Adverse effects: the use of COX2 inhibitors for arthritis has been questioned.

The ability of the cyclooxygenase-2 (COX2) inhibitors rofecoxib (Vioxx; Merck) and celecoxib (Celebrex; Pfizer) to treat arthritis with fewer side effects than traditional non-steroidal anti-inflammatory drugs (NSAIDs) has been questioned by the French health products safety agency Agence Française de Sécurité Sanitaire des Produits de Santé (AFSSAPS).

As Anne Castot, head of AFSSAPS's co-ordination and management of pharmacovigilance explains, “Concerns about high prescribing levels of COX2 inhibitors, including for elderly patients and those at risk of gastrointestinal (GI) problems, led us to initiate re-evaluation at European Union (EU) level. Considering data from pharmacovigilance and clinical trials worldwide, the benefit/risk ratio remains positive, but prescribing and labelling need caution.”

The decision highlights the worldwide debate about the use of COX2 inhibitors. The UK drug watchdog, the National Institute for Clinical Excellence (NICE), recommended in 2001 that all NSAIDs be used cautiously, and COX2 inhibitors preferentially only in patients at high risk of GI side effects. In the United States, COX2 inhibitors are used more widely for pain relief, and US $6 billion annually is spent on rofecoxib and celecoxib alone. But fears have recently been raised about reports linking rofecoxib to heart attacks. Some insurers now insist that doctors get 'prior approval' before prescribing such drugs.

This stems in part from a blurring of the divide between the roles of COX1 and COX2. The central idea that inhibition of COX2 is responsible for the therapeutic actions of the drugs, whereas COX1 inhibition is responsible for the side effects, is still valid, although there are some caveats. Reports have indicated that COX2 is more widespread than first suspected and that it is not necessarily restricted to inflammatory sites. For example, the kidney contains high levels of COX2 under 'normal conditions', and so COX2 inhibitors could cause hypertension and fluid retention in some patients.

Jon Deeks, Senior Medical Statistician at the Institute of Health Sciences, Oxford, systematically reviewed COX2 inhibitors (Deeks, J. J. et al. BMJ 325, 619 (2002)), and says there is a general problem in assessing the two large trials — VIGOR for rofecoxib and CLASS for celecoxib — which investigated COX2 inhibitors and serious GI adverse events compared with NSAIDs.

“CLASS found fewer adverse events with celecoxib versus NSAIDs during the trial's first six months,” says Deeks. But analysis of the data at 12 months is more complicated due to issues such as patient drop-out. Therefore, Deeks believes that the six-month data should be the primary consideration, as other trials strongly indicate that COX2 inhibitors favourably alter outcomes such as endoscopically detectable ulcers. “Australia's decision is most closely linked to the evidence,” says Deeks, “approving them for six months based on evidence from CLASS, but with less commitment to longer treatment.”

Rod Flower, Professor of Biochemical Pharmacology at St Bartholomew's and the Royal London School of Medicine and Dentistry, says that CLASS, VIGOR and other trials, “strongly suggest that selective COX2 inhibitors produce less side effects.”

But he adds that these studies are performed in carefully monitored patient groups and that patients also often self-medicate with over-the-counter NSAIDs to try to achieve a better effect, which eliminates the potential benefits of selective inhibition. “When released into the general population, polypharmacy may result in these drugs having no apparent significant benefit. Perhaps this underlies the French announcement?”