in re-
hs-CRP and TC:HDL-C ratio was approximately ninefold solved.
higher than that of men in the lowest quintiles of these As indicated earlier, models containing both hs-CRP
analytes. Data from the WHS demonstrated similar find- and TC or the TC:HDL-C ratio were better able to predict
ings such that women in the highest quintile of both future first coronary events than those containing hs-CRP
hs-CRP and TC:HDL-C ratio had a RR more than eight- alone. The RRs of future first coronary events for men and
fold higher than that of women in the lowest quintiles women as well as lipid concentrations were computed in
(Fig. 3). In all of these analyses, risk prediction models quintiles from the PHS and WHS databases, respectively,
 Clinical Chemistry 47, No. 3, 2001 407

Fig. 3. RRs of first coronary event among apparently healthy men (left) and women (right) associated with different hs-CRP concentrations and
TC:HDL-C ratios.
Adapted from Ridker and co-workers (23, 24 ).

and are presented in Fig. 3. Because the computed RRs statin (32 ) are effective in decreasing the incidence of
did not vary significantly between men and women, a future coronary events in those with increased hs-CRP
single risk assessment algorithm is suggested for both concentration. These studies suggest that the two exam-
genders (Table 1) (37 ). The hs-CRP concentrations were ined drugs possess antiinflammatory characteristics.
derived from ongoing population-based surveys. It is Among apparently healthy men in the PHS with in-
important to note that it is not necessary in this case to creased hs-CRP (!2.1 mg/L), aspirin use decreased the
report the actual hs-CRP concentration to the clinician but risk of future MI by almost 60% (23 ). In contrast, aspirin
only the patient’s RR. The clinical laboratory should play use was associated with a much smaller, although statis-
an active role in the interpretation and implementation of tically significant, 14% decrease in future MI among men
this clinical application. Providing incomplete informa- with low hs-CRP ("0.55 mg/L). Although the magnitude
tion or just the actual hs-CRP concentration will only of reduction in future risk of MI depended on the concen-
frustrate and prevent the clinician from correctly inter- tration of hs-CRP, it is important to note that all subjects
preting the data and managing the patient. benefited from aspirin use. These findings suggest that
aspirin was acting not only as an antiplatelet agent but
Potential Preventive Therapies also as an antiinflammatory drug.
Although no specific therapies have been developed to Similar findings were also noted with pravastatin use
decrease hs-CRP and there is no direct evidence that risk in the CARE study (32 ). As indicated earlier, CARE is a
of future cardiovascular events is diminished by reducing prospective study of men and women with average lipid
hs-CRP, studies have shown that aspirin (23 ) and prava- concentrations who have suffered an MI. Participants

Table 1. RR estimates for future coronary events in men and women associated with quintiles of
hs-CRP and TC:HDL-C ratio.a
Quintile of hs-CRP, mg/L

1 2 3 4 5
Men Women (<0.7) (0.7–1.1) (1.2–1.9) (2.0–3.8) (3.9–15.0)
Quintile of TC:HDL-C ratio
1 <3.4 <3.4 1 1.2 1.4 1.7 2.2
2 3.4–4.0 3.4–4.1 1.4 1.7 2.1 2.5 3
3 4.1–4.7 4.2–4.7 2 2.5 2.9 3.5 4.2
4 4.8–5.5 4.8–5.8 2.9 3.5 4.2 5.1 6
5 >5.5 >5.8 4.2 5 6 7.2 8.7
a
RR estimates and TC:HDL-C ratio were derived from the PHS (23) and the WHS (24) databases. hs-CRP concentrations were derived from ongoing population-based
surveys.
408 Rifai and Ridker: hs-CRP and CHD

were randomized between 40 mg of pravastatin per day assays designed to detect active inflammation and infec-
and placebo and followed for 5 years (38 ). Study partici- tion. The dynamic range of these assays spans from 3
pants with high hs-CRP (!9.9 mg/L or 90th percentile) at mg/L ($90th percentile of the general population) to well
baseline experienced a reduction of 54% in the incidence over 200 mg/L. Such traditional assays, however, do not
of recurrent coronary events compared with a reduction have appropriate sensitivity in the range required for the
of 25% in those with low hs-CRP ("9.9 mg/L or 90th determination of cardiovascular risk in apparently
percentile), although baseline lipid values were almost healthy men and women.
identical in the two groups. Moreover, during the 5-year To achieve the desired limit of quantification, manu-
follow-up, pravastatin lowered mean hs-CRP by almost facturers and investigators have continued to use immu-
40%. This represented a 22% difference at 5 years in nochemical techniques in their attempts to measure hs-
median hs-CRP between the pravastatin and placebo CRP, but with modifications to increase the detectable
groups. Furthermore, the magnitude of change in hs-CRP signal. Several approaches have been used, including the
appeared to be unrelated to that of LDL-C in both the labeling of anti-CRP antibodies with either an enzyme
pravastatin and placebo groups. These findings suggest (ELISA) or a fluorescent compound, and attaching the
that pravastatin may have antiinflammatory characteris- antibodies, either monoclonal or polyclonal, to polysty-
tics that are independent from its lipid-lowering property. rene beads (50 –55 ). The latter approach was popular
Clinical trials are currently ongoing to further explore the among manufacturers because it enabled the adaptation
interaction between pravastatin, aspirin, and the inflam- to commonly used automated analyzers in clinical chem-
matory response in primary and secondary prevention istry laboratories. Currently, hs-CRP concentrations as
settings. low as 0.15 mg/L ("2.5th percentile of the general pop-
ulation) can be reliably measured. It is important to note,
Interrelationships with Other CHD Risk Factors however, that not all hs-CRP assays possess a similar
Several CHD risk factors appear to modulate the inflam- sensitivity or lower limit of quantification (56 ). For prac-
matory response and affect hs-CRP concentration. Obe- tical considerations, it is advisable to have a single CRP
sity, for example, is directly associated with increased assay in the clinical laboratory that is capable of measur-
hs-CRP concentrations, an intriguing observation consid- ing low and high concentrations. However, if that is
ering that interleukin-6, the primary stimulant of the de impossible, clinicians should be aware of the availability
novo hepatic synthesis of CRP, is secreted by adipose of two different CRP assays and request hs-CRP for
tissue (39, 40 ). Therefore, the attenuation of the inflamma- cardiovascular risk prediction purpose.
tory response may represent a mechanism by which diet Because the hs-CRP value of an individual patient is
and weight loss reduce cardiovascular risk. Cigarette interpreted in the context of cutpoints established by
smoking has also been shown to increase the concentra- prospective clinical studies, standardization of hs-CRP
tion of several inflammatory markers, including hs-CRP, assays is crucial. Poor agreement among methods will
interleukin-6, and soluble intercellular adhesion mole- lead to misclassification and mismanagement of patients.
cule-1. Increases of both interleukin-6 (41 ) and soluble Recent reports have indicated that the measurement of
intercellular adhesion molecule-1 (42 ) were shown to be low hs-CRP concentrations is not consistent among vari-
associated with increased risk of future first coronary ous methods, suggesting that standardization efforts are
events in both men and women. Smoking cessation de- needed (51, 56 ). In one case where a significant bias was
creases these markers. Diabetic patients are reported to noted between two methods (51 ), the manufacturers of
have increased hs-CRP values (43 ); In this regard, links both reagent systems claimed to have their calibrators
between hs-CRP and the insulin resistance syndrome traceable to the WHO reference materials. Unfortunately,
have also been reported (44 ). In addition, experimental this is not an unusual occurrence. Although manufactur-
findings suggest that increased blood pressure promotes ers attempt to standardize their assays using the WHO
endothelial expression of cytokines and inflammatory calibrators, they often fail to follow the appropriate value
activation (6, 45, 46 ). These observations suggest that per- transfer protocol from the reference materials to their own
haps better control of diabetes and hypertension may calibrators (57 ). Invariably, this leads to suboptimal stan-
attenuate the contribution of the inflammatory response dardization.
to overall cardiovascular risk. Finally, physical exercise An in-house hs-CRP ELISA method (52 ), utilizing
has been shown to have a beneficial effect in terms of polyclonal antibodies from Calbiochem, was used in
reducing the concentration of several inflammatory mark- MRFIT, the Cardiovascular Health Study, and the Rural
ers (47, 48 ). Taken together, the available evidence thus Health Promotion Project as well as in the early work
supports the hypothesis that hs-CRP concentrations cor- from the PHS. The analytical performance and clinical
relate with endothelial dysfunction (49 ). efficacy of the ELISA assay were compared with those of
an automated and commercially available latex-enhanced
Analytical Considerations in the Measurement of hs-CRP method (Dade Behring) (51 ) used at present in several
Historically, CRP has been measured in clinical laborato- prospective studies, including the PHS, WHS, Women’s
ries by immunoturbidimetric and immunonephelometric Health Initiative, Nurses’ Health Study, Health Profes-
 Clinical Chemistry 47, No. 3, 2001 409

sionals’ Study, and Texas/Air Force Coronary Atheroscle- 5. Navab M, Hama SY, Nguyen TB, Fogelman AM. Monocyte adhe-
rosis Prevention Study (58 ). The two assays were evalu- sion and transmigration in atherosclerosis. Coron Artery Dis
ated using plasma samples from the PVD cohort of the 1994;5:198 –204.
6. Nagel T, Resnick N, Atkinson WJ, Dewey CF Jr, Gimbrone MA Jr.
PHS in a nested case-control design. Excellent analytical
Shear stress selectively upregulates intercellular adhesion mole-
agreement between the two methods was reported cule-1 expression in cultured human vascular endothelial cells.
(slope # 0.99; intercept # 0.36 mg/L; r # 0.95) (58 ). In J Clin Invest 1994;94:885–91.
addition, for both methods, the calculated RRs of devel- 7. Nakashima Y, Raines EW, Plump AS, Breslow JL, Ross R. Upregu-
oping future PVD increased significantly with each in- lation of VCAM-1 and ICAM-1 at atherosclerosis-prone sites on the
creasing quartile of hs-CRP. The calculated interquartile endothelium in the ApoE-deficient mouse. Arterioscler Thromb
Vasc Biol 1998;18:842–51.
increase in RR of PVD was 31% (95% CI, 5.2– 62.2) for the
8. Moreno PR, Falk E, Palacios IF, Newell JB, Fuster V, Fallon JT.
ELISA and 34% (95% CI, 8.2– 66.1) for the latex-enhanced Macrophage infiltration in acute coronary syndromes. Implications
method. Furthermore, all but two participants were clas- for plaque rupture. Circulation 1994;90:775– 8.
sified into concordant quartiles or varied by only one 9. van der Wal AC, Becker AE, van der Loos CM, Das PK. Site of
quartile. This study demonstrated comparable clinical intimal rupture or erosion of thrombosed coronary atherosclerotic
efficacy of the two methods and linked the earlier and the plaques is characterized by an inflammatory process irrespective
of the dominant plaque morphology [see comments]. Circulation
current data, thus assuring consistency among reported
1994;89:36 – 44.
hs-CRP values. On the basis on this report, the US Food 10. Richardson PD, Davies MJ, Born GV. Influence of plaque configu-
and Drug Administration approved the use of this latex- ration and stress distribution on fissuring of coronary atheroscle-
enhanced method in the risk assessment of cardiovascular rotic plaques [see comments]. Lancet 1989;2:941– 4.
disease. Therefore, this latex-enhanced method is usually 11. Pepys MB, Baltz ML. Acute phase proteins with special reference
used as the reference procedure when comparison studies to C-reactive protein and related proteins (pentaxins) and serum
of various hs-CRP assays are conducted. At present, only amyloid A protein. Adv Immunol 1983;34:141–212.
12. Liuzzo G, Biasucci LM, Gallimore JR, Grillo RL, Rebuzzi AG, Pepys
a few hs-CRP methods are commercially available. How-
MB, Maseri A. The prognostic value of C-reactive protein and
ever, several assays are currently under development or serum amyloid a protein in severe unstable angina [see com-
evaluation and are expected to be available for routine ments]. N Engl J Med 1994;331:417–24.
clinical use in the very near future. 13. Rus HG, Vlaicu R, Niculescu F. Interleukin-6 and interleukin-8
protein and gene expression in human arterial atherosclerotic
conclusion wall. Atherosclerosis 1996;127:263–71.
14. Sukovich DA, Kauser K, Shirley FD, DelVecchio V, Halks-Miller M,
hs-CRP is a very promising novel biochemical marker for
Rubanyi GM. Expression of interleukin-6 in atherosclerotic lesions
the prediction of future first or recurrent coronary events. of male ApoE-knockout mice: inhibition by 17$-estradiol. Arterio-
American and European prospective studies have been scler Thromb Vasc Biol 1998;18:1498 –505.
highly consistent regarding the ability of hs-CRP to pre- 15. Maseri A. Inflammation, atherosclerosis, and ischemic events—
dict future CHD risk in both men and women. Potential exploring the hidden side of the moon [Editorial; comment]. N Engl
preventive therapeutic modalities to attenuate coronary J Med 1997;336:1014 – 6.
risk in those with increased hs-CRP concentrations have 16. Biasucci LM, Liuzzo G, Grillo RL, Caligiuri G, Rebuzzi AG, Buffon A,
et al. Elevated levels of C-reactive protein at discharge in patients
been suggested. The potential use of hs-CRP as a means to
with unstable angina predict recurrent instability. Circulation
improve the cost-to-benefit ratio in statin therapy is 1999;99:855– 60.
currently under investigation. Although standardization 17. de Winter RJ, Bholasingh R, Lijmer JG, Koster RW, Gorgels JP,
of hs-CRP measurement at the lower concentration range Schouten Y, et al. Independent prognostic value of C-reactive
among various methods should be addressed, a robust protein and troponin I in patients with unstable angina or non-Q-
and Food and Drug Administration-approved method is wave myocardial infarction. Cardiovasc Res 1999;42:240 –5.
18. Ferreiros ER, Boissonnet CP, Pizarro R, Merletti PF, Corrado G,
currently available. Several other sensitive assays are
Cagide A, Bazzino OO. Independent prognostic value of elevated
under development and are expected to be commercially C-reactive protein in unstable angina. Circulation 1999;100:
available soon. 1958 – 63.
19. Morrow DA, Rifai N, Antman EM, Weiner DL, McCabe CH,
Cannon CP, Braunwald E. C-Reactive protein is a potent predic-
References tor of mortality independently of and in combination with
1. Libby P, Ridker PM. Novel inflammatory markers of coronary risk: troponin T in acute coronary syndromes: a TIMI 11A substudy.
theory versus practice [Editorial; comment]. Circulation 1999; J Am Coll Cardiol 1998;31:1460 –5.
100:1148 –50. 20. Haverkate F, Thompson SG, Pyke SD, Gallimore JR, Pepys MB.
Production of C-reactive protein and risk of coronary events in
2. Ross R. Atherosclerosis—an inflammatory disease [see com-
stable and unstable angina. European Concerted Action on Throm-
ments]. N Engl J Med 1999;340:115–26.
bosis and Disabilities Angina Pectoris Study Group [see com-
3. Libby P. Molecular bases of the acute coronary syndromes. ments]. Lancet 1997;349:462– 6.
Circulation 1995;91:2844 –50. 21. Koenig W, Sund M, Frohlich M, Fischer HG, Lowel H, Doring A, et
4. Cybulsky MI, Gimbrone MA Jr. Endothelial expression of a mono- al. C-Reactive protein, a sensitive marker of inflammation, pre-
nuclear leukocyte adhesion molecule during atherogenesis. Sci- dicts future risk of coronary heart disease in initially healthy
ence 1991;251:788 –91. middle-aged men: results from the MONICA (Monitoring Trends
410 Rifai and Ridker: hs-CRP and CHD

and Determinants in Cardiovascular Disease) Augsburg Cohort 38. Sacks FM, Pfeffer MA, Moye LA, Rouleau JL, Rutherford JD, Cole
Study, 1984 to 1992. Circulation 1999;99:237– 42. TG, et al. The effect of pravastatin on coronary events after
22. Kuller LH, Tracy RP, Shaten J, Meilahn EN. Relation of C-reactive myocardial infarction in patients with average cholesterol levels.
protein and coronary heart disease in the MRFIT nested case- Cholesterol and Recurrent Events Trial investigators [see com-
control study. Multiple Risk Factor Intervention Trial. Am J Epide- ments]. N Engl J Med 1996;335:1001–9.
miol 1996;144:537– 47. 39. Visser M, Bouter LM, McQuillan GM, Wener MH, Harris TB.
23. Ridker PM, Cushman M, Stampfer MJ, Tracy RP, Hennekens CH. Elevated C-reactive protein levels in overweight and obese adults
Inflammation, aspirin, and the risk of cardiovascular disease in [see comments]. JAMA 1999;282:2131–5.
apparently healthy men [published erratum appears in N Engl 40. Yudkin JS, Stehouwer CD, Emeis JJ, Coppack SW. C-Reactive
J Med; see comments]. 1997;336:973–9. protein in healthy subjects: associations with obesity, insulin
24. Ridker PM, Hennekens CH, Buring JE, Rifai N. C-Reactive protein resistance, and endothelial dysfunction: a potential role for cyto-
and other markers of inflammation in the prediction of cardiovas- kines originating from adipose tissue? Arterioscler Thromb Vasc
cular disease in women. N Engl J Med 2000;342:836 – 43. Biol 1999;19:972– 8.
25. Ridker PM, Buring JE, Shih J, Matias M, Hennekens CH. Prospec- 41. Ridker PM, Rifai N, Stampfer MJ, Hennekens CH, Buring JE.
tive study of C-reactive protein and the risk of future cardiovascu- Plasma concentration of interleukin-6 and the risk of future
lar events among apparently healthy women [see comments]. myocardial infarction among apparently healthy men. Circulation
Circulation 1998;98:731–3. 2000;101:1767–72.
26. Tracy RP, Lemaitre RN, Psaty BM, Ives DG, Evans RW, Cushman 42. Ridker PM, Hennekens CH, Roitman-Johnson B, Stampfer MJ,
M, et al. Relationship of C-reactive protein to risk of cardiovascular Allen J. Plasma concentration of soluble intercellular adhesion
disease in the elderly. Results from the Cardiovascular Health molecule 1 and risks of future myocardial infarction in apparently
Study and the Rural Health Promotion Project. Arterioscler Thromb healthy men [see comments]. Lancet 1998;351:88 –92.
Vasc Biol 1997;17:1121–7. 43. Ford ES. Body mass index, diabetes, and C-reactive protein
27. Roivainen M, Viik-Kajander M, Palosuo T, Toivanen P, Leinonen M, among U.S. adults [see comments]. Diabetes Care 1999;22:
Saikku P, et al. Infections, inflammation, and the risk of coronary 1971–7.
heart disease. Circulation 2000;101:252–7. 44. Festa A, D’Agostino R Jr, Howard G, Mykkanen L, Tracy RP,
28. Ridker PM, Cushman M, Stampfer MJ, Tracy RP, Hennekens CH. Haffner SM. Chronic subclinical inflammation as part of the insulin
Plasma concentration of C-reactive protein and risk of developing resistance syndrome: the Insulin Resistance Atherosclerosis
peripheral vascular disease [see comments]. Circulation 1998; Study (IRAS). Circulation 2000;102:42–7.
97:425– 8.
45. Liu Y, Liu T, McCarron RM, Spatz M, Feuerstein G, Hallenbeck JM,
29. Danesh J, Whincup P, Walker M, Lennon L, Thomson A, Appleby P,
Siren AL. Evidence for activation of endothelium and monocytes in
et al. Low grade inflammation and coronary heart disease: pro-
hypertensive rats. Am J Physiol 1996;270:H2125–31.
spective study and updated meta-analyses [see comments]. BMJ
46. McCarron RM, Wang L, Siren AL, Spatz M, Hallenbeck JM.
2000;321:199 –204.
Monocyte adhesion to cerebromicrovascular endothelial cells
30. Mendall MA, Strachan DP, Butland BK, Ballam L, Morris J,
derived from hypertensive and normotensive rats. Am J Physiol
Sweetnam PM, Elwood PC. C-Reactive protein: relation to total
1994;267:H2491–7.
mortality, cardiovascular mortality and cardiovascular risk factors
47. Manson JE, Hu FB, Rich-Edwards JW, Colditz GA, Stampfer MJ,
in men. Eur Heart J 2000;21:1584 –90.
Willett WC, et al. A prospective study of walking as compared with
31. Morrow DA, Rifai N, Antman EM, Weiner DL, McCabe CH, Cannon
vigorous exercise in the prevention of coronary heart disease in
CP, Braunwald E. Serum amyloid A predicts early mortality in acute
women. N Engl J Med 1999;341:650 – 8.
coronary syndromes: A TIMI 11A substudy. J Am Coll Cardiol
2000;35:358 – 62. 48. Smith JK, Dykes R, Douglas JE, Krishnaswamy G, Berk S. Long-
term exercise and atherogenic activity of blood mononuclear cells
32. Ridker PM, Rifai N, Pfeffer MA, Sacks FM, Moye LA, Goldman S, et
in persons at risk of developing ischemic heart disease. JAMA
al. Inflammation, pravastatin, and the risk of coronary events after
1999;281:1722–7.
myocardial infarction in patients with average cholesterol levels.
Cholesterol and Recurrent Events (CARE) Investigators. Circula- 49. Fichtlscherer S, Rosenberger G, Walter DH, Breuer S, Dimmeler S,
tion 1998;98:839 – 44. Zeiher AM. Elevated C-reactive protein levels and impaired endo-
33. Ridker PM, Hennekens CH, Rifai N, Buring JE, Manson JE. thelial vasoreactivity in patients with coronary artery disease.
Hormone replacement therapy and increased plasma concentra- Circulation 2000;102:1000 – 6.
tion of C-reactive protein. Circulation 1999;100:713– 6. 50. Wilkins J, Gallimore JR, Moore EG, Pepys MB. Rapid automated
34. Cushman M, Legault C, Barrett-Connor E, Stefanick ML, high sensitivity enzyme immunoassay of C-reactive protein. Clin
Kessler C, Judd HL, et al. Effect of postmenopausal hormones Chem 1998;44:1358 – 61.
on inflammation-sensitive proteins: the Postmenopausal Estro- 51. Ledue TB, Weiner DL, Sipe JD, Poulin SE, Collins MF, Rifai N.
gen/Progestin Interventions (PEPI) Study. Circulation 1999; Analytical evaluation of particle-enhanced immunonephelometric
100:717–22. assays for C-reactive protein, serum amyloid A and mannose-
35. Ridker PM, Glynn RJ, Hennekens CH. C-Reactive protein adds to binding protein in human serum. Ann Clin Biochem 1998;35:745–
the predictive value of total and HDL cholesterol in determining 53.
risk of first myocardial infarction [see comments]. Circulation 52. Macy EM, Hayes TE, Tracy RP. Variability in the measurement of
1998;97:2007–11. C-reactive protein in healthy subjects: implications for reference
36. Ridker PM, Rifai N, Pfeffer MA, Sacks F, Braunwald E. Long-term intervals and epidemiological applications. Clin Chem 1997;43:
effects of pravastatin on plasma concentration of C-reactive 52– 8.
protein. The Cholesterol and Recurrent Events (CARE) Investiga- 53. Kapyaho K, Welin MG, Tanner P, Karkkainen T, Weber T. Rapid
tors. Circulation 1999;100:230 –5. determination of C-reactive protein by enzyme immunoassay using
37. Rifai N, Ridker PM. Proposed cardiovascular risk assessment two monoclonal antibodies. Scand J Clin Lab Invest 1989;49:
algorithm using high-sensitivity C-reactive protein and lipid screen- 389 –93.
ing [Opinion]. Clin Chem 2001;47:28 –30. 54. Eda S, Kaufmann J, Molwitz M, Vorberg E. A new method of
 Clinical Chemistry 47, No. 3, 2001 411

measuring C-reactive protein, with a low limit of detection, suit- implications for clinical and epidemiological applications. Clin
able for risk assessment of coronary heart disease. Scand J Clin Chem 2000;46:461– 8.
Lab Invest Suppl 1999;230:32–5. 57. Johnson A, Sampson E, Blirup-Jensen S, Scvendsen P. Recom-
55. Borque L, Bellod L, Rus A, Seco ML, Galisteo-Gonzalez F. Devel- mendations for the selection and use of protocols for assignment
opment and validation of an automated and ultrasensitive immu- of values to reference materials. Eur J Clin Chem Clin Biochem
noturbidimetric assay for C-reactive protein. Clin Chem 2000;46: 1996;34:279 – 85.
1839 – 42. 58. Rifai N, Tracy RP, Ridker PM. Clinical efficacy of an automated
56. Roberts WL, Sedrick R, Moulton L, Spencer A, Rifai N. Evaluation high-sensitivity C-reactive protein assay. Clin Chem 1999;45:
of four automated high-sensitivity C-reactive protein methods: 2136 – 41.