ROCHESTER, Minnesota—A retrospective, population-based cohort study by Mayo Clinic researchers of glucocorticoids (GC) and cardiovascular (CV) events in patients with rheumatoid arthritis (RA) concludes that steroid exposure increases CV risk in patients who are rheumatoid factor (RF)-positive but not in patients who are RF-negative. The report by John M. Davis, III, MD, et al is in Arthritis & Rheumatism.¹

"These findings suggest that glucocorticoids interact with RF status to modulate the occurrence of CV events in patients with RA," writes Dr. Davis, with the division of rheumatology at the Mayo Clinic, in Rochester, Minnesota.

Study Correlates GC Exposure and CV Events

The study included analysis of the medical records of 603 adult patients with incident RA between 1955 and 1995, with a median follow-up of 13 years. GC exposure was analyzed as tertiles of cumulative exposure and defined as recent use (<3 months) versus past use (>3 months), and as an average daily dose (< or >7.5 mg/d). CV events were defined as myocardial infarction, heart failure, or death from CV causes. Severity of RA was controlled for by analyses for interaction between cumulative GC exposures and presence or absence of erosions, rheumatoid nodules, or any joint surgery.

This analysis showed that RF-negative patients were at no increased risk of CV events regardless of cumulative GC dose, but RF-positive patients' CV risk increased with GC dose, particular with cumulative doses >7000 mg, with average daily doses of >7.5 mg, or with GC exposure within the past 3 months.

Dr. Davis suggests that the absence of effect in the RF-negative patients suggests that "the typically cited adverse effects of glucocorticoids (insulin resistance, increased blood pressure, etc) are relatively less consequential. Otherwise, one would expect these potentially harmful effects to similarly affect the entire RA cohort." He speculates that the increased CV risk in RF-positive patients is related to their higher levels of inflammatory activity.

The authors conclude, "Our study is the first to demonstrate that glucocorticoids interact with RF status to modulate the occurrence of CV events among individuals with RA. The mechanisms underlying this interaction are unknown, and further research is needed to fully explicate the relationship between glucocorticoid exposure and cardiac disease in patients with RA."

But is the RF difference real, and does it matter?

Miguel A. Gonzalez-Gay, MD, PhD, who most recently reviewed GC-related CV and cerebrovascular events in rheumatic diseases in Arthritis Care & Research² told CIAOMed that the Mayo Clinic study is subject to the limitations of all retrospective studies and that the presence of RF may be largely a marker of disease severity, including higher risk of extraarticluar manifestations.

"Since seropositive RA patients have generally more severe disease, it is logical to think that this subgroup of patients may have a higher incidence of cardiovascular events, regardless of steroid therapy. Due to this, results like these should be very carefully considered. I think it is difficult to exclude the problem of confounding by indication," said Dr. Gonzalez-Gay, with the rheumatology division at Hospital Xeral-Calde in Lugo, Spain.

"I think steroids are potentially useful in RA, in particular when they are prescribed in early phases of the disease. They have a number of beneficial effects and have not been associated with atherosclerosis in my studies on RA using echo-ultrasound. I do not think [steroids] should be used in a different way according to RF status. They must be used taking into account disease severity (regardless of RF result)," Dr. Gonzalez-Gay added.

E-mail any comments to .

References

1. Davis JM, III, Kremers HM, Crowson CS, et al. Glucocorticoids and cardiovascular events in rheumatoid arthritis. A population-based cohort study. Arthritis Rheum. 2007;56:820-830.
2. Gonzalez-Gay MA. Glucocorticoid-related cardiovascular and cerebrovascular events in rheumatic diseases: myth or reality? Arthritis Care & Research. 2007;57:191-192.