Rheumatology Advance Access originally published online on November 28, 2006
Rheumatology 2007 46(4):561-562; doi:10.1093/rheumatology/kel398
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
EDITORIALS |
Celecoxib and CVS risk—lessons from the APC and PreSAP studies
Department of Rheumatology, University Hospital of North Durham, North Road, Durham
Correspondence to: D. J. Armstrong. E-mail: oswald17727{at}hotmail.com
Since the return of Robin Goodfellow to the land of the faeries, we mortal rheumatologists have had to scan the general medical press ourselves for items of arthritic interest. One wonders what Robin might have made of the two large trials of celecoxib for the prevention of colonic adenomatous polyps recently published in the New England Journal of Medicine [1, 2], and of the generally negative view of the accompanying editorial [3] with respect to balancing risks and benefits. If coxibs are too dangerous to use to prevent pre-malignant polyps (he might have asked), can one safely assume they are too dangerous to prescribe for sore knees?
A total of over 3000 patients who had previously had at least one colorectal adenoma removed took part in the Adenoma Prevention with Celecoxib (APC) and Prevention of colorectal Sporadic Adenomatous Polyps (PreSAP) studies, designed to examine the effect of daily celecoxib use on the recurrence of polyps. In APC [1], 2035 patients were randomized to receive placebo, 400 mg daily or 800 mg daily of celecoxib. Thirty per cent of each group also took low dose aspirin. Colonoscopy was performed 1 yr and 3 yrs after randomization, and patients were closely monitored for serious cardiovascular complications. Recurrence of adenomata over 3 yrs was reduced from 60.7% in placebo to 43.2% and 37.5%, respectively in the two celecoxib groups, but with an odds ratio for CVS events of 2.6 (1.1–6.1) and 3.4 (1.5–7.9). As a result of interim analysis suggesting increased CVS events in the celecoxib arms, the study was discontinued early [4]. However, when patients receiving celecoxib at either dose were divided into those with established coronary or cerebrovascular disease and those without, the findings were noticeably different—8.8% of patients with established atherosclerotic disease had a CVS event compared with only 2.1% of those without. The figure of 2.1% was higher than placebo (0.7%) in the healthy group, but still lower than placebo in those with established CVS disease (3.0%). Notably, the incidence of CVS events was considerably lower than that of gastrointestinal ulceration or haemorrhage—over 11% in all three groups.
In the PreSAP study [2], 1738 patients were enrolled, and randomized to receive either celecoxib 400 mg daily or placebo. Aspirin use was less than in the APC project (around 17% in both groups), but endoscopic and CVS surveillance was similar. Although an interim analysis did not show any statistically significant increase in CVS events, the study was ended prematurely at the same time as the APC research. In terms of efficacy for prevention of adenoma recurrence, the results confirmed those of the previous study—cumulative rate of adenoma recurrence fell from 49.2% ± 2.2 in the placebo group to 33.7% ± 4.1 in the celecoxib arm, and indeed the incidence of advanced adenomas was almost halved. The picture for development of serious CVS events was also similar, but not statistically significant—1.9% of the placebo group compared with 2.5% of the celecoxib arm, an odds ratio of 1.30 (0.65–2.62). Although it is more difficult to tease out the effect of previous CVS history within the placebo and celecoxib arms from the figures, in the groups together, the arrow was pointing in the same direction as the previous study—6.1% of those with a CVS history suffered a serious CVS event, compared with 1.7% of those without.
The conclusions of the editorialists are unarguable. Regular colonoscopy and removal of polyps is associated with negligible CVS risk, while taking regular celecoxib is associated with a (small) risk. Although the studies confirm that coxibs significantly reduce the risk of polyp recurrence, patients will need regular endoscopic monitoring anyway, at which polyps will be removed, therefore, the routine use of celecoxib is not recommended. Indeed, their hypothetical risk–benefit analysis confirmed one would be better taking aspirin, which will be good for the heart as well as the bowel.
What can the jobbing rheumatologist learn from this? Several points stand out. In the first instance, patients taking 800 mg of celecoxib a day had no more gastrointestinal ulcers or haemorrhages than placebo (in fact they had marginally fewer), and even those also taking aspirin had only a small and non-significant increase over placebo. In the absence of CVS worries, one imagines this would be regarded as a major finding from the studies. Secondly, for patients without IHD taking 400 mg celecoxib daily (the absolute maximum recommended dose), while the incidence of serious CVS events is greater than placebo, it is still at the 1–2% level (compared with over 10% who will have an ulcer or bleed). Even in patients with established IHD at entry (a group in whom most doctors would not now consider prescribing a coxib), and who received 800 mg of celecoxib daily for 3 yrs (a sustained and substantial overdose), more than 91% suffered no serious CVS side effect. Finally, many rheumatology prescriptions of celecoxib are for 200 mg daily or on a PRN basis, which the studies do not directly address.
Evidence for an increased cardiovascular risk with long-term use of coxibs (and almost certainly other NSAIDs) is substantial, but these studies go some way towards properly quantifying it, and should be viewed with the perspective they deserve. Few outside the rheumatology community appreciate the substantial improvements in pain control and quality of life that an effective anti-inflammatory can make in both rheumatoid and osteoarthritis, even in patients on quantities of immunosuppressive or corticosteroid therapy. Most of us have patients who still rue the withdrawal of Vioxx, which appeared almost unparalleled in efficacy.
The use of NSAIDs is contra-indicated in patients with active peptic ulcer disease. Even in those in whom there is no known history, 3.5% will develop a symptomatic ulcer in the first year of taking the full dose regularly, rising to 6.0% in patients who also take aspirin [5]. We therefore ask patients specifically about ulcer disease, and may consult their past records, before considering the use of an NSAID. In exactly the same way, with a well-quantified risk, we can enquire about a patient's CVS history, and involve the patient in the discussion of the risks and benefits of using a coxib. Some areas require clarification; for example, the influence of CVS risk factors (as opposed to established disease) remains open to debate, although most physicians would be reluctant to prescribe a high dose of anti-inflammatory to an obese middle-aged smoker with hypertension, even if there was not yet established disease. Nevertheless, knowledge is power, and the more knowledge we gain about COX-2 inhibitors (and we await the results of the MEDAL study into the effects of etoricoxib with interest), the better able we are to prescribe them with confidence to the correct patients. In many ways, these are remarkably safe and effective drugs, and correctly prescribed, deserve their place in the formulary for the foreseeable future. I hope Robin would agree ...
D.J.A. has received honoraria, speakers fees and unrestricted educational grants from MSD, Pfizer and Novartis.
References
- Bertagnoli MM, Eagle CJ, Zauber AG, et al. (2006) Celecoxib for the prevention of colorectal adenomas. N Engl J Med 335:873–84.
- Arber N, Eagle CJ, Spicak J, et al. (2006) Celecoxib for the prevention of adenomatous polyps. N Engl J Med 355:885–95.
[Abstract/Free Full Text] - Psaty BM and Potter JD. (2006) Risks and benefits of celecoxib to prevent recurrent adenomas. N Engl J Med 355:950–2.
[Free Full Text] - Soloman SD, McMurray JJ, Pfeffer MA, et al. (2005) Cardiovascular risk associated with celecoxib in a clinical trial for colorectal adenoma prevention. N Engl J Med 352:1071–80.
[Abstract/Free Full Text] - Gastrointestinal toxicity with celecoxib vs nonsteroidal anti-inflammatory drugs for osteoarthritis and rheumatoid arthritis: the CLASS study: a randomized controlled trial. Celecoxib Long-term Arthritis Safety Study. JAMA (2000) 284:1247–55.
[Abstract/Free Full Text]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||