Intended for healthcare professionals


Efficacy and safety of COX 2 inhibitors

BMJ 2002; 325 doi: (Published 21 September 2002) Cite this as: BMJ 2002;325:607

New data are encouraging but the risk:benefit ratio remains unclear

  1. Roger Jones (roger.jones{at}, Wolfson professor of general practice
  1. Department of General Practice and Primary Care, Guy's, King's, and St Thomas's School of Medicine, London SE11 6SP

    Papers pp 619, 624

    Non-steroidal anti-inflammatory drugs (NSAIDs) reduce pain and improve function in people with mechanical and inflammatory arthropathies and are beneficial in many other conditions, but these benefits come at a price. In the United Kingdom, every year over 2000 people die as a result of upper gastrointestinal damage induced by NSAIDs, and these agents can also have unwanted effects on the lower bowel, lungs, kidneys, and cardiovascular system. Conversely, some NSAIDs may have useful antithrombotic actions and increasing evidence shows that they may inhibit the development of colonic neoplasia and other gastrointestinal cancers.13 The introduction of new anti-inflammatory agents, with more specific inhibitory effects on cyclo-oxygenase 2 (COX 2) pathways, promised equivalent efficacy with greater safety and tolerability.

    Two large pivotal trials have been published, in which the efficacy and safety of the COX 2 inhibitors celecoxib and rofecoxib were compared with various traditional NSAIDs. Unfortunately the celecoxib long term arthritis safety study (CLASS),4 in which celecoxib was compared with ibuprofen and diclofenac and found to be associated with a lower incidence of symptomatic ulcers and complications related to ulcers, became the subject of almost unprecedented criticism on the grounds of design of study, analysis of data, and selective presentation of results. 5 6 In the Vioxx gastrointestinal outcomes research (VIGOR) trial rofecoxib was compared with naproxen in patients with rheumatoid arthritis.7 Significantly fewer clinically important upper gastrointestinal side effects occurred in the rofecoxib group, but an unexpected substantial excess of serious cardiovascular events occurred among the rofecoxib patients.8 Clinicians were left with numerous questions. Are the costs of these new drugs justified in terms of their greater tolerability and safety? Are they as effective as traditional NSAIDs? Are they really safe? Are they appropriate for patients with histories of upper gastrointestinal symptoms or ulceration, and are they a better alternative to co-prescription of a proton pump inhibitor in protecting against upper gastrointestinal damage?

    Some answers to these questions may have emerged from two studies published in this issue (see pp 619, 624). Deeks and colleagues report the findings of a systematic review and meta-analysis of randomised trials comparing celecoxib with a traditional NSAID or placebo.9 They identified nine trials including 15 172 patients with osteoarthritis and rheumatoid arthritis, in which celecoxib was compared with at least one NSAID (diclofenac, naproxen, or ibuprofen) or with a placebo (five trials). CLASS contributed over half of the patients analysed. They found equivalent efficacy between celecoxib and the comparator NSAIDs, but significantly greater tolerability, in terms of withdrawals from studies as a result of gastrointestinal adverse effects, with celecoxib and a lower incidence of upper gastrointestinal complications, including symptomatic ulcers, perforation, and haemorrhage. These results seem also to apply to the subgroup of patients taking low dose aspirin as antithrombotic prophylaxis. This study was not able to examine longer term sequelae, did not comment on deaths, and did not analyse cardiovascular events.

    In an observational cohort study in Canada, Mamdani and colleagues compared rates of upper gastrointestinal haemorrhage resulting in admission to hospital in patients aged over 66 years who started treatment with either traditional NSAIDs or COX 2 inhibitors, with each other and with a large control group.10 Over a follow up period of less than six months it seems that the risk of upper gastrointestinal haemorrhage for the selective COX 2 inhibitors is significantly lower than for the traditional non-selective NSAIDs, and that celecoxib seems to be associated with a lower risk of bleeding than rofecoxib. But the study does not contain information about death rates, does not address the question of gastrointestinal haemorrhage managed outside hospital or causing otherwise unexplained sudden death in elderly people, nor does it include data on other non-gastrointestinal side effects of NSAIDs.

    These results may provide some comfort, but many questions remain unanswered. It may, for example, be inappropriate to regard traditional or new NSAIDs as homogeneous comparator groups in these studies, because some drugs may have greater antithrombotic properties than others. Assuming equivalent efficacy, the risk to benefit ratio for COX 2 inhibitors depends critically on the cumulative effects of other, non-gastrointestinal side effects, for which the data remain controversial. At present it is still difficult to give patients an honest, accurate, and understandable account of the balance between relief of pain and improved function on the one hand, and the likelihood of serious adverse effects on the other. The National Institute for Clinical Excellence provides useful guidance on the prescription of COX 2 inhibitors, emphasising their most appropriate use in high risk patients, in whom cost effectiveness is also likely to be maximised.11 More information is, however, still required for prescribers to be able to make rational decisions about the use of these agents, particularly in older people in whom comorbidity is common and for whom the stakes are high. The stakes are also very high for the manufacturers of these drugs, who must ensure the highest standards of research governance in future studies.


    • Competing interests RJ has received fees for speaking and consultancy from Astra Zeneca and Wyeth Pharmaceuticals.


    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    9. 9.
    10. 10.
    11. 11.
    View Abstract