Editorial I: Postoperative NSAIDs and COX-2 inhibitors: cardiovascular risks and benefits

S. F. Jones

Waikato Hospital, Hamilton, New Zealand

I. Power*

School of Clinical Sciences and Community Health, The University of Edinburgh Royal Infirmary, Little France, Edinburgh EH16 4SA, UK

* Corresponding author. E-mail: Ian.Power{at}ed.ac.uk

"I would have everie man write what he knowes and no more."—Montaigne

The identification of a second isoenzyme of cyclo-oxygenase (COX-2), induced at the site of injury or inflammation, raised the prospect of analgesia without some of the adverse effects associated with the traditional non-steroidal anti-inflammatory drugs (NSAIDs). Preferential COX-2 inhibition was identified in some existing drugs, such as meloxicam, and a new family of drugs, the coxibs, was developed with still greater selectivity: celecoxib, rofecoxib and lumaricoxib exhibit COX-1:COX-2 inhibitory ratios of 1:30, 1:276 and 1:433, respectively. The selective COX-2 inhibitors were developed to exploit a minor conformational difference between the cyclo-oxygenase isoforms, binding to the same sites but able to be accommodated only by the COX-2 enzyme because of its distinguishing side pocket.

Theoretical concerns about a possible increase in the risk of adverse cardiovascular events appeared to be borne out when a 5-fold increase in the incidence of myocardial infarction (MI) was seen in the Vioxx Gastrointestinal Outcome Research (VIGOR),1 an increase in morbidity that substantially offset the reduction in major gastrointestinal complications. This study, using rofecoxib 50 mg daily for a median of 9 months in an intrinsically high-risk rheumatoid population, precluded the use of aspirin which, according to international guidelines, was indicated in a significant proportion of subjects. It was argued at the time that the findings did not represent a prothrombotic effect of rofecoxib but rather a protective effect of naproxen in the control population.

Epidemiological database studies, which reflect actual drug use and include higher-risk patients, also found a correlation between normal or high-dose rofecoxib use and adverse cardiovascular outcomes.2 3 Major limitations of these studies included the variable use of aspirin. Aspirin may protect against cardiovascular disease, but established use implies a higher baseline risk. Furthermore, some non-selective inhibitors, including ibuprofen and indomethacin, may impede access of aspirin to platelet COX-1 enzyme and inhibit this protective effect.4

Optimistic of an improved gastrointestinal risk–benefit ratio, the sponsoring companies sought to extend the indications for the selective COX-2 inhibitors. Their ability to inhibit both angiogenesis, upon which both local and metastatic tumour growth are dependent, and inflammation led to studies in oncoprophylaxis and Alzheimer's disease. These studies were placebo controlled, eliminating the confounding influence of active comparators. The finding of a 1.7-fold increased risk of MI or cerebrovascular accident with rofecoxib 25 mg daily compared with placebo in the APPROVe adenoma prevention study5 prompted Merck to withdraw Vioxx from the market. The increased risk represented an additional seven cases per 1000 patient-years and is consistent with the VIGOR results. In December 2004, Pfizer announced a significant dose-dependent 2.3- to 3.4-fold increase in serious adverse cardiovascular events in the interim analysis of their APC study evaluating the prevention of adenomatous polyps with celecoxib 200 mg or 400 mg twice daily.6 They emphasized that this result was not in keeping with an extensive database or with two other large long-term placebo-controlled studies: PreSAP, comparing celecoxib 400 mg daily with placebo, and ADAPT, the Alzheimer's disease and prevention with treatment study. The latter study found no increase with celecoxib 200 mg daily but a statistically significant increase in cardiovascular risk with naproxen, the non-selective NSAID which had purportedly attenuated the risk in the VIGOR study. However, one smaller study of Alzheimer's disease had also reported an increased composite cardiovascular risk with celecoxib 200 mg twice daily, although an unequal distribution of baseline risk favoured the placebo group.6 So why might these drugs be prothrombotic?

Several explanations have been proposed and it is likely that more than one mechanism contributes. Initially, the most prominent theory was of prostacyclin–thromboxane imbalance affecting the platelet–endothelial interface.2 These two eicosanoids, both products of the same cyclo-oxygenase-dependent arachidonic acid cascade, have mutually opposing vascular effects. Thromboxane, produced from platelets activated by exposure to adenosine, collagen or epinephrine, promotes haemostasis by vasoconstriction and promotion of platelet aggregation. On the other hand, endothelial prostacyclin is a powerful vasodilator and inhibitor of platelet aggregation. Low-dose aspirin protects against arterial thromboembolism as a result of a unique selectivity, inhibiting platelet thromboxane while leaving systemic prostacyclin synthesis intact. Aspirin achieves this through acetylation and irreversible covalent binding to platelet COX-1 within the portal circulation. Residual aspirin is converted to low doses of the relatively inactive salicylic acid on first pass through the liver. As platelets have no nuclei they cannot synthesize fresh cyclo-oxygenase and are inhibited for the remainder of their lifespan, although substantial offset of the effect occurs within 48 h.7 Platelets are not inhibited by selective COX-2 inhibitors as they do not express this isoform. However, endothelial prostacyclin, albeit counterintuitively, appears to be synthesized primarily by the ‘inducible’ COX-2 and therefore is inhibited by this group of drugs. Thus selective COX-2 inhibitors tip the balance in the opposite direction to aspirin, preserving haemostasis but potentially at the expense of increased vascular occlusion. In the case of the non-selective inhibitors, the loss of prostacyclin resulting from COX-2 inhibition is balanced by COX-1-mediated platelet inhibition. Interestingly, acetaminophen also preserves platelet function but has been shown to reduce systemic prostacyclin,8 thus exhibiting a cardiovascular profile resembling that of a weak COX-2 inhibitor. However, the clinical significance of this is unclear.

A second mechanism which might contribute specifically to the risk of MI is the effect that COX-2 inhibitors have on myocardial preconditioning. This is the phenomenon whereby an initial period of subcritical myocardial ischaemia offers both initial and more sustained protection against subsequent ischaemia, with a reduction in infarct size demonstrated in animal models. Certain drugs, including volatile anaesthetic agents, may also induce preconditioning, while others block the effect. A critical late step in the intracellular processing appears to be dependent upon COX-29 and is blocked by selective inhibitors as well as by traditional agents, although not by low-dose aspirin. Also within the myocardium, COX-2-derived prostaglandins have been shown to mediate the statin-induced attenuation of reperfusion injury.10 These and the third mechanism will be shared by selective and non-selective NSAIDs, with no modulating benefit from COX-1 inhibition.

The third mechanism which may account for an increased risk of cardiovascular and cerebrovascular complications is the cardiorenal mechanism.11 The cyclo-oxygenase inhibitors have profound effects on renal function. Salt and water retention, together with loss of vasodilator prostaglandins, increases blood pressure. A recent meta-analysis of 45 000 patients in 19 trials showed this to be greater with coxibs than with non-selective inhibitors and greater with rofecoxib than with celecoxib.12 The efficacy of antihypertensive medication, especially angiotensin-converting enzyme inhibitors and ß-blockers, is impaired. Congestive heart failure may be exacerbated or induced, as in a recent study which found odds ratios for hospital admission with congestive heart failure of 1.8 for rofecoxib and 1.4 for non-selective inhibitors.11 No increase was found with celecoxib. These conditions may indirectly increase the risk of myocardial ischaemia, MI and stroke. There is evidence from several papers that fluid retention, hypertension, heart failure and renal impairment are more marked with rofecoxib than with comparator coxibs or traditional NSAIDs,11 13 14 allowing the possibility that cardiovascular complications might be inherently associated with rofecoxib or, alternatively, reflect dose inequivalence.

A number of potentially significant pharmacokinetic and pharmacodynamic differences have been identified to account for the apparent lower risk with celecoxib. The shorter half-life of celecoxib may allow intermittent recovery of renal function and endothelial prostacyclin, especially with once-daily dosing as employed in the PreSAP and ADAPT studies. The lower selectivity of celecoxib might imply a degree of protective platelet inhibition. Furthermore, whereas high-dose rofecoxib may inhibit aldosterone metabolism and thus exacerbate hypertension, celecoxib inhibits the metabolism of carbonic anhydrase with a possibly protective diuretic consequence.12

Valdecoxib and the parenteral prodrug parecoxib have also been associated with a significant increase in mixed serious adverse events in the postoperative setting, including a proportional increase in cerebrovascular complications (2.9% vs 0.7%) and MI (1.6% vs 0.7%) after supramaximal dosing with 40 mg twice daily for 14 days following coronary artery bypass grafting (CABG).15 Four of the five MIs in the treatment arm occurred within 24 h of surgery. A subsequent study of 1636 CABG patients found a similar increase after therapeutic dosing for up to 10 days, whereas no increase was found in 1050 mixed general and orthopaedic patients.16 All the CABG patients were given low-dose aspirin.

Following considerable criticism over the delay in recognizing the potential risk of adverse cardiovascular events, the US Food and Drug Administration (FDA), in an advisory issued in April 2005, have required boxed warnings about possible increased cardiovascular risk for both selective COX-2 inhibitors and non-selective prescriptions. The rationale behind this was that COX-2 inhibitors collectively increased cardiovascular risk compared with placebo but not when compared with non-selective inhibitors.17 A recent case–control analysis supports this concern.18 Recent surgery for CABG is now listed by the FDA as a contraindication for both COX-2 inhibitors and traditional NSAIDs although, importantly, they commented that short-term use of NSAIDs otherwise does not appear to increase cardiovascular risk.17

The Committee on Safety of Medicines in the UK,19 the Therapeutic Goods Administration in Australia20 and Medsafe in New Zealand21 have also issued advice. This includes recommendations that selective COX-2 inhibitors should not be prescribed for patients with or with increased risk of cardiovascular disease,1921 that they should be used at the lowest dose and for the shortest time necessary,17 20 that moderate as well as severe congestive heart failure be considered a contraindication,19 that gastroprotective agents be considered for patients switched to non-selective NSAIDs,19 that a lower gastrointestinal risk has not been clearly demonstrated for those patients taking low dose aspirin 19 21 and that etoricoxib appears to be associated with more frequent and severe hypertension.19 21

The disparity between the increase in postoperative MI following CABG as opposed to general or orthopaedic surgery may reflect the recent creation of a coronary anastomosis in the former group and may be explained by a different mechanism for most postoperative MIs.22 In the general population acute MI generally results from thrombosis of an acutely ruptured, unstable and previously subcritically stenosing plaque, while postoperative MI occurs distal to a stable but more critically stenosing lesion and is initiated by prolonged haemodynamic stress, inducing a supply–demand imbalance. A longer duration of ischaemia results in a higher risk of adverse outcome.22 Therefore it is noteworthy that ketorolac, a non-selective inhibitor, has been shown to reduce haemodynamic stress and limit the duration of S–T segment changes following major arthroplasty.23 Furthermore, a large retrospective but non-randomized study of ketorolac compared with opioid therapy for acute pain found that the incidence of MI in the ketorolac cohort was less than half that in the opioid cohort.24 The clinical significance of these findings is uncertain, and it is not known whether any true differences are mediated via a reduction in haemodynamic stress, which is likely to be shared by selective COX-2 inhibitors, or via a platelet inhibitory effect. In view of these tantalizing suggestions of possible cardiovascular protection and a total absence of documented cardiovascular harm in the postoperative period, except after CABG, we contend that it may be premature to discourage the practice of balanced analgesia with opioid and cyclo-oxygenase inhibitors that has become established as the cornerstone of contemporary acute pain management. Further study is certainly warranted, although exclusion of high-risk patients may limit the validity of such research.

Meanwhile, we must avoid COX-2 selective agents in patients after recent CABG and probably also after other procedures with an arterial anastomosis, such as vascular surgery, free tissue transfer and solid organ transplantation. We should probably choose a non-selective NSAID in preference to a selective agent in patients with, or with conditions increasing the risk of, cardiovascular disease unless there are other overriding considerations. We must also remain cognisant of the other precautions, especially the potential for either group of drugs to impair renal function in the postoperative period. Conversely, we must not forget that there is morbidity as well as misery associated with unrelieved acute pain and potentially serious side effects from all other available analgesic modalities

References

1 Bombardier C, Laine L, Reicin A et al. Comparison of upper gastrointestinal toxicity of rofecoxib and naproxen in patients with rheumatoid arthritis. N Engl J Med 2000; 343: 1520–8[Abstract/Free Full Text]

2 Solomon DH, Schneeweiss S, Glynn R et al. Relationship between selective cyclooxygenase-2 inhibitors and acute myocardial infarction in older adults. Circulation 2004; 109: 2068–73[Abstract/Free Full Text]

3 Levesque LE, Brophy JM, Zhang B. The risk for myocardial infarction with cyclooxygenase-2 inhibitors: a population study of elderly adults. Ann Intern Med 2005; 142: 481–9[Abstract/Free Full Text]

4 MacDonald TM, Wei L. Effect of ibuprofen on cardioprotective effect of aspirin. Lancet 2003; 361: 573–4[CrossRef][ISI][Medline]

5 Bresalier RS, Sandler RS, Quan H et al. Cardiovascular events associated with rofecoxib in a colorectal adenoma chemoprevention trial. N Engl J Med 2005; 352: 1092–1102[Abstract/Free Full Text]

6 Solomon SD, McMurray JJV, Pfeffer MA et al. Cardiovascular risk associated with celecoxib in a clinical trial for colorectal adenoma prevention. N Engl J Med 2005; 352: 1071–80[Abstract/Free Full Text]

7 Sonksen JR, Kong KL, Holder R. Magnitude and time course of impaired primary haemostasis after stopping chronic low and medium dose aspirin in healthy volunteers. Br J Anaesth 1999; 82: 360–5[Abstract/Free Full Text]

8 Green K, Drvota V, Vesterqvist O. Pronounced reduction of in vivo prostacyclin synthesis in humans by acetaminophen (paracetamol). Prostaglandins 1989; 37: 311–15[CrossRef][Medline]

9 Bolli R. The late phase of preconditioning. Circ Res 2000; 87: 972–83[Abstract/Free Full Text]

10 Birnbaum Y, Ye Y, Rosanio S et al. Prostaglandins mediate the cardioprotective effects of atorvastatin against ischaemia-reperfusion injury. Cardiovasc Res 2005; 65: 345–55[CrossRef][ISI][Medline]

11 Mamdani M, Juurlink DN, Lee DS et al. Cyclo-oxygenase-2 inhibitors versus non-selective non-steroidal anti-inflammatory drugs and congestive heart failure outcomes in elderly patients: a population based cohort study. Lancet 2004; 363: 1751–6[CrossRef][ISI][Medline]

12 Aw T-J, Haas SJ, Liew D, Krum H. Meta-analysis of cyclooxygenase-2 inhibitors and their effects on blood pressure. Arch Intern Med 2005; 165: 490–6[Abstract/Free Full Text]

13 Zhao SZ, Reynolds MW, Lefkowith J. et al. A comparison of renal-related adverse drug reactions between rofecoxib and celebrex, based on the World Health Organization/Uppsala Monitoring Centre safety database. Clin Ther 2001; 23: 1478–91[CrossRef][ISI][Medline]

14 Welton A, Fort JG, Puma JA et al. Cyclooxygenase-2-specific inhibitors and cardiorenal function: a randomized, controlled trial of celecoxib and rofecoxib in older hypertensive osteoarthritis patients. Am J Ther 2001; 8: 85–95[CrossRef][Medline]

15 Ott E, Nussmeier NA, Duke PC et al. Efficacy and safety of the cyclooxygenase 2 inhibitors parecoxib and valdecoxib in patients undergoing coronary artery bypass surgery. J Thorac Cardiovasc Surg 2003; 125: 1481–92[Abstract/Free Full Text]

16 US Food and Drug Administration. BEXTRA valdecoxib tablets. Available online at: http://www.fda.gov/cder/foi/label/2004/21341lbl.pdf

17 Analysis and recommendations for Agency action regarding NSAIDs and cardiovascular risk. Available online at: http://www.fda.gov/cder/drug/infopage/COX2/NSAIDdecisionMemo.pdf

18 Hippisley-Cox J, Coupland C Risk of myocardial infarction in patients taking cyclo-oxygenase-2 inhibitors or conventional non-steroidal anti-inflammatory drugs: population based nested case-control analysis. Br Med J 2005; 330: 1366–72[Abstract/Free Full Text]

19 Updated advice on the safety of selective COX-2 inhibitors. Available online at: http://www.mca.gov.uk/aboutagency/regframework/csm/csmhome.htm

20 Expanded information on COX-2 inhibitors for doctors and pharmacists. Available online at: http://www.tga.gov.au/media/2005/050214_cox2.pdf

21 Medsafe media release. Stringent conditions for COX-2 inhibitors. Available online at: http://medsafe.govt.nz/hot.htm

22 Landesberg G. The pathophysiology of perioperative myocardial infarction: facts and perspectives. J Cardiothorac Vasc Anesth 2003; 17: 90–100[CrossRef][ISI][Medline]

23 Beattie WS, Warriner CB, Etches R et al. The addition of ketorolac to a patient-controlled analgetic morphine regime reduced postoperative myocardial ischemia in patients undergoing elective total hip or knee arthroplasty. Anesth Analg 1997; 84: 715–22[Abstract]

24 Kimmel SE, Berlin JA, Kinman JL et al. Parenteral ketorolac and risk of myocardial infarction. Pharmacoepidemiol Drug Saf 2002; 11: 113–19[CrossRef][ISI][Medline]





This Article
Full Text (PDF)
E-Letters: Submit a response to the article
Alert me when this article is cited
Alert me when E-letters are posted
Alert me if a correction is posted
Services
Email this article to a friend
Similar articles in this journal
Similar articles in ISI Web of Science
Similar articles in PubMed
Alert me to new issues of the journal
Add to My Personal Archive
Download to citation manager
Disclaimer
Request Permissions
Google Scholar
Articles by Jones, S. F.
Articles by Power, I.
PubMed
PubMed Citation
Articles by Jones, S. F.
Articles by Power, I.