Possible increased risk of cardiovascular events attributable to an effective and popular class of anti-inflammatory drugs is raising considerable concern amongst the many thousands of patients taking these COX-2 inhibitors. Behind the rhetoric and scientific disputes of validity of the analyses, let's look at what the take-away points may be.

The medications under fire are Celebrex and Vioxx, the available non-steroidal anti-inflammatory drugs collectively known as NSAIDs. Their primary use is in the control of the inflammation and resultant pain from arthritis. NSAIDs work through the blocking of an enzyme called cyclooxygenase. Cyclooxygenase (COX) protects the stomach from the ulcer-producing effects of stomach acid. Not surprisingly, blocking COX with NSAIDs has been known to increase the risk of stomach ulcers.

It turns out that there are at least two COX enzymes, creatively called COX-1 and COX-2. The stomach one is COX-1, and the peripheral one, the one that benefits arthritis, is COX-2. Celebrex and Vioxx belong to the COX-2 class, and because of their presumed protection against ulcer formation, have been very popular to the tune of projected combined sales of $6 billion per year.

A provocative reanalysis of the patient databases that included aspirin, NSAIDs and COX-2 inhibitors was published last week in the Journal of the American Medical Association purporting to show at least a two fold increased risk of cardiac events in patients taking the COX-2 inhibitors. This has understandably raised many questions and concerns amongst patients taking and the doctors prescribing these medications.

First things first. This is NOT a stand-alone research study. It is an analysis of existing databases containing disparate populations of patients that may not be directly comparable. The fact that some of the studies were heavily weighted to patients with certain kinds of arthritis, which inherently may alter their concomitant illnesses and risks of heart disease, needs to be kept in mind.

Differences in cardiac event rates, even if validated in subsequent evaluations, do not necessarily mean cause and effect. By that I mean that the difference in heart attack risk can be from either of two reasons. One explanation is that a type of medication is actually causing the event. The second explanation is that one of the medications is protecting from the event and the remaining agent is actually neutral. Sort of the lowering the water versus raising the bridge to get a tall ship under the bridge model. All we suspect now from this report is that there may be a difference. It's a hypothesis that now needs validation by targeted research.

As a cardiologist, I have concerns almost daily about the NSAIDs, including the COX-2 inhibitors. In susceptible patients, there is the risk of exacerbation of high blood pressure, kidney problems, fluid retention and heart failure, as well as the usual litany of drug interactions.

Let's leave this subject with my usual caveat. As illustrated by the recent cerivastatin recall and this COX-2 controversy, all interventions have risk and benefit, some known and others unknown or undiscovered. It is the responsibility of the doctor and the patient to be aware of the clear indications and potential adverse effects of any therapy. That's what doctor-patient relationships should be.