Medical journals should seek to publish papers that stimulate debate
An earlier publication of Graham's paper, and the debate it would
have generated, may have saved many from the adverse effects of rofecoxib
(1). Peer-reviewed journals can play an important role in creating
awareness of drug adverse effects by encouraging debate on the possible
hazards of pharmaceutical products. White and colleagues in analyzing 10
trials involving 7934 patients with osteoarthritis or rheumatoid arthritis
suggested that valdecoxib, and the non-steroidal anti-inflammatory drugs
(NSAID), have no adverse cardiovascular effects, when in fact their data
actually show the opposite (2).
The data presented is clearer if the percentage incidence is
presented to three decimal places rather than rounded up to a single
decimal point as in the original paper. In considering all the 7934
patrients reviewed, the cardiovascular event rate with placebo is 0.175%,
that with NSAID is 0.575%, with valdecoxib 10mg is 0.324% and with
valdecoxib 40 mg is 0.562%. Amongst non-aspirin users, cardiovascular
event rate is 0% in placebo, 0.359% with NSAID, 0.151% with valdecoxib 10
mg, 0.224% with valdecoxib 20 mg, and 0.324% with valdecoxib 40 mg.
Thus, it is clear that the cardiovascular event rate with valdecoxib is
higher than in the placebo group, and probably increases the higher the
dose of drug used. Incidence from the valdecoxib 80 mg group may not be
reliable since much smaller number of patients were studied. The adverse
cardiovascular effect of valdecoxib 40 mg appears equivalent to that of
Because of its cardioprotective effect, therapy with low-dose aspirin
appears to abolish the adverse cardiovascular effect of the coxibs, making
the cardiovascular event rate of aspirin users equivalent whether the
patient is on placebo, valdecoxib or NSAID (3). The smaller number of
such patients also make these data less conclusive.
The mortality data confirms the suspicion that valdecoxib and the
NSAID do have adverse cardiovascular effects. There were 4 cardiovascular
deaths in the valdecoxib group, 3 cardiovascular deaths in the NSAID group
and none in the group on placebo. The adverse cardiovascular effect of the
coxibs may be due its inhibition of prostaglandin I2 formation, with
resulting unfavorable effects on platelet aggregation and endothelium
Given the influential role peer-reviewed journals have, the absence
of a feedback mechanism for debating published articles will add to the
problems highlighted by Fontanarosal et al and further delay recognizing
the hazards of pharmaceutical products (5).
1. 1. Graham D, Campen D, Hui M, Spence M, Cheetham C, Levy G, et al.
Risk of acute myocardial infarction and sudden cardiac death in patients
treated with cyclo-oxygenase-2 selective and non-selective non-steroidal
anti- inflammatory drugs: nested case-control study. Lancet
2005;365:Published online Jan 25, 2005.
2. White WB, Strand V, Roberts R, Whelton A. Effects of the
Cyclooxygenase-2 specific inhibitor Valdecoxib versus Nonsteroidal anti-
inflammatory agents and placebo on cardiovascular thrombotic events in
patients with arthritis. Am J Ther 2004; 11: 244-50.
3. Topol EJ, Falk GW. A coxib a day won’t keep the doctor away.
Lancet 2004; 364: 639-40.
4. FitzGerald GA. Coxibs and cardiovascular disease. N Engl J Med
2004; 351: 1709-11.
5.Fontanarosa PB, Rennie D, DeAngelis CD. Postmarketing surveillance-
Lack of vigilance, lack of trust. JAMA 2004; 292: 2647-50.
Competing interests: No competing interests