Y a-t-il des différences de tolérance cardiovasculaire entre AINS classiques et coxibs?

Abstract

Three placebo-controlled studies have demonstrated deleterious cardiovascular (CV) effects of rofecoxib, celecoxib, and pare/valdecoxib. It remains to be determined whether this CV toxicity is specific to coxibs, or shared with all non-steroidal anti-inflammatory drugs (NSAID). Seven meta-analyses show that, in comparison with non-specific NSAIDs, the risk of thrombotic CV accident is increased with rofecoxib and celecoxib, but not with valdecoxib or lumiracoxib. Concerning the risk of thrombotic CV accident, seven of the ten observational studies which have evaluated the risk, have found an increased risk for the non-specific NSAID in comparison with non-exposed subjects,. The seven observational studies, which evaluated the risk of coxibs, have all found an increased risk with rofecoxib, and two with celecoxib. Three studies out of six have shown an increase of risk with rofecoxib and one study out of five with celecoxib. Two of the three studies, which have compared rofecoxib with celecoxib, have found an increased risk with rofecoxib. Concerning the risk of arterial hypertension, oedemas or congestive cardiac insufficiency, a meta-analysis and a randomised trial have shown a deleterious effect of rofecoxib in comparison with celecoxib and non-specific NSAID. Two studies have shown a deleterious effect of the non-selective NSAID and three a deleterious effect of rofecoxib in comparison with non-exposed subjects. Three studies have demonstrated a deleterious effect of rofecoxib in comparison with non-specific NSAID. No study has shown any deleterious effect of celecoxib in comparison with subjects non-exposed or exposed to non-specific NSAID. These studies suggest that all the NSAID, specific or not, increase the CV and renal risk. This risk seems variable from a compound to another one and must be evaluated, for each patient, according to the susceptibility and associated risk factors. While waiting for other long-term controlled studies, the available data show the existence of a risk of CV secondary effect linked to the class of NSAID, specific (coxibs) or not.