The heart of the matter

Results of recent trials have added to the diversity of views in the on-going discussion about the relation of cardiovascular risk with cyclooxygenase (COX)-2-selective inhibitors (coxibs) (such as celecoxib and rofecoxib) and smoking.

A recent study by McAdam et al. challenged our view with regard to the increase in risk of cardiovascular (heart and blood vessel) disease associated with the use of coxibs. Previous trials such as the Vioxx Gastrointestinal Outcomes Research (VIGOR) and Adenomatous Polyp Prevention on Vioxx (APPROVe) indicated an increase in risk for myocardial infarction (heart attack) and stroke with the use of Vioxx (rofecoxib), raising concerns about the safety of other coxibs e.g. celecoxib. This was based on the assumption of a “class effect” of coxibs related to preferential inhibition of prostacyclin over thromboxane and thus a tendency toward clotting. Prostacyclin restrains platelet activation (for clotting) and therefore inhibiting it results in clotting; whereas on the other hand, thromboxane leads to platelet aggregation and inhibiting it reduces clotting. Selective inhibition of prostacyclin over thromboxane will therefore result in a tendency to clot, causing heart events; and the assumption that all drugs in the coxib class share this effect is the basis of the concern over the safety of coxibs other than rofecoxib (e.g. celecoxib). However, the results presented by McAdam et al. support the increasing body of evidence that this assumption is not only highly simplistic but is also lacking in support from available clinical data. Contrary to the conventional wisdom of increase in Thromboxane A2 from the COX-1 pathway in smokers, the authors demonstrated that COX-2 contributed to the production of Thromboxane A2 in smokers. Recent data also suggest that celecoxib (but not rofecoxib) actually improves the function of blood vessels when given to people with known cardiovascular diseases, by means of improving the availability of nitric oxide with resultant effects such as dilatation of blood vessels and reduced aggregation of platelets (reduced clotting). Different coxibs have different chemical structure and properties and therefore it remains undetermined whether all coxibs share the same effects. Recent studies show that even non-steroidal anti-inflammatory drugs (NSAIDs), which have considerable COX-1 effect and would be expected to have fewer cardiovascular side effects than coxibs, have been observed to have a trend toward increased heart events, therefore not supporting the hypothesis that only coxibs (that block COX-2) cause adverse effects to the heart. Celecoxib was found in one study to have no increased risk for heart attack and sudden cardiac death, but this risk was present for rofecoxib and several NSAIDs. The different effects of different coxibs are further demonstrated in research that shows that celecoxib and rofecoxib, both being in the coxib class, do not affect the same signal transduction pathway in cells. Smoking increases oxidative stress, causes inflammation, increases activation of platelet (causing clotting) and reduces the level of nitric oxide (a substance that expands the blood vessels and reduces clotting). Anti-inflammatory treatment with celecoxib reduces oxidative stress in patients with heart disease, but this effect was not observed for rofecoxib. In a recent study by McAdam et al., the administration of rofecoxib (which blocks COX-2 and will supposedly result in platelet activation) did not cause an increase in platelet activation, thus raising further questions about the hypothesis on how coxibs cause heart events by blocking prostacyclin and activating platelets. At least certain coxibs may exhibit clotting effect not only by reduction of prostacyclin but also by inhibition of the anti-clotting pathway. In summary, the McAdam et al study provides new insights into the role of different COX pathway substances (e.g. prostacyclin and thromboxane) in the effects on the heart from smoking. It challenges the hypothesis that the imbalance between thromboxane and prostacyclin alone accounts for all the adverse effects of coxibs and NSAIDs. It therefore remains unclear how selective in its COX-2 activity an individual drug has to be to predict its risk for causing adverse effects to the heart. The present study indicates that coxibs are associated with both harm and benefit. More research is needed to look at the net effect of an individual drug. None of the reported trials so far has specifically examined the cardiovascular outcomes of celecoxib or any other coxibs, resulting in many prior inappropriate conclusions. A large-scale safety trial involving more than 20,000 patients is currently planned to be carried out to address cardiovascular safety of coxibs and non-selective NSAIDs as well as the interplay of smoking, coxibs and cardiovascular risk. Until definite trials are completed, careful analysis is needed in determining the overall risk of the therapy to the patient. (Source: Circulation 2005;112:941-945.)