Arthritis & Rheumatism, Volume 62,
November 2010 Abstract Supplement
Abstracts of the American College of
Rheumatology/Association of Rheumatology Health Professionals
Annual Scientific Meeting
Atlanta, Georgia November 6-11, 2010.
Novel Cardiovascular Risk Prediction Models in Patients with Systemic Lupus Erythematosus.
Kawai2, Vivian K., Oeser4, Annette M., Solus3, Joseph, Rho3, Young H., Raggi1, Paolo, Bian3, Aihua, Gebretsadik3, Tebeb
Women with systemic lupus erythematosus (SLE) have accelerated atherosclerosis and increased coronary heart disease (CHD), but the Framingham risk score (FRS) underestimates risk. Several new approaches have improved prediction of CHD risk in the general population. These include the Reynolds risk score (RRS), that is of particular interest because it includes a measure of inflammation (C-reactive protein), and novel risk scores that substitute "coronary age" in place of chronological age. We examined the hypothesis that these new CHD risk prediction models would better identify increased risk of CHD in women with SLE, particularly in those with subclinical coronary atherosclerosis.
We studied 121 women with SLE and 65 age-matched female-controls without a previous cardiovascular event. Demographic and clinical data were recorded and FRS and RRS calculated. Coronary artery calcium (CAC) was measured by computed tomography using the Agatston score, and coronary age was derived as described by McClelland. Coronary age-modified risk scores (camFRS, camRRS) were calculated in which chronological age in the existing FRS and RRS calculations was replaced with coronary age. Risk scores were compared in patients with SLE and controls, and in SLE patients with and without CAC, using Wilcoxon rank sum tests. We examined the association between risk scores and CAC using proportional odds logistic regression models. Paired analyses were performed using Wilcoxon signed-rank test or McNemar's test, as appropriate. All analyses used a 5% two-sided significance level.
Coronary artery calcium was present in 21 patients with SLE (17%) and 4 controls (6%) (P=0.033). However, despite more frequent and severe atherosclerosis in SLE, the FRS, camFRS, RRS and camRRS did not differ significantly among women with SLE and controls (p>0.05). In women with SLE with CAC the FRS (OR=1.42, 95%CI (1.061.90), P=0.017) and RRS (OR=1.29, 95% CI (1.111.51), P=0.001) were higher than in those without CAC, however few patients (<=5%) with CAC had 10-year risk scores >=10% - the threshold used to determine moderate CHD risk. Coronary age-modified risk scores were significantly lower than their respective conventional risk scores in lupus patients without CAC, and higher in those with CAC (both P values <0.001). Coronary age-modified risk scores assigned more patients with CAC to a risk category of >=10% (camFRS 48%, cam RRS 29%) than conventional scores (FRS 0%, RRS 5%) (P =0.004 and P=0.073, respectively).
The RRS, a novel 10-year risk prediction model that includes C-reactive protein in the risk calculation, was of limited use in women with SLE. Incorporation of coronary age, calculated from CAC scores, into risk prediction models increases the number of patients with CAC with a predicted CHD risk above the 10% 10-year threshold. The ability of models that include coronary age to predict risk of cardiovascular events will be of interest.
To cite this abstract, please use the following information:
Kawai, Vivian K., Oeser, Annette M., Solus, Joseph, Rho, Young H., Raggi, Paolo, Bian, Aihua, et al; Novel Cardiovascular Risk Prediction Models in Patients with Systemic Lupus Erythematosus. [abstract]. Arthritis Rheum 2010;62 Suppl 10 :1867