Ten years ago, cardiologists wouldn’t have ordered tests for C-reactive protein for patients with suspected cardiovascular disease. But results of the landmark Justification for the Use of Statins in Prevention: an Intervention Trial Evaluating Rosuvastatin (JUPITER) study, released in November, have confirmed the role of CRP as a biomarker of risk for cardiovascular disease and established it as a means of monitoring the impact of cholesterol-lowering therapy, not only in people with known risks, but also in asymptomatic individuals previously considered at average risk for myocardial infarction, stroke, or death from cardiovascular causes (N Engl J Med 2008; 359:2195–2207). The study, which made headline news, underscores the importance of inflammation in cardiovascular disease and could result in changes to recommended cardio-protective practices and patient management.
“This is not just a regular clinical trial. It’s a major turning point in preventive cardiology; one of the most significant developments since the inception of the National Cholesterol Education Program,” noted Nader Rifai, PhD, director of clinical chemistry at Children’s Hospital Boston and professor of pathology at Harvard Medical School. Rifai has been a leading researcher in evaluating the utility of CRP and its associated assays, and has collaborated extensively with Paul Ridker, MD, principal investigator of the JUPITER study.
The emphasis on CRP comes when there are many analytically sound high-sensitivity assays available for this biomarker that produce reliable results at detection limits <0.3 mg/L. But there are challenges, too. While cardiologists may understand the distinction between hs-CRP and traditional CRP tests with higher detection limits, this crucial difference may be less well-known by other physicians and patients, who may hear about the importance of CRP and demand the test. Lab reporting practices appear to have blurred, rather than clarified, the distinction between the tests, according to experts. Therefore, thought leaders in the field are urging close collaboration between laboratorians and physicians to ensure the test is ordered, reported, and interpreted properly.
“If a physician doesn’t have complete information about the tests, he or she may order CRP without realizing that an hs-CRP assay is needed to detect levels lower than the acute phase range,” said Mary M. Kimberly, PhD, chief of the Lipid Reference Laboratory at CDC. “Laboratorians can play a role in working with and educating physicians about the differences between the two.”
All About JUPITER
The JUPITER study involved just over 17,800 people in 26 countries, all with baseline LDL-C levels <130 mg/dL, CRP levels ≥2.0 mg/L, and no history of cardiovascular disease. Participating men were at least 50 years old; female subjects were 60 years old or older. Patients were randomly assigned to take daily either 20 mg of rosuvastatin (Crestor, AstraZeneca) or placebo. Follow-up visits, including a battery of lab tests, were scheduled after 13 weeks, and at 6-month intervals until 60 months after randomization, starting at the sixth month. Primary endpoints were first major cardiovascular event—defined as myocardial infarction, stroke, arterial revascularization, or hospitalization for unstable angina—or confirmed death from cardiovascular causes. The study was stopped after less than 2 years because of significantly favorable outcomes in the rosuvastatin treatment arm, including a 44% reduction in primary endpoints overall, a 54% lower risk of heart attack, a 48% lower risk of stroke, as well as a 50% drop in LDL-C levels, and a 37% decrease in CRP levels. The researchers estimated that applying the JUPITER screening and treatment strategy to the overall U. S. population for 5 years could prevent more than 250,000 heart attacks, strokes, revascularization procedures, and cardiovascular-related deaths.
“Our results are relevant for patient care and the prevention of heart attack and stroke,” noted Ridker, who is director of the Center for Cardiovascular Disease Prevention at Brigham and Women’s Hospital in Boston and Eugene Braunwald professor of medicine at Harvard Medical School. “Physicians can no longer assume that patients are at low risk for heart disease simply because they have low cholesterol. We have confirmed that patients with increased hs-CRP are at high risk even if cholesterol levels are low, and we now have evidence that a simple and safe therapy cuts that risk and saves lives.”
A Growing Body of Evidence
CRP is an acute phase reactant that is produced by hepatocytes and regulated primarily by inflammatory cytokines, including interleukin-6 and tumor necrosis factor-alpha. Its serum concentration can increase 1,000 times above baseline in the event of infection, trauma, and surgery. Chronic elevated levels have been linked to malignancies and autoimmune disorders such as rheumatoid arthritis, lupus, and inflammatory bowel disease.
Discovered in the 1930s, high levels of CRP were first associated with acute coronary events in the 1950s. Interest in this relationship lead to a series of investigations and by the late 1990s, researchers linked CRP to cardiac risk and hypothesized that it could be useful for monitoring the impact of drug therapy. Most notably, a post hoc analysis of data from the Air Force/Texas Coronary Atherosclerosis Prevention Study (AFCAPS/TexCAPS) showed that statin therapy reduced vascular event rates in people with elevated CRP and low LDL-C. This intriguing finding lead to the prospective JUPITER study.
More than 30 studies have shown CRP to be an independent predictor of cardiovascular events, and the marker also has been proven to add prognostic value to cardiovascular risk assessment models such as the Framingham Risk Score and the Reynolds Risk Score. That level of evidence could lead to revised treatment guidelines, perhaps with a recommendation for statin therapy in patients with low LDL-C levels who previously might only have been educated about diet and exercise and not prescribed a statin. Considerations in changing any guidelines would likely involve the costs and benefits of CRP screening for a wider population along with the costs, risks, and benefits of taking statins, according to Alan Remaley, MD, PhD, section chief of the Lipoprotein Metabolism Section of the Pulmonary and Vascular Medicine Branch at NIH. Updated NCEP guidelines are in the process of being updated. At the time the JUPITER results were released, Elizabeth Nabel, MD, director of the National Heart, Lung, and Blood Institute, indicated that the findings would be “part of the rigorous scientific review to distill the scientific evidence and generate an evidence-based, comprehensive set of clinical guidelines for primary care practitioners to help adult patients reduce their risk for cardiovascular disease.”
Current American Heart Association/CDC guidelines, issued in 2003, identified CRP as the best of all existing inflammatory markers for use in clinical practice. The guidelines recommended testing CRP in individuals at intermediate risk, defined as a 10% to 20% risk of coronary heart disease per 10 years. The guidelines left it to the physician’s discretion to have CRP testing performed in adults without known cardiovascular disease. The benefits of any therapy based on CRP levels in intermediate risk individuals and of obtaining CRP measures in asymptomatic patients were noted as uncertain. The guidelines called for measuring CRP twice, optimally two weeks apart, with scores categorized as follows: low risk, <1 mg/L; average risk, 1.0 mg/L to 3.0 mg/L; and high risk, >3.0 mg/L. At the time the JUPITER study was released, AHA indicated that JUPITER and other studies would be included in ongoing updates of prevention guidelines.
Draft guidelines of the National Academy for Clinical Biochemistry issued in 2006 recommended not testing for CRP levels in individuals with less than a 5% 10-year predicted risk. The guidelines also noted that about 10% of individuals with a 10-year risk between 5% and 10% might be reclassified to a higher risk group based on hs-CRP test results, but more information was needed. Finally, in patients with intermediate risk between 10% and 20% where there is uncertainty about whether to start preventive therapies, CRP might be helpful in further stratifying risk. The guidelines also noted that the benefits of therapy based on CRP levels are unclear and that there is insufficient data about using CRP to evaluate the effects of primary prevention efforts. Final guidelines are expected to be published in Spring 2009.
What Difference in Practice?
Absent any guideline changes it is unclear how JUPITER results will translate into changes in clinical practice. “It’s a fabulous study, but the real question is, will it change medical behavior?” noted Remaley. He wondered how swayed some clinicians will be, since the absolute risk reduction for patients with low LDL-C and high CRP who received rosuvastatin was “quite small,” about 0.9%, versus 1.8% in the placebo group. He added that it may be necessary, as was done for mammography screening of breast cancer and other screening tests that initially were controversial, to perform a controlled study in the field to determine how the use of CRP in routine medical practice impacts cardiovascular outcomes in this patient population.
When it published the JUPITER study, the New England Journal of Medicine ran an online survey about how the results should impact lab screening and statin use. The responses were split down the middle: of 2,553 participants, 49% indicated that, based on the JUPITER trial findings, the approach to laboratory screening of apparently healthy adults should be changed, versus 51% not in favor of a change. Similarly, 48% thought the therapeutic use of statins should be changed based on JUPITER results, versus 52% who indicated no change was warranted. An accompanying editorial suggested that the existing recommended practice of selectively testing only asymptomatic individuals with intermediate risk might be a better strategy than routine measurement of patients heretofore considered at average risk. (See Survey Results, below)
New England Journal of Medicine Website Survey
Physicians Weigh In
Do you believe, on the basis of the JUPITER trial results, that the approach to laboratory screening of apparently healthy adults should be changed?
Do you believe, on the basis of the JUPITER trial results, that the therapeutic use of statins in apparently healthy adults should be changed?
Reprinted by permission from Massachusetts Medical Society: online, accessed January 8, 2009. Total responses=2,553.
As noteworthy as the JUPITER results may be in establishing CRP as a significant marker of inflammation, the trial was not designed to, and didn’t address the precise roles either inflammation or CRP play in cardiovascular disease. According to Ridker, CRP has been shown to be predictive of future events, but correlates “only modestly with underlying extent of artherosclerosis.” Similarly, “the inflammation detected by hs-CRP appears to represent … a propensity for plaque rupture more than an overt acceleration of atherosclerosis” (Arterioscler Thromb Vasc Biol 2008; 28:1222–1224).
The role of CRP in artherogenesis remains a hot subject of investigation. It is unclear whether CRP is an innocent bystander merely indicating the presence of inflammation or an active participant in the inflammatory process that is integral to artherogenesis, although recent evidence suggests the latter. CRP has been associated with several proartherogenic properties, including activating endothelial cells to express adhesion molecules, inducing the secretion of interleukin-6 and endothelin-1, and activating macrophages to express cytokine and tissue factor (Circulation 2003; 107:370–372).
This uncertainty does not diminish the significance of CRP in assessing cardiovascular risk, according to Rifai. “For CRP’s use as a marker it doesn’t matter if it’s a causative factor or a marker. We know it is an exquisite marker and has additive value to other risk measures,” he noted.
Since rosuvastatin lowers both LDL-C and CRP, JUPITER also didn’t answer how much of an impact use of an anti-inflammatory agent alone would have on cardiovascular risk. As Timothy Gardner, MD, president of the AHA pointedly noted, JUPITER “was not designed to answer the question of whether the impact on risk was due to a reduction in inflammation (marked by hs-CRP) or a reduction in LDL. Statins lower both LDL cholesterol and hs-CRP. Thus, the findings … cannot determine whether lowering cholesterol, reducing inflammation, or a combination of both is responsible for the effects seen in this paper.” Some observers, including Remaley, have questioned whether aspirin therapy not being an exclusion criterion for JUPITER subjects might have weakened or confounded the results.
Even as the debate continues, researchers in Japan have found a continuous and graded relationship between CRP and cardiovascular risk, but at considerably lower levels than found in JUPITER and other North American and European studies (Arterioscler Thromb Vasc Biol 2008; 1385–1391). The Japanese subjects had a median CRP of 0.43 mg/L at baseline, compared with 4.2 mg/L in the JUPITER rosuvastatin arm and 4.3 mg/L in the placebo arm.
Challenges for Laboratorians
Regardless of the ultimate impact of JUPITER, the study unquestionably has heightened the profile of CRP. But this new appreciation of the biomarker comes at a time when issues with the hs-CRP assay, its utilization, and reporting have become apparent. hs-CRP assays first were developed in the 1990s concurrent with accumulating evidence about the utility of CRP. These assays, with claimed detection levels as low as 0.05 mg/L in contrast to traditional acute phase reactant assays with 3 to 250 mg/L detection limits, initially used ELISA.
Second- and third-generation tests use automated immunoturbidimetric and immunoluminometric methods, and today there are about 30 assays available world-wide, according to Rifai. A 2003 analysis of the tests found a number of issues related to their analytical variability, including varying detection limits, imprecision, antigen excess, and matrix effects (Clinical Chemistry 2003; 49:1258–1271). Further analysis is needed to assess the extent to which these concerns exist today, Rifai said.
CDC has an interest in developing a standardization program for hs-CRP assays, according to Kimberly, who has spearheaded two early efforts in that regard. The first involved evaluation of two secondary reference materials, and the second dealt with two approaches for assessing the commutability of a reference material. Separately, NIST is developing a reference measure procedure, but a deliverable is not expected for at least 2 years, according to David Bunk, PhD, research chemist in the analytical chemistry division.
As the industry moves toward harmonization and eventual standardization of hs-CRP assays, laboratorians have an important role in understanding the performance parameters of the assay being used, and in ensuring the accuracy of that particular test. “Labs need to be careful and not jump around among assays,” said Kimberly. “They need to establish one assay and set up a reference range. If they happen to change to another assay, they should be aware that there may not be assay-to-assay agreement.”
Options exist, however, for labs to verify their methods and assess their performance in comparison to peers. CAP sponsors three CRP-related proficiency tests: CRP with acute phase detection levels is includ-ed in the immunology surveys; hs-CRP with mean levels that fall within each of the three AHA/CDC risk categories is included in the cardiac risk surveys; and a linearity survey that tests values from 0.4 mg/L to 4.3 mg/L is part of the instrumentation surveys.
Beyond any technical issues with the hs-CRP assay, there are troubling concerns about the utilization and reporting of hs-CRP and CRP results. According to data from Information Dynamics, requests for hs-CRP have risen markedly since 2001, an increase one would expect because of the growing evidence about the biomarker. However, requests for CRP tests with acute phase detection limits have increased as well. “There hasn’t been a change in any clinical protocol that would warrant this increase, which indicates people are still using acute phase CRP assays to assess for cardiac risk. But it’s useless in that situation,” emphasized Rifai.
Growth in CRP and hs-CRP Assay Utilization
In 1999, high sensitivity CRP assays with detection limits as low as 0.05 mg/L were just being introduced. Usage of the hs-CRP assay grew with increasing evidence about its role in detecting cardiovascular risk. Yet utilization of the acute phase CRP assay also has risen without any new recommended uses. Some experts believe this is due to confusion about the purposes of the two tests.
Click here for figure
Reprinted by permission from American Association of Clinical Chemistry: Clinical Chemistry 2009; 55: 201
Absent computerized physician order entry systems or a clear means of differentiating between the tests on lab requisitions, physicians may inadvertently order the wrong test. “There’s no question that many physicians may not understand that there’s a distinct difference between CRP and hs-CRP,” said Gary Horowitz, MD, director of the clinical chemistry lab at Beth Israel Deaconess Medical Center in Boston and associate professor of pathology at Harvard Medical School. “I believe it’s the responsibility of labs to make sure physicians understand the difference, on the request and in the report.” Horowitz also is chair of the CAP chemistry resource committee, which oversees proficiency testing for hs-CRP.
A separate concern is that some labs are still reporting hs-CRP assay results in mg/dL, even though both AHA/CDC and NACB guidelines recommend expressing results in mg/L. This practice could result in misidentifying a person as being at low risk who actually is at high risk, or of having the test reported as beyond the limit of detection. Repeatedly sending out results with values below the detection limit will frustrate physicians, according to Rifai. “They’ll stop ordering the test and someone will be harmed,” he warned.
Some organizations have resolved this issue by using only the hs-CRP assay, or by reporting both CRP and hs-CRP assay results in mg/L. “For a period of time we were still reporting CRP in mg/dL but about a year-and-a-half ago, began reporting both CRP and hs-CRP in mg/L,” said Joseph McConnell, PhD, laboratory director of cardiovascular laboratory medicine at the Mayo Clinic. McConnell and his colleagues held many educational sessions for physicians and also included a notice on lab results that the change was impending. Other organizations and manufacturers use names for the test that emphasize the “cardio” aspect of hs-CRP, or locate hs-CRP in the cardiac panel on both printed and electronic order sets.
As laboratorians inform physicians about the appropriate use of hs-CRP assay results, Stephen Suffin, MD, notes that it also is important to emphasize guideline recommendations to perform the test sequentially in order to establish baseline levels of the biomarker. “Since even trivial inflammation can disturb the baseline, it’s important to make sure that false conclusions are not drawn,” he noted. Suffin is corporate medical director of clinical pathology and interim chief laboratory officer at Quest Diagnostics, based in Madison, NJ.
As the understanding of the role of CRP in cardiovascular disease grows and the import of the JUPITER trial becomes clearer, laboratorians will need to stay on top of the analytical issues related to this marker. Until these issues can be sorted out, laboratorians have a key role in ensuring the test is ordered and reported correctly so that physicians can accurately assess their patients’ cardiac risk.
Disclosures: Dr. Ridker is listed as a co-inventor on patents held by the Brigham and Women’s Hospital that relate to the use of inflammatory biomarkers including CRP in the detection and treatment of cardiovascular disease.