Cyclooxygenase-2--specific inhibitors: are they safe?

The basic tenet of the cyclooxygenase-2 (COX-2) hypothesis rests on the fact that sparing of inhibition of COX-1 should result in greater safety than if both COX isoforms are inhibited. This increase in safety should be most evident in those organs and tissues in which COX-1 alone has important, necessary physiologic functions (e.g., the stomach and platelets). Data from large clinical trials are now available to support the superior gastrointestinal safety of COX-2 inhibitors, not only for endoscopic endpoints but also for clinically significant outcomes. Additionally, lack of effect on platelets has been demonstrated at doses many times higher than being used clinically. Unfortunately, the COX-2 inhibitors still retain some of the side effects seen with traditional dual COX inhibitors (nonsteroidal anti-inflammatory drugs), namely, effects on the kidney that may manifest as an increased incidence of hypertension, edema, and associated clinical states. Similarly, effects on reproductive functions, endothelial function, and wound healing are theoretically possible but need to be evaluated in well-controlled clinical trials.