Selective COX-2 inhibitors, eicosanoid synthesis and clinical outcomes: a case study of system failure

Elucidation of differences between the active sites of COX-1 and COX-2 allowed the targeted design of the selective COX-2 inhibitors known as coxibs. They were marketed as non-steroidal anti-inflammatory drugs (NSAIDs) that had improved upper gastrointestinal (GI) safety compared with older non-selective NSAIDs such as diclofenac and naproxen. Two GI safety studies conducted with arthritis patients demonstrated that in terms of upper GI safety, celecoxib was not superior to diclofenac (CLASS study) but rofecoxib was superior to naproxen (VIGOR study). However, the VIGOR study revealed also that rofecoxib had increased cardiovascular (CV) risk compared with naproxen. This clinical outcome was supported by the existence of plausible eicosanoid-based biological mechanisms whereby selective COX-2 inhibition could increase CV risk. Nevertheless, the existence of CV risk with rofecoxib was successfully discounted by its pharmaceutical company owner, Merck & Co, with the assistance of specialist opinion leaders and rofecoxib achieved widespread clinical use for 4-5 years. Rofecoxib was withdrawn from the market when several clinical trials in colorectal cancer and post-operative pain revealed increased CV risk with not only rofecoxib, but also coxibs. The commercial success of rofecoxib provides a case-study of failure of the medical journal literature to guide drug usage. Attention to ethical issues may have provided a more useful guide for prescribers.