Bruce Rothschild, MD
Does long-term use of celecoxib increase risk for cardiovascular (CV) events? Is risk dose-dependent?
Included studies compared celecoxib with placebo and had ≥3 years planned follow-up. Outcome was a composite of CV death, myocardial infarction, stroke, heart failure, or thromboembolic event.
Public literature was searched for randomized controlled trials (RCTs), and the National Institutes of Health and Pfizer were contacted. Investigators for each trial were asked for individual patient data; review authors adjudicated deaths and CV events using uniform definitions. 6 trials (n = 7950, mean age 59 to 75 y, median follow-up 5 to 37 mo) in patients with conditions other than arthritis were included in the intention-to-treat meta-analysis.
Meta-analysis showed that celecoxib increased risk for combined CV outcomes overall and with twice-daily doses (200 mg or 400 mg) (Table). CV risk increased with dose (linear trend from placebo to 400 mg once daily to twice-daily doses of 200 mg and 400 mg, P < 0.001); the effect was greatest for patients with high vs moderate vs low baseline CV risk (interaction, P = 0.034).
Celecoxib increases risk for cardiovascular events in patients with conditions other than arthritis. Risk is greatest with higher doses of twice-daily celecoxib.
Celecoxib vs placebo for composite of CV events*
*CV = cardiovascular; HR = hazard ratio; other abbreviations defined in Glossary. Composite of CV events adjusted for baseline CV risk: CV death, myocardial infarction, stroke, heart failure, or thromboembolic event.
†Analyses adjusted for baseline CV risk exclude 1 trial (n = 86).
Rothschild B. Review: Celecoxib increases risk for cardiovascular events; risks are greatest with higher doses. Ann Intern Med. ;149:JC3–15. doi: 10.7326/0003-4819-149-6-200809160-02015
Download citation file:
Published: Ann Intern Med. 2008;149(6):JC3-15.
Acute Coronary Syndromes, Cardiology, Coronary Heart Disease, Coronary Risk Factors, Emergency Medicine.
Results provided by:
Copyright © 2019 American College of Physicians. All Rights Reserved.
Print ISSN: 0003-4819 | Online ISSN: 1539-3704
Conditions of Use