Rheumatoid Arthritis: A New CAD Risk Factor

Abstract & Commentary

Synopsis: Rheumatoid arthritis should be recognized as a marker of increased risk for myocardial infarction.

Source: Solomon DH, et al. Circulation. 2003;107: 1303-1307.

It is well known that there is increased morbidity and mortality in individuals with rheumatoid arthritis (RA), the most common systemic autoimmune disease, affecting 2 million Americans, most of whom are women. Recently, several studies have suggested increased rates of cardiovascular disease in these patients, potentially contributing to the reduced longevity in RA. In that atherothrombosis is an inflammatory process, it is a rational hypothesis that the chronic inflammation in RA may adversely affect blood vessels. Furthermore, it is well known that inflammatory markers, such as C-reactive protein (CRP) and other cytokines, are elevated in RA.

This observational investigation from the the Nurses’ Health Study (NHS), sought to assess whether RA was associated with increased cardiovascular events. In the NHS, women were enrolled in 1976 between the ages of 30 and 55 years and underwent questionnaire follow-up every 2 years. A total of 141,342 women remained after baseline exclusion of RA, CV disease, and cancer, of which 7786 women reported RA on at least 1 biannual questionnaire between 1978 and 1996. Solomon and colleagues pursued the RA diagnosis by a connective tissue disease screening questionnaire, followed up by careful review of medical records in the 2170 women with symptoms on the questionnaire suggestive of RA. The cardiovascular end points were myocardial infarction and stroke, both fatal and nonfatal, all verified by medical review. Associated confounding risk factors were adjusted for, including all potential major coronary artery disease (CAD) risk factors, as well as physical activity, BMI, folate, omega-3, and vitamin intake. Use of corticosteroids and nonsteroidal antiinflammatory drugs (NSAIDs) was examined for possible associations. (The data on corticosteroids are available only since 1994 and for NSAIDs, 1990). The primary analysis examined the age-specific incidence rates of stroke and myocardial infarction. The duration of follow-up was calculated as the interval between the 1976 questionnaire and the first diagnosis of myocardial infarction, stroke, or death, or conclusion of the study in mid-1996. Relative risks were computed for individuals with and without RA who had cardiovascular end points. A pooled logistic regression technique was used to adjust for multiple potential confounders, including a wide variety of risk factors. Several other analyses were used, all of which produced the same results as the main analysis.

Based upon medical record reviews, 527 women developed RA during the follow-up out of the 114,000 women in the observational study, representing more than 2 million years of follow-up. There were 2296 myocardial infarctions and 1326 strokes. The women with RA were similar to those without RA with respect to CAD risk factors, with some variation. Three percent of the RA patients reported corticosteriod use since the data were documented beginning with the 1994 survey. The age-adjusted risk of myocardial infarction was 2-fold for patients with RA compared to those without (RR, 2.07). After adjustment for multiple confounders, the relative risk was identical at 2.0, roughly comparable for fatal and nonfatal infarction. These risk ratios were highly significant. However, adjusted stroke rates did not achieve significance, although there was a trend approaching 50% more stroke in the RA women. Duration of disease seemed to be important; women with less than 10 years of known RA had an adjusted relative risk of 1.16 (NS), while those with > 10 years of exposure to RA had a risk of myocardial infarction of 3.1. Solomon et al point out that the relative increased risk of RA of 2- to 3-fold is comparable to that suggested in smaller studies in the literature. They emphasize that the risk association remained after adjustment for known CAD risk factors. Laboratory analyses and markers of inflammation are not available from the NHS. Solomon et al point out, however, that, "many of the cells comprising the inflammatory infiltrate in the joint lining are likewise found in atherosclerotic plaque." CRP and other cytokine markers of inflammation are known to be elevated in RA. Furthermore, one recent study reported a decrease of cardiovascular mortality in patients treated with methotrexate, suggesting that immunosuppressive therapy was beneficial for the vascular wall. Solomon et al point out that there may be inadvertent confounders, such as reduced physical activity, the various medications taken for RA, and perhaps differential use of cardiovascular prevention medications. They emphasize that medications used for RA have the potential for both inducing and protecting from thrombotic events and atherogenesis. Efforts to control for use of corticosteriods and NSAIDs were carried out, with the risk of MI persisting. Solomon et al concluded, "RA should be recognized as a marker of increased risk for myocardial infarction." They believe that treatment medications for RA and perhaps less-than-adequate CAD prevention therapy may contribute to the increased risk in these women. They concluded, "It would be prudent to consider aggressive cardiac prevention measures in patients with RA to address established coronary heart disease risk factors."

Comment by Jonathan Abrams, MD

This is an interesting analysis, which seems not surprising given our current knowledge and the focus on inflammation and its relationship to atherothrombosis and unstable plaque. In this large, observational cohort, it is difficult to establish the effects of various potent drugs used for RA. Furthermore, the recent controversy about COX-2 inhibitors and the suggestion that nonselective NSAID may be safer point out the complexity of anti-inflammatory medications. Aspirin use was less common in the RA patients, 41.5% vs 52.2%. Hormone replacement therapy was slightly greater in RA, 33% vs 27%; as expected, NSAID use approached 67% in the RA patients vs 22% in non-RA women. Physical activity, measured in estimated mets per week, was somewhat less in the RA women but not strikingly so. No data were provided for CRP in these women, and one wonders if the banked plasma stored at the beginning of the study could be analyzed for CRP. It is clear that many to most patients with RA will have elevations of CRP and other inflammatory markers. Many believe that CRP itself may play an adverse role in the vasculature and is not just a marker of increased vascular risk. Thus, chronic elevation of cytokines in this autoimmune disease over many years may induce or exacerbate events in the vessel wall in promoting atherothrombosis and perhaps unstable plaque.

I agree firmly with the conclusions of Solomon et al that aggressive CAD preventive measures should be considered in RA, and I would suggest that this approach be mandated. We are currently treating diabetes without obvious overt vascular disease as a CAD risk equivalent (ie, diabetics should be treated with the same target goals of blood pressure and lipid modification as individuals with established CAD or a previous stroke). It is logical that for the RA patient, one should aim for an LDL target of 100 mg/dL or less; aspirin should be used in all; COX-2 NSAID use should be avoided until the current controversy is resolved; blood pressure should be controlled, with a target of 130/80 or less; an optimal "heart healthy diet" is recommended; and as much physical activity as can be carried out should be. With the complexity of this illness and its treatment, it may be difficult to carry out randomized trials looking at CAD risk prevention, particularly since the major adverse vascular effects of RA are seen only after 10 or more years of exposure. Thus, RA is now added to the pantheon of conditions for which very aggressive CAD prevention approaches are warranted. In this case, it is unlikely that a "smoking gun" trial of aggressive prevention therapy vs "usual care" will be carried out.

Dr. Abrams is Professor of Medicine, Division of Cardiology, University of New Mexico, Albuquerque