A Message from the Heart
A new test could save the lives of millions who don't even know they're in danger
`So how's your LDL doing?" "It's down--I've been laying off the chips. You?" Who ever would have anticipated that Americans would someday chat casually about their cholesterol? Very soon, another test for heart disease, with at least as much lifesaving value as a cholesterol check, will become equally commonplace. Commit "C-reactive protein" to memory. If you haven't already seen it on a lab printout, you will.
There's nothing new about this protein. It was identified in 1930, and scientists have known for decades that the liver pumps it into the bloodstream when something is going on that results in inflammation. That happens, to cite obvious examples, when you have a cold or a gum infection, or after you cut yourself or have surgery. Or, less obviously, if you are developing cardiovascular problems, because inflammation is now believed by most heart researchers to be the process behind most cardiovascular disease.
Aha, you say. Wouldn't that make CRP a logical indicator for the condition that turns arteries into plaque-filled death traps? That is just what it is turning out to be. An eight-year study of nearly 28,000 women, published in the current issue of the New England Journal of Medicine, found that CRP foretold heart attacks and strokes better than levels of "bad" cholesterol (LDLs, or low-density lipoproteins) did. Combining tests of both cholesterol and CRP, says lead author Paul Ridker, director of the Center for Cardiovascular Disease Prevention at Brigham and Women's Hospital in Boston, would have immense predictive power and enormous impact. "It's high time to move beyond cholesterol," Ridker says firmly.
The study, which is being showcased and debated this week at the American Heart Association's annual science meeting, is the latest and largest of well over 1,000 studies assessing CRP's predictive value. There is a considerable range of opinion. But as the pile has grown, experts have moved steadily toward a consensus that the CRP test could be instrumental in identifying many of the estimated 25 million people who have atherosclerosis but don't know it because by normal standards they're not at high risk. "It will add a new component," says Sidney Smith, the AHA's chief medical officer and head of cardiology at University of North Carolina Hospitals. It doesn't hurt that the test costs only $10 to $25 and has been covered by Medicare since January, although private health insurers generally won't pay for it.
It's not easy to find and treat such individuals, who often don't display the warning flags doctors look for. Half of all heart attacks strike people who don't have a cholesterol problem. Their LDL levels are not high and their levels of protective "good" cholesterol (HDLs, or high-density lipoproteins) are not low. And at least 25 percent of heart attacks happen to individuals with no major risk factors. In addition to having acceptable cholesterol levels, these are people who don't smoke, don't have abnormally high blood pressure, and are not diabetic. Their weight is not always ideal, but they are not obese. Yet somehow the heart disease process has been triggered.
It's a profile that neatly fits 60-year-old Marian Boesen, except that the resident of Sterling, a small city in northwestern Illinois, got lucky. The intake form she gave Peter Toth, a family practitioner at a local clinic, on her first visit last year showed that both parents had had strokes--her father at 55, her mother at 69. "Bells went off," says Toth. But Boesen was more focused on thyroid and migraine problems at the time, and didn't have her cholesterol and other risk factors checked. "I just wasn't interested," she says. Last August, however, she paid $125 on her own for an ultrasound scan, advertised in a newspaper flier, of various portions of her body. "Down deep," she says, "I think I felt if my family was prone to strokes I'd best get checked." The report came back normal--except the scan of her carotid arteries, which revealed a small amount of plaque on one side.
Prevention. "Any buildup in the carotid, you jump," says Toth. Last month Boesen succumbed to his importuning and had a complete blood work-up. Only one number jumped out: high CRP. Toth counseled Boesen on diet, exercise, and weight but also put her immediately on aspirin, which may lower inflammation in heart disease the way it does for sore joints. And he started her on Zocor, a statin--not because statin drugs can lower cholesterol, but because studies (including one coauthored by Ridker) indicate that they can also bring down CRP.
Doctors don't routinely send patients like Boesen for ultrasound screening, notes Toth, and standard cholesterol lab tests wouldn't have provided the slightest hint of a problem. "You wouldn't have thought there was anything wrong with her," he says, adding that of the patients he refers out for coronary bypass surgery and angioplasty, about 75 percent have elevated CRP and about one third of those patients do not have abnormal cholesterol.
Toth, unhappily, is not the typical primary-care physician. He has a Ph.D. in biochemistry, has spoken at length with Ridker, keeps up with the studies, and has become something of a CRP evangelist: "I am now going to incorporate CRP in evaluating risk even for patients who don't have a strong family history or other high-risk profile," starting at age 25 or 30. Kenneth Cooper, president of the Cooper Aerobics Center in Dallas and a personal physician to George W. Bush, is another proselytizer. A CRP report is part of the center's blood work-up for everyone who comes for fitness testing, and it was Cooper who introduced a CRP check into the president's annual physical starting last year. (It was satisfyingly low in 2001 and again this year.) "I think we're going to find that inflammation is more of a factor than we ever used to think," says Cooper. "The best measure of inflammation may be CRP."
Such enthusiasm for CRP testing is anything but universal. "Some day it may be useful in clinical practice," says Daniel Levy, principal investigator for the Framingham Heart Study. But, he adds emphatically: "Not yet." Research, he points out, hasn't established that steps taken specifically to reduce CRP also cut the risk of heart disease. And he worries about veering away from the real problem, which he sees as the simple fact that people need to stop smoking and need to address their cholesterol and blood pressure. "Only 27 percent of people with high blood pressure take steps, such as drugs, to control it," he says. "It's even worse for cholesterol."
Many physicians hesitate to order up CRP tests because the scientific dust hasn't settled yet. The cutoff points that define risk categories haven't been determined. There isn't even an agreed-upon laboratory standard for measuring CRP. The traditional CRP assay couldn't sniff out low levels of inflammation, so a few years ago more-sensitive tests were developed--quite a few, in fact. "About 30 different methods for analyzing CRP are available today, all of them approved by the Food and Drug Administration," says Nader Rifai, a longtime Ridker coauthor and director of clinical chemistry at Children's Hospital Boston. Last year, a comparison in the journal Clinical Chemistry of nine methods of determining CRP found that the results did not always agree. Nor are there clinical guidelines to determine which patients should be tested and how to interpret the results.
That is all about to change. At a meeting last March, the federal Centers for Disease Control and Prevention and the American Heart Association agreed to issue joint guidelines on measuring and using CRP, exactly what primary-care doctors and cardiologists in offices and clinics need. And last week, as Ridker's paper neared publication, Thomas Pearson, a University of Rochester (N.Y.) Medical Center cardiologist in charge of writing the guidelines, had just finished sorting through the comments submitted by various experts. "There are some surprising deficiencies in the evidence," notes Pearson, "and [Ridker's] study does patch a couple of those holes." He hopes that the guidelines will be published by late winter. "We want to get them out as quickly as possible," he says.
Nothing is settled. The guidelines haven't even gone through a first draft, and both the CDC and AHA will have to approve them. But three levels of risk, taken directly from the New England Journal study, are likely: low for CRP below 1 milligram per liter, moderate for CRP between 1 and 3 mg/L, and high for CRP above 3 mg/L. That's less complicated than the four or five risk groups in most CRP studies. And two readings, especially if the first is extremely high, will be recommended.
Whom to test? "The real question," says Pearson, "is, What is the point at which you should say, `This is useful,' and then apply it to patients?" No guidelines can possibly answer for each clinic or individual practitioner. Among the hottest debates at this week's AHA meeting will be which patients should get a CRP test. "I measure CRP selectively," says Lori Mosca, director of preventive cardiology at New York-Presbyterian Hospital and author of a companion editorial on the Ridker-led study in the New England Journal.
An example, Mosca suggests, might be a middle-aged man with high blood pressure and LDL cholesterol between 130 milligrams per deciliter and 160 mg/dL. "He doesn't qualify for statin therapy based on national guidelines, but maybe he has a family history of early heart disease." A CRP test that comes back high might push her toward giving him a statin drug, and even determine whether to refer him for stress testing. "I wouldn't use it on very high-risk people, and I wouldn't use it on very low-risk people," says Mosca. "I use it on people with a moderate level of risk, and then only after conventional measures have been tried."
Conceivably, the guidelines could go beyond Ridker's own opinion of which patients might be good candidates for a CRP test. For high-risk patients, he says--those whose LDL cholesterol exceeds 160 mg/dL, for example--"we already have guidelines encouraging us to treat them aggressively." So the only reason to test them is that a high CRP "might tip the balance for a patient reluctant to take medication or a physician reluctant to apply them." Ridker is not in favor of putting most Americans through CRP screening. That position may help placate critics who have observed that he is co-owner of patents on the test. According to Ridker, he might collect a penny or two per test if the patents are fully enforced--a possibility he describes as highly unlikely.
Part of the whom-to-test puzzle is that CRP's predictive ability won't be equal for everybody. Even in Ridker's analysis, it is weaker, for example, at predicting heart disease in women on hormone replacement therapy. "HRT raises CRP," says Thomas Pearson. "We don't know why. In HRT users, the level is raised to the point that it doesn't show that much of a dose-response relationship." (High LDL doesn't work as well in these women in predicting heart disease, either.)
CRP may be more than just a marker for heart risk. It may be a causative agent. Studies suggest that it is produced not only in the liver but also inside artery walls where cellular troops are gathering to initiate an inflammatory response. CRP may promote the formation of "adhesion molecules" that, sirenlike, lure white blood cells out of the bloodstream to stick to artery walls. From there, they squirm inside the wall, where they contribute to a growing pocket of inflammation. CRP also may lower the level of nitrous oxide--"a major player in keeping blood vessels healthy," says Ridker.
The clear message is that it may be as vital to bring down elevated levels of CRP as it is to lower high levels of LDL or raise low levels of HDL. Conveniently, most of the measures that seem to whittle CRP are the same ones already known to benefit the heart. Statins, for example, cut LDLs and boost HDL--and have been shown to reduce CRP levels by a healthy amount. And aspirin, firmly established as a cheap and effective way to cut the risk of heart attack and stroke, may lower CRP as well, although that has not yet been proved in clinical studies.
Exercise works, too. A study in the current issue of Arteriosclerosis, Thrombosis, and Vascular Biology shows a reduction in CRP of about 30 percent for men who went from sedentary to walking 30 minutes five times a week, even if the men didn't lose weight. "The big benefit is in just getting off the sofa," says Tim Church, the lead author and medical director of the research arm of the Cooper Aerobics Center.
But losing weight also helps. In fact, the inflammatory model of heart disease helps explain why shedding pounds has benefits beyond relieving the heart of the burden of lugging extra weight and supplying blood to the added tissue. Researchers are finding that fat cells are far from passive. They dump interleukin-6 into the blood, which promotes an inflammatory response and happens to be the key cellular signal to the liver to churn out CRP. It is as if the bodies of overweight and obese individuals are in a constant state of low-grade inflammation. So it follows that losing weight lowers CRP and may help cool off the inflammatory process.
Key trial. A key study remains undone: If a CRP test has independent value for predicting heart disease, then people with good cholesterol levels but high CRP levels should have fewer cardiovascular events if they bring down their CRP. That's the study Levy and others want, and so does Ridker. "That trial is absolutely critical," he agrees. "If the answer is no, we should look for other ways to apply our healthcare dollars."
Last week Ridker got his wish: approval of a CRP trial of 15,000 American men older than 50 and women older than 55. None will have a history or major risk of heart disease. All will have LDL levels below 130 mg/dL and CRP levels above 2 mg/L. Half will be given a placebo. The other half will get a statin drug to lower their CRP, which if the CRP model holds true should reduce their chances of dying, having a heart attack or stroke, needing angioplasty, or other cardiovascular event. Ridker hopes to begin the three- to four-year study in the spring.
Whether the trial goes thumbs up or down, no one argues that CRP should be the only index of potential heart disease--or even that cholesterol and CRP are enough. In 10 years we may be talking about homocysteine, sub-subcategories of cholesterol, and even genetic markers for heart disease as breezily as we now drop LDL readings into conversation. Each person's mix of risks is different.
How else to figure someone like Fred Stancel? The Denver resident is a 50-year-old vegetarian with no family history of heart disease. He doesn't smoke. He isn't overweight. His blood pressure is fine. His LDL level is extremely low. He bikes 4 1/2 miles to and from work and another 50 miles on weekends. He has participated in Ride the Rockies, a weeklong bicycle road trip that crisscrosses the Continental Divide. Yet in July of last year, he had a heart attack, followed by triple bypass surgery. "I didn't fit the normal risk profile," says Stancel with deliberate understatement. And eight months later, one of the bypass grafts showed unmistakable signs of closing back up. Last month, after reading about Ridker's work on CRP, Stancel paid his own way to Boston. No explanation was to be found there; his CRP was in the low-risk zone.
The way Peter Toth thinks about CRP, all patients are puzzles and CRP is just one more tool that can fill in a missing piece. He seethes when critics in academic medicine, as he sees it, want to wait until the last T is crossed before CRP meets their approval.
"There's a whole lost group, people whose risk isn't being picked up," Toth says with passion. "It's the practicing cardiologists out there who have to deal with the consequences. We are the ones who have to go into the hospital at 2 or 3 in the morning to handle a myocardial infarction. We're the ones who have to talk to families and tell them, `I'm really sorry--we lost your father and we don't know why.' CRP gives us a whole new crystal ball. Who can object to a $15 to $20 test that improves the chances of finding and preventing heart attacks?"
A telltale protein
The level of C-reactive protein (CRP) in the blood responds to a cycle of inflammation now understood to cause heart disease.
[Drawing is not available.]
[labels]
Early inflammation
An inflamed pocket of fatty "soft plaque" begins to form.
Coronary artery
Normal coronary artery
CRP molecules
Triggered by the inflamed plaque, molecules of CRP build up in the blood.
Advanced inflammation
Stimulated by chemicals that promote inflammation, the pocket of plaque continues to grow.
The pocket bulges into the artery channel.
CRP concentration rises.
Stephen Rountree--USN&WR
The CRP effect
A high level of C-reactive protein (CRP) dramatically raises the chance of heart attack or stroke.
[Complete chart data are not available.]
[labels]
Odds of a cardiovascular event in the next 10 years
0-1 pct.
2-4 pct.
5-9 pct.
10 pct. or more
High risk
Low risk
CRP below 1.0*
CRP 1.0-3.0
CRP above 3.0
* milligrams per liter
Sources: Framingham Heart Study; Paul Ridker, Brigham and Women's Hospital
This story appears in the November 25, 2002 print edition of U.S. News & World Report.