"Although these results are in mice, not people, they raise an exciting possibility which can be tested in humans," says senior author Garret FitzGerald, MD, Director of Penn's Institute for Translational Medicine and Therapeutics,
Ever since the association of selective inhibitors of COX-2 ?Vioxx, Bextra, and Celebrex ?with an increased incidence of heart attack and stroke, there has been intense interest in understanding the mechanism involved. Clarification of this issue offers the prospect of conserving the clinical benefit of these drugs for patients with arthritis, while managing the risk.
Almost 10 years ago, FitzGerald noticed that both Celebrex and Vioxx depressed in healthy individuals a protective fat called prostacyclin, while leaving unaltered a harmful one called thromboxane. This led him to predict that drugs in this class might confer a cardiovascular risk before either reached the US market.
In the present studies, the investigators used multiple genetically manipulated mice ?including mice that mimicked the impact of either COX-2 inhibitors or low-dose aspirin and compared them with treating healthy mice with COX-2 inhibitors, such as Celebrex. They found that genetic disruption of COX-2; inhibition of the enzyme by different inhibitors; and disruption of prostacyclin's effects by removing its receptor all had the same effect ?a predisposition to clotting and an elevation of blood pressure. "This provides compelling evidence in support of the original hypothesis," says co-author Colin Funk, PhD, who has collaborated with Fit zGerald at Penn over the last decade on this line of research. Funk is now the Canada Research Chair of Physiology at Queen's University, Ontario. "One does not need additional explanations to understand what we have seen in clinical trials. COX-2 inhibitors confer a small, but absolute cardiovascular risk using the same mechanism by which they relieve pain and inflammation".
The investigators also addressed the likely benefit of adding aspirin to diminish this effect of the inhibitors. Surprisingly, this appeared to reduce not only the clotting response, but also the rise in blood pressure caused by drugs like Celebrex.
"Despite some chatter to the contrary, this issue of an aspirin effect has not been addressed directly in any of the clinical trials of COX-2 inhibitors," says FitzGerald. "However, although these studies indicate that it would limit the cardiovascular risk, it would also be expected to add to the risk of stomach problems, undermining the reason for choosing COX-2 inhibitors in the first place."
A surprising finding came when the investigators turned to a drug target that might substitute for COX-2 ?an enzyme called microsomal prostaglandin E synthase (mPGES)-1. Other investigators had shown previously that deletion of this enzyme seemed as effective as treatment with NSAIDs in models of pain and inflammation. This has prompted several large pharmaceutical companies to develop drugs targeting this enzyme. Such inhibitors will soon enter human trials.
FitzGerald and his colleagues showed that deletion of mPGES-1, in contrast to deletion or inhibition of COX-2, did not predispose the animals to thrombosis or elevate blood pressure. A clue to this surprising finding was that while mPGES-1 deletion suppressed profoundly another product of the COX-2 pathway called PGE2, the deletion of mPGES-1 actually elevated prostacyclin, the complete reverse of what was observed with COX-2 inhibitors. "Selective inhibitors of mPGES-1 may retain much of the benefit of drugs like Vioxx and Celebrex, while diminishing the risk of heart attack and stroke by having precisely the opposite effect on prostacyclin," says FitzGerald.