Study shows how COX-2 inhibitors lead to heart disease

University of Pennsylvania School of Medicine researchers have found additional evidence that may help explain how selective inhibitors of COX-2 might predispose individuals to heart disease and stroke.

In Circulation Research, they report that a COX-2-derived fatty substance-a prostaglandin called prostacyclin-controls the blood-vessel response to stresses such as high-blood pressure, thereby further linking COX-2 inhibitors to an increased risk of heart attack or stroke. This knowledge, along with a growing literature on physiological responses to COX-2 inhibitors, should help in the development of a rational approach to clinical risk management for this class of drugs.

Two randomized trials of COX-2 inhibitors-the gold standard of clinical evidence-conducted in 2004 at other institutions suggested that risk of cardiovascular disease might increase gradually during continued treatment with drugs such as Celebrex and Vioxx, even in individuals initially at low risk of the disease.

"The risk of heart attack and stroke became progressively evident during treatment with either Celebrex or Vioxx during the APPROVe and APC trials last year," says Garret FitzGerald, MD, lead author of the study published online last week. FitzGerald is the Director of the Institute for Translational Medicine and Therapeutics at Penn.

These studies were designed to determine whether COX-2 inhibitors limited the development of benign growths in the large bowel of patients who-to the best of study authors' knowledge-were at low risk of heart disease. "While the results of these trials are not conclusive, they are compatible with a gradual transformation of increased cardiovascular risk during continued dosing with either Celebrex or Vioxx," says FitzGerald. "We need to determine how this might occur, and whether we can manage this risk by developing tests that reflect the process."

Earlier animal studies by Penn researchers and others showed that suppression of the protective fat prostacyclin, which is made by COX-2, could predispose individuals to a rise in blood pressure which, in turn, can accelerate hardening of the arteries, or atherosclerosis. COX-2 inhibitors such as older NSAIDs have been shown to raise blood pressure in people. In addition, the Penn group has shown in previous studies that shutting down prostacyclin hastens initiation and early development of atherosclerosis.

The current research expands on this notion. R. Daniel Rudic, PhD, and Derek Brinster, MD, and others in FitzGerald's laboratory, report that COX-2-derived prostacyclin also controls the changes that occur in the muscular lining of blood vessels in response to pressure-related changes in blood flow.

They used two animal models to test their ideas. In one, they looked at changes in a blood vessel that had been transplanted into mice of a different genetic make-up; in fact, the model mimicks the events of human organ transplant rejection. Here, they found that they had, in effect, removed a brake on the response of the blood vessel to the challenge of transplantation by deactivating prostacyclin by genetically deleting its receptor. The result was that muscle cells proliferated dramatically, which normally reduces the openness of the blood vessel. However, the openness of the blood vessel was not changed, through a process of structural reorganization of the blood vessel called vascular remodeling.

In the second model, they reduced blood flow in arteries in the neck and looked at the downstream effects in the blood vessel. This time, instead of suppressing prostacyclin receptor signaling by genetic deletion, they did so by giving a COX-2 inhibitor. Indeed, they saw the same effect. Cells in the muscular lining of the vessel wall multiplied (just like in the transplant model). Additionally, despite the vessel growth caused by the COX-2 inhibitor, openness of the blood vessel was again preserved. This occurred despite lower blood flow caused by the COX-2 inhibitor. Thus, prostacylin may act to remodel blood vessels to preserve adequate blood flow.

"What is really convincing here is how similarly the two models responded and how the genetics of the pharmacological approach to disrupting the effects of COX-2 had the same effect," says Rudic. In further studies-also described in the paper and performed in collaboration with Thomas Coffman, of Duke University-FitzGerald's group showed that the consequences of shutting down COX-2-derived prostacyclin could be limited, in part, by removing a receptor activated by thromboxane A2, the fatty product of COX-1 in platelets. This mirrors a similar balancing effect between COX-1 and COX-2, which has been noted in the case of blood clotting, blood pressure, and atherosclerosis. This suggests that suppression of thromboxane with low-dose aspirin could reduce the risk of heart disease if taking COX-2 inhibitors.

These findings suggest that during prolonged dosing with COX-2 inhibitors, several consequences of drug action-a rise in blood pressure, initiation, and early development of atherosclerosis, and now the architectural and functional response of blood vessels to such stress-could all interact in a reinforcing fashion to transform the risk of heart attack and stroke, even in previously healthy individuals. "We need to determine whether these mechanisms are operative in people, and if so, we should be able to develop tests which reflect this process," says FitzGerald. "This may allow us to detect the small number of individuals at risk of rapidly developing heart disease and stop the drugs before they run into trouble. We could also determine how quickly risk might dissipate on stopping the drugs. Certainly, the development of a rational approach to risk management will be key to giving Celebrex or other COX-2 inhibitors safely, even to healthy patients, for extended periods."