Statin effect on thrombosis not so positive
medwireNews: Statins may not prevent venous thromboembolic events despite previous research suggesting otherwise, say researchers.
A meta-analysis of published and unpublished randomized controlled trials revealed that in 22 trials of statin versus control treatment, venous thrombosis occurred in 0.9% of people on statins compared with 1.0% of people not taking the drugs. This suggests that statins have little if any effect on venous thomboembolic events, report Kazem Rahimi (University of Oxford, UK) and team.
These results did not change when the authors excluded the findings of the JUPITER trial, a randomized controlled analysis conducted in 2009, which demonstrated that rosuvastatin halves the risk for venous thromboembolic events in apparently healthy adults.
Speaking to medwireNews, Rahimi remarked: "Unlike previous reports suggesting statins may reduce the risk of clots developing in the veins, this study suggests that there is no such effect, or if there were an effect, it is very small. Therefore, it suggests that there is no need to change guidelines."
He explained that despite the strikingly positive findings of the JUPITER trial and other similar studies, most current guidelines are cautious and say these findings may be due to chance. "In light of our study, I'd assume that the guidelines would say that statins should not be prescribed for the prevention of venous thromboembolic events, unless further studies change that," he said.
As reported in PLoS Medicine, there was no evidence of heterogeneity among the trials in terms of statins' effects on either deep vein thrombosis or embolism in the 22 trials of statins versus control medication.
In the seven trials that compared a more intensive with a standard statin regimen, there was no evidence that high-dose statin therapy (80 mg/day) reduced the risk for venous thromboembolic events compared with standard dose statin therapy (10-40 mg/day).
In an accompanying editorial, Frits Rosendaal (Leiden University Medical Center, the Netherlands) says that although the study findings are more credible than observational findings, one of its limitations is bias due to outcome misclassification.
Commenting on this, Rahimi conceded: "I think in principle he is right and this is something that we discuss in our paper. There might be some bias, but the magnitude of the bias is going to be minimal. That does not change the overall result nor the conclusion of our findings.
"Even if a quarter of the events we've collected were completely wrong, that would change the order of magnitude only by 3%."
medwireNews (www.medwire-news.md) is an independent clinical news service provided by Springer Healthcare Limited. © Springer Healthcare Ltd; 2012
By Piriya Mahendra, medwireNews Reporter