Precaution, cyclooxygenase inhibition, and cardiovascular risk James M. Ritter 1 , Idris Harding 2 and John B. Warren 3 1 Department of Clinical Pharmacology, King’s College London School of Medicine at Guy’s, King’s College & St Thomas’ Hospitals, St Thomas’ Hospital, Lambeth Palace Road, London SE1 7EH, UK 2 Division of Cardiac and Vascular Sciences, St George’s Hospital, Blackshaw Road, London SW17 OQT, UK 3 Top Flat, 25 South Audley Street, London W1K 2PB, UK Cardiovascular risk led to the withdrawal of Vioxx (rofecoxib) in 2004. Some related drugs also increase cardiovascular risk and cyclooxygenase (COX)-2 inhibi- tors that remain on the market, including unselective non-steroidal anti-inflammatory drugs (NSAIDs), have not been exonerated. This article reviews if new evi- dence should change clinical and regulatory practice. Substantial COX-2 inhibition increases the incidence of cardiovascular disease unless concomitant platelet thromboxane production is inhibited by >95%. This can be investigated using whole blood assays, an approach used recently to show that acetaminophen (paracetamol) is a COX-2-selective inhibitor. The epide- miology available suggests acetaminophen, though readily available over-the-counter, does increase cardi- ovascular risk. Current evidence is inadequate to recom- mend many potential alternatives to Vioxx as safe. We argue that the precautionary principle, ‘first do no harm’, should underpin the regulation and prescribing of NSAIDs. Labelling which identifies these risks for prescribers and consumers should be mandatory. Introduction Non-steroidal anti-inflammatory drugs (NSAIDs) are the most widely prescribed group of therapeutic drugs world- wide, and include many common analgesics and anti- inflammatory agents. They inhibit cyclooxygenase (COX) enzymes, which accounts for desirable and undesirable effects (Figure 1). The pharmacological profile depends upon the dose and selectivity for COX-1 or COX-2 of the particular agent. COX-1-selective inhibitors reduce plate- let TXA 2 synthesis and can be as used as anti-thrombotic agents; they can also be anti-inflammatory and analgesic, but often cause gastrointestinal toxicity. Non-selective COX-inhibitors are anti-inflammatory, analgesic and also cause gastrointestinal toxicity. COX- 2-selective NSAIDs such as rofecoxib, valdecoxib, pare- coxib and celecoxib, developed to reduce the burden of gastrointestinal toxicity, raise concerns regarding cardio- vascular safety, though the mechanism is not known. Possibilities include increased arterial blood pressure, increased atherogenesis, or increased thrombotic tendency [1–5]. To reduce platelet aggregation, platelet TXA 2 synthesis must be inhibited by >95% [6]. Low-dose aspirin achieves this with little effect on basal prostacyclin biosynthesis, [4] and is also anti-thrombotic [7]. Prostacyclin, synthesised mainly by COX-2 [8], has opposing effects to those of TXA 2 . Its biosynthesis is increased in atherosclerosis where it is protective [2,3,5]. Moncada and Vane postulated a balance between thromboxane and prostacyclin thirty years ago, though this concept has been disputed [9].A protective role for prostacyclin is supported by studies in mice deficient in the prostacyclin receptor [10], consistent with a prothrombotic action of COX-2 inhibitors through reduced synthesis of prostacyclin. COX-2 inhibition may also elevate blood pressure, accelerate atherosclerosis and undermine plaque stability [1,4]. PGE 2 , synthesised by COX-1 or COX-2, is a vasodilator, but may be proathero- genic [9]. The regulatory response to early signals of the cardio- vascular risk of Vioxx [11] was tardy and limited [12]. Drug regulators and prescribers take a relaxed view of the safety of many NSAIDs, but the implications for patient safety are considerable. There are plausible mechanisms for car- diovascular toxicity of NSAIDs consistent with the primary pharmacology of COX-2 inhibition. Improved biochemical assessments predict the COX-1/2 profile of NSAIDs, which appears to predict cardiovascular risk accurately. The precautionary principle ‘first do no harm’ applies to Euro- pean pharmaceutical regulation [13]. We argue that this should apply to NSAIDs as a class in terms of pharmacov- igilance and prescribing, unless data from appropriate clinical trials provide sufficient reassurance for specific products. We propose that drug regulation should mandate labelling of NSAIDs that identifies these risks to prescri- bers and consumers. This opinion piece reviews how much cardiovascular morbidity and mortality might have been attributable to Vioxx and whether such toxicity is idiosyn- cratic, or is shared with other NSAIDs including acetami- nophen. Quantifying the cardiovascular morbidity and mortality associated with Vioxx Vioxx was considered to be safe by the US Food and Drug Administration (FDA) in 2004, but withdrawn from the market by Merck in September that year because of poten- tial cardiovascular toxicity. In 2005, a Texan court awarded $250 million to the widow of Robert Ernst, who died while taking the drug (http://news.bbc.co.uk/1/hi/ business/4168332.stm). Estimates of the relative risk Opinion Corresponding author: Warren, J.B. (jbwarren@btinternet.com). 0165-6147/$ – see front matter ß 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.tips.2009.07.007 503