To the Editor: Langton et al claim that “the celecoxib studies have not demonstrated an increased risk of thrombosis”.1 However, the European Agency for the Evaluation of Medicinal Products concluded that “there is a trend towards a higher MI [myocardial infarction] risk associated with the use of celecoxib compared with naproxen and diclofenac”, and decided that a warning statement was required for all cycloxygenase 2 (COX-2)-selective drugs.2 It is surprising that celecoxib may be worse than diclofenac, because diclofenac is similarly COX-2-selective as celecoxib.3,4

A retrospective analysis of the full CLASS study data for people not taking aspirin found the rates of serious thromboembolic cardiovascular events were celecoxib 1.4% and diclofenac 1.6%, as against 0.7% for ibuprofen.5 These differences were not individually statistically significant, but CLASS was underpowered for cardiovascular events. However, pooling the results for the two similarly COX-2-selective drugs versus ibuprofen reveals a significant difference (relative risk [RR], 2.1; 95% CI, 1.1–3.9). In the full CLASS data, celecoxib did not have a lower rate of complicated ulcers (RR, 0.83; 95% CI, 0.46–1.5) and there was a trend towards more serious adverse events of all types (RR, 1.17; 95% CI, 0.99–1.39) than in the combined ibuprofen and diclofenac groups.3,6

We conclude that the case against all COX-2-selective drugs has not been proven beyond doubt because they have not been studied adequately. However, on the balance of probabilities, they are all likely to have a similar propensity to rofecoxib to increase thrombotic cardiovascular events to some extent. This prothrombotic effect may be reduced by combining them with aspirin, but then the main gastrointestinal benefit is likely to be lost,3,4 so use of such combinations is not justified.

Overall, celecoxib is no more effective, more expensive, no safer (and possibly less safe) than non-selective drugs. Meloxicam has not been shown to be any better. COX-2-selective drugs should not be used unless a subpopulation can be identified for whom these drugs have an advantage over the non-selective drugs. In theory, COX-2-selective drugs may be useful for a tiny group of people who are at greater risk of serious harm from gastrointestinal injury than from vascular events. However, there is no proven way to identify such people and there are no relevant trials to guide us. For example, no trials have been done in patients with a history of peptic ulcer.

All COX-2-selective drugs should be removed from the market until they have been properly evaluated.