Medical Device Daily Washington Editor
BALTIMORE – Last week’s meeting of the American Association of Clinical Chemistry (AACC; Washington) brought into sharp relief a constant dilemma for medical science. As researchers peer more closely into the origins of disease, the roles of each of the numerous biochemicals involved emerge.
But the picture develops very slowly, and the migration of that knowledge into clinical practice is perhaps as incremental a process as the original research.
In discussing cardiac biomarker guidelines just released by the National Academy of Clinical Biochemistry , part of AACC, Roger Blumenthal, MD, the director of preventive cardiology at Johns Hopkins University Hospital (Baltimore), pointed out that guidelines are anything but static documents, reminding the audience that perhaps now the most difficult task for this branch of medical science is to find biomarkers that can identify asymptomatic patients in primary care settings.
Blumenthal said that coronary heart disease (CHD) claims about 959,000 Americans each year, more than cancer, accidents and acquired immunodeficiency syndrome (AIDS) combined. According to the American Heart Association (Dallas), half of men and 64% of women who died suddenly of CHD in 2003 exhibited no previous symptoms, thus driving the search for more accurate markers.
At present, risk calculations can be as simple as a combination of age, cholesterol levels (total and high-density vs. low-density lipoprotein), blood pressure and smoking (Journal of the American Medical Association in 2001), and as complex as a battery of tests, including electrocardiograms, stress tests, tests of immune-system agents and coronary calcium build-up. Blumenthal stated that measures of coronary artery calcium (CAC) indicated that a risk analysis based on the Framingham Heart Study, one of the most widely used in recent years, left out a number of individuals who probably should be on statins to lower cholesterol.
However, much of the focus of Friday morning’s sessions was on C-reactive protein, a much-ballyhooed substance that seemed to demonstrate substantial predictive power for myocardial infarction in a 1997 study of just under 1,000 men. Blumenthal said that the Physician’s Health Study “got a lot of people interested in CRP” after publication of the results in the New England Journal of Medicine (NEJM) that year.
Five years later, the author of the Physician’s Health Study, Paul Ridker, the director of the center for cardiovascular disease prevention at Brigham and Women’s Hospital (Boston), published data again in NEJM indicating that CRP boosted the predictive power of the Framingham, but Blumenthal said that this effect was seen “in some studies, but not in all.”
A comparison of CAC and CRP in the Oct. 15, 2002, edition of Circulation, penned by Robert Park, MD, et al at the Harbor-UCLA Research and Education Institute (Los Angeles) indicates that the former is “a potent predictor of cardiac events whereas CRP was only a modest predictor,” but Blumenthal stated that other data indicate that “the two may be very complementary” in evaluating risk for asymptomatic patients.
Still, CRP is not finished as a biomarker in Blumenthal’s mind.
In closing comments he said that “[d]ue to the overwhelming number of articles on CRP and the overall consistency, in my mind, when you look at a high CRP, that’s associated with a doubling of the Framingham score,” describing it as a secondary risk category.
One attendee asked Blumenthal if the case against CRP was not moving a bit too quickly and inquired as to how to proceed with a patient with mixed physical risk factors. Blumenthal replied that “if you’re on the fence” concerning biomarkers such as CRP, “familial history is so suggestive.” A patient whose uncle, as an example, died suddenly and without warning might be a good candidate for cholesterol-lowering therapy and routine monitoring. Blumenthal noted that those in medical science often have a tendency “to believe our data a little too much” when dealing with an individual patient.
Thomas Pearson, MD, the department chair of community and preventive medicine at the University of Rochester (Rochester, New York), highlighted the state of flux for those in the business of finding biomarkers for CHD, noting that “if you are unhappy with the evidence base for therapeutics, just try diagnostics.”
Pearson alluded to a glut of “evidence” in printed form in commenting that “the big push at NIH these days is translational medicine.” The difficulty now is that at present, “the typical clinician has about 1,500 guidelines” by which to steer his or her practice, he noted.
As for CRP, Pearson said that this protein “has the analyte and assay characteristics most conducive to use in clinical practice,” but noted that late data suggest that this predictive power “has crept down.” He noted further that data published in the Nov. 28, 2005, edition of the Archives of Internal Medicine suggest that CRP levels did not exhibit any more predictive power than lipoprotein assays. The study in question, conducted by Peter Wilson, MD, of the Emory University School of Medicine (Atlanta) et al, performed cardiovascular disease (CVD) risk assessment on 1,949 men and 2,497 women without CVD from the Framingham Heart Study. A follow-up eight years later suggested that elevated CRP level added nothing to the typical doctor’s office risk factor assessment for this group.
Pearson noted that one of the difficulties in developing risk tools is that the total cost of any such undertaking is the sum of the cost of tests, the costs of any prescriptions used in such a trial, the cost of any side effects, and the cost of any diseases that are not caught by that risk tool. As a consequence, he noted, “evaluation of diagnostics are few and far between.”
On the other hand, Pearson was also reluctant to announce the utter demise of CRP, recommending it as “a useful adjunct to risk predictions in patients for further testing or therapy ... especially if you’re on the fence” about that patient’s risk.