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.

The Relative Importance of Current Versus Past Use of Medication in Excess Rates of Adverse Cardiovascular Events among Patients with SLE.

Petri1,  Michelle A., Magder2,  Laurence S.

Timonium, MD
Department of Epidemiology and Preventive Medicine, University of Maryland School of Medicine, Baltimore, MD


Previous studies have suggested that some of the excess risk of adverse cardiovascular events (CVE) in patients with SLE may be due to treatments such as prednisone. Other studies have suggested that treatment with hydroxychloroquine results in reduced risk. In this study, we investigated whether past or current drug exposures have greater association with CVE.


We examined CVE rates in a large clinical cohort, in subgroups defined by medication history. CVE was defined as myocardial infarction (MI), stroke, coronary procedure, incident angina, or claudication. The analysis was based a data set which had one record per patient-month of participation in the cohort. Each record contained data regarding the clinical history up until that time, the most recently measured levels of disease activity, medications taken at that time, and whether a CVE occurred during that month. This file was analyzed using pooled logistic regression.


134 CVE events were observed from among 1874 patients who were followed for 9485 person-years during periods between 1987 to 2010. Events included 65 strokes, 27 MI, 29 cases of angina or coronary procedures, and 13 cases of claudication. The Table shows rates of CVE in subgroups defined by current and past medication use.

Table. Rates of CVE in Subgroups defined by Current and Past Medication Use

SubgroupNumber of CVERate of CVE per 1000 person-yearsRate Ratios based on a model that adjusts for age (95% CI)P-value
Prednisone exposure    
        None2213.31.0 (Ref. Group) 
        Past (not current)237.90.6 (0.4, 1.2).14
        Currently taking (any dose)8818.21.6 (1.0, 2.5).057
        Currently taking >=20 mg/day2535.44.0 (2.2, 7.1)<.0001
Plaquenil Use    
        None4617.91.0 (Ref. Group) 
        Past (not current)2020.31.1 (0.7, 1.9).65
        Currently used but for <6 consec. mo.1416.91.0 (0.6, 1.9).95
        Current use, >6 consec. mo5410.60.5 (0.4, 0.8).0019
NSAID Use    
        None5112.61.0 (Ref. Group) 
        Past (not current)4215.01.0 (0.7, 1.6).86
        Current4115.61.1 (0.7, 1.6).78
Immunosuppresant Use    
        None5612.11.0 (Ref. Group) 
        Past (not current)310.10.8 (0.2, 2.5).67
        Current7516.51.5 (1.0, 2.5).31

In multivariable models, cumulative (past) corticosteroid exposure was not associated with CVE after controlling for current use. Similarly, cumulative exposure to hydroxychloroquine was not protective after controlling for current use. Also, in multivariable models, the association between current corticosteroid use and CVE persisted after controlling for the most recent measure of disease activity.


Current (but not past) use of high-dose corticosteroids was associated with high rates of CVE. Current (but not past) use of hydroxychloroquine for 6 or more months was associated with lower rates of CVE. This study strongly supports long-term use of hydroxychloroquine in SLE.

To cite this abstract, please use the following information:
Petri, Michelle A., Magder, Laurence S.; The Relative Importance of Current Versus Past Use of Medication in Excess Rates of Adverse Cardiovascular Events among Patients with SLE. [abstract]. Arthritis Rheum 2010;62 Suppl 10 :1879
DOI: 10.1002/art.29644

Abstract Supplement

Meeting Menu