It took 30 years for researchers to agree that lower serum cholesterol would reduce the risk of heart disease, but they seem to be moving considerably faster when it comes to inflammation. In the past five years, cardiologist Paul M. Ridker and his colleagues at Harvard Medical School's Brigham and Women's Hospital in Boston have managed to amass a compelling amount of data suggesting that a marker for inflammation, known as the C-reactive protein (CRP), is a potent risk factor for cardiovascular disease, and that drugs that reduce CRP levels will reduce heart-disease risk, as well.

The extent to which this idea is catching fire with medical researchers is made manifest by Ridker's phenomenal citations. By the summer of 2001, Ridker's articles were perched atop the Science Watch listing of the hottest papers in medicine. "Long-term effects of pravastatin on plasma concentration of C-reactive protein," (P.M. Ridker, et al., Circulation, 100 (3):230-5, 20 July 1999) was at #1, followed by "C-reactive protein and other markers of inflammation in the prediction of cardiovascular disease in women," (P.M. Ridker, et al., New Engl. J. Med., 342(12):836-43, 2000.) at #6. Indeed, in the past decade Ridker has published a dozen papers that have each attracted more than 200 citations, while his seminal work on inflammation and heart disease (P.M. Ridker, et al., "Inflammation, aspirin, and the risk of cardiovascular-disease in apparently healthy men," New Engl. J. Med., 336(14): 973-9, 1997) has racked up a remarkable 1,000+ citations in five years (see table on page 4). And that doesn’t even include two very recent papers published in the New England Journal of Medicine, one demonstrating how inflammation can be used to target statin therapy (P.M. Ridker, et al., New Engl. J. Med., 344(26):1959-65, 2001), and one demonstrating that inflammation as detected by CRP is a more potent predictor of vascular risk than LDL cholesterol (P.M. Ridker, et al., New Engl. J. Med., in press).

Ridker’s work has also attracted attention in the wider world: last year Time magazine included him, representing the field of cardiology, in an elite selection of researchers dubbed "America’s best in science and medicine."

Ridker, 43, graduated from Brown University in 1981 and received his medical degree from Harvard Medical School five years later. In 1992 he obtained a Masters in Public Health from the Harvard School of Public Health, and that same year began as an attending cardiologist at Brigham and Women's Hospital and an instructor at Harvard Medical School. Today Ridker is an associate professor of medicine at Harvard and director of the Center for Cardiovascular Disease Prevention at Brigham and Women's Hospital.

Ridker spoke to Science Watch correspondent Gary Taubes from his Boston office.

  What was the origin of the idea that inflammation might play an important role in heart disease?

 The idea that inflammation might be important for atherosclerosis goes back almost 70 years, in many ways predating the cholesterol hypothesis. That inflammation was critical to the initiation and progression of atherothrombosis was re-discovered by the vascular biology community about 15 years ago, with Russell Ross, now deceased, being a guiding light in this field. The influence of this work was enormous for the direction of our research group, because we began to ask how we might translate this molecular information about inflammation into a population setting where we could actually test it. So that began a whole series of NIH-funded studies, starting in 1994, in which we began to ask whether there were molecular biomarkers of inflammation, measurable on a population basis, that we could use to predict heart disease in large-scale prospective cohorts—in tens of thousands of initially healthy individuals. We measured several upstream markers of inflammation: interleukin-6, TNF-alpha, soluble ICAM-1, soluble VCAM-1, P-selectin, E-selectin, CD40 ligand, and macrophage inhibitory cytokine or MIC-1. What was extraordinary was that almost all of these mediators of the systemic immune response seemed to be elevated among apparently healthy individuals at high risk of future heart attacks or strokes. However, while intriguing from a pathophysiologic perspective, none of these biomarkers had clinical appeal. All had limitations in the laboratory—issues, for instance, that would matter greatly at the public-health level, such as how easy they are to measure in routine clinical settings.

  So where does C-reactive protein come in?

 We wanted a simple downstream clinical marker of inflammation that could be moved out of the lab and into the clinic. And CRP turned out to be a wonderful candidate. CRP is a pentraxin and a quintessential acute-phase reactant. It was discovered in the 1930s at Rockefeller University, when people were very early in the process of understanding innate immunity. A group of very famous scientists in the 1940s, like Oswald Avery and Mac McCarty, had found that CRP could be used to detect acute inflammation and help to understand the timing of the body’s initial response to infection. Interestingly, in the early 1950s, there was even a brief time when CRP was being used to help diagnose acute ischemia. Then it fell out of general interest when cholesterol came around and pretty much dominated the next 50 years of cardiology research.

  How did you come upon it?

 We were just trying to find an appropriate inflammatory biomarker with adequate characteristics for clinical use. A few clinical reports of CRP in acute ischemia had appeared, most notably from Wayne Alexander and Attilio Maseri. These papers were important, but they could not tell us whether inflammation preceded or was simply a result of coronary artery occlusion.

  Which study convinced you that CRP was the key and that inflammation was significant in heart disease?

 Our first major CRP publication was in the New England Journal of Medicine in 1997. In that study, which derived from a cohort of 22,000 healthy middle-aged men, we measured CRP levels using a new high-sensitivity approach developed in the laboratory of Russ Tracy at the University of Vermont. What we found was that baseline CRP levels in these completely healthy men were a very potent predictor of risk for first-ever heart attack or stroke during an eight-year follow-up period. That study essentially demonstrated that inflammation could be measured in individuals with no prior cardiovascular history and that it had profound prognostic ability. Looking back, that one paper really constituted hard clinical evidence corroborating a whole field of inflammation research at the laboratory level. Then, in a second NEJM article, we showed that CRP also predicts disease in healthy middle-aged women. But more importantly we measured a whole series of inflammatory biomarkers simultaneously, along with traditional risk factors, and we found that inflammation was at least as important as cholesterol. In particular, CRP appeared to be a stronger predictor of future heart attack risk than LDL cholesterol, and CRP was detecting risk even when lipid levels were low. We then published a series of papers showing that LDL levels and CRP don’t relate to each other, so you cannot predict your inflammatory response on the basis of plasma cholesterol levels. From an epidemiologic perspective, we knew this had to be true.

  If you could explain in a little more detail, why does it have to be true?

 In order for CRP to be such a dominant predictor of risk, yet independent of lipids, it means you have two different processes at work, both of which are necessary, neither of which is sufficient. While the lipid levels relate to the build-up of plaques in the artery, the inflammatory response—as detected by CRP—seems to be describing plaque instability and the propensity of plaque to actually rupture. Those are two different issues. That’s why people with low cholesterol levels but high CRP, for instance, are still at very high risk.

  Does this say anything about how standard therapies—aspirin, for instance—are actually working? And what about statins? There has been talk for years that they lower heart disease risk through mechanisms other than lowering cholesterol.

 Well, aspirin is also an anti-inflammatory drug. And the Physicians Health Study—that original 1997 paper—happened to be a randomized trial of aspirin. So we were simultaneously able to show not only that CRP levels predicted the risk of heart attack, but that the magnitude of the benefit that any individual got from aspirin was directly related to the CRP level. As for statins, we were able to demonstrate very early that these agents not only lowered CRP levels, but also appeared to have a greater relative efficacy in preventing heart attacks among those with elevated CRP levels. These data provided a clear clinical demonstration that statins, in addition to being potent lipid-lowering drugs, also had important anti-inflammatory properties. What is also of interest is that the magnitude of LDL reduction achieved with statin therapy does not predict the magnitude of CRP reduction, so these drugs may be working through mechanisms we are only now beginning to understand. In one recent study, we showed that the magnitude of risk reduction associated with random allocation to statin therapy is just as large for individuals with high CRP as it is for individuals with high levels of cholesterol. Moreover, we found that those with elevated CRP levels but low LDL levels—a group which on a population basis includes almost 30 million Americans—actually appear to benefit from statins just as much as did those with overt hyperlipidemia. These data raise the possibility that the very way we prescribe these drugs needs to be carefully reconsidered. It may well be that people with low cholesterol, who therefore don’t qualify for drug therapy, nonetheless benefit markedly from their use because of these vascular anti-inflammatory effects. Our most recent data push this concept even further; we now have evidence that the CRP phenomena predict vascular risk across a full spectrum of LDL levels and add prognostic information at all levels of the Framingham Risk Score.

  How was this research accepted initially?

 Initially, much of it was simply not believed. What has been incredibly gratifying is that these concepts have rapidly become a focal point for both clinical and laboratory research around the world. We are now seeing many reports that CRP is not just a marker of this process, but a direct mediator of atherothrombosis and perhaps even a target for therapy. Moreover, we are seeing a number of targeted anti-inflammatory therapies now being considered as possible agents for vascular protection.

  Beside statins, are there other ways to lower CRP levels?

 There are, but whether they're better or not, we don't know. I’m a preventive cardiologist by training, and I bring a strong public-health perspective to my work. Exercise, weight loss, and smoking cessation will probably turn out to be ways to control CRP levels without medication. Because we think inflammation is part and parcel of arteriosclerosis, it may be that almost anything that reduces heart-attack risk at a population level will in turn lower CRP levels.

  Do we know what causes the inflammation to begin with?

 That’s the Holy Grail at the moment, and there are many competing hypotheses. A favored hypothesis is that atherosclerosis is simply a fundamentally inflammatory disease. From this perspective, we have to think of heart disease as an inflammatory disorder in the same way we think of rheumatoid arthritis or lupus in those terms. This viewpoint suggests that we are witnessing evolutionary biology in action—that an adaptive response in the past is now maladaptive in our current modern environment. For millions of years, humans had to survive infection, trauma, and starvation, as these were the things that killed our forebears. So we have systems like those of innate immunity and insulin resistance that evolved to protect us from infection and starvation. A million years ago, the only goal was to live long enough to reproduce. Arteriosclerosis only became a problem in the mid-to-late 20^th century when death from infection and starvation became rare and we began living well into our 70s and 80s. In some ways, the heightened inflammatory response we have inherited and which is part and parcel of arteriosclerosis must be seen as a by-product of our success at learning how to live to a ripe old age.

  What's next on your research agenda?