Correspondence University Hospital, Rämistrasse 100, CH-8091, Zürich, Switzerland

Email frank.ruschitzka@usz.ch

Original article

Nussmeier NA et al. (2005) Complications of the COX-2 inhibitors parecoxib and valdecoxib after cardiac surgery. N Engl J Med 352: 1081–1091   PubMed

Synopsis

Background

Despite the reported beneficial effects after noncardiac surgical procedures, the safety of selective cyclo-oxygenase 2 (COX2) inhibitors has been questioned. Studies have shown that their use after cardiac surgery and in nonsurgical situations can increase the frequency of adverse events.

Objective

To establish whether patients can be safely treated with the COX2 inhibitors valdecoxib and its intravenous prodrug, parecoxib, after coronary artery bypass grafting (CABG) surgery.

Design

This international, multicenter, randomized, placebo-controlled study enrolled patients aged 18–80 years who had undergone elective CABG surgery with cardiopulmonary bypass. Patients had to be in NYHA class I–III or have an ejection fraction of 0.35 or more. Exclusion criteria included myocardial infarction in the 7 days before study start, a thromboembolic event in the 3 months before start, duodenal or gastic ulcer in the 60 days before start or CABG surgery lasting more than 3.5 h.

Intervention

Patients were randomly assigned to one of three treatment groups: 40 mg intravenous parecoxib on day 1 after surgery and 3 days of 20 mg parecoxib every 12 h, followed by 20 mg valdecoxib every 12 h until day 10; 3 days of placebo followed by 20 mg valdecoxib every 12 h until day 10; or placebo for 10 days. After surgery and throughout the study, every patient received 75–325 mg aspirin daily, and opioid medication if necessary. Use of nonsteroidal anti-inflammatory drugs (NSAIDs), prophylactic antiemetic agents, sedating antihistamines, epidural or intrathecal opioids and localized analgesics was not permitted.

Outcome measures

The primary endpoint was a combination of renal events, gastrointestinal complications, surgical wound complications and cardiovascular events (including peripheral vascular, cerebrovascular and cardiac events).

Results

Of the 1,671 patients enrolled, 555 were randomized to the parecoxib and valdecoxib treatment strategy, 556 to placebo and valdecoxib and 560 to placebo only. There were significantly more adverse events in patients who had received COX2-inhibitor therapy than in patients who had received placebo; 40 (7.4%) patients suffered adverse events in each COX2 group (parecoxib and valdecoxib, and placebo and valdecoxib) versus 22 (4.0%) patients in the placebo group (risk ratio 1.9; 95% CI 1.1–3.2; P = 0.02 for each comparison with placebo group). Specifically, patients treated with parecoxib and valdecoxib had significantly more cardiovascular events than patients treated with placebo (11 [2.0%] patients versus 3 [0.5%] patients, risk ratio 3.7, 95% CI 1.0–13.5, P = 0.03). There was no significant difference in cardiovascular events, however, when results from both COX2 inhibitor groups were pooled and compared with the placebo group. Although higher in the COX2 inhibitor groups than placebo group, the difference in the frequency of renal events, gastrointestinal complications and surgical wound complications was not significant.

Conclusion

The study confirmed the findings of a previous, smaller study and the authors conclude that patients should not receive COX2 inhibitors after CABG surgery.

Commentary

Although NSAIDs offer advantages in the management of postoperative pain, their use is limited because of adverse effects. By leaving cyclo-oxygenase 1 (COX1)-mediated platelet aggregation and gastric protection unopposed, drugs selectively blocking the COX2 isoenzyme promised similar pain control and fewer gastrointestinal side effects than nonselective COX1 and COX2 inhibitors.

The results of this trial, however, add to concerns that the promise of gastrointestinal safety with COX2 inhibitors comes with increased cardiovascular toxic effects. An earlier trial of CABG-surgery patients receiving higher doses of parecoxib and valdecoxib also suggested a trend towards increased cardiovascular events.¹ Interestingly, a similar trial of patients undergoing general surgery² revealed no cardiovascular risks with parecoxib or valdecoxib, which suggests that atherosclerotic vascular disease, exposure to additional cardiopulmonary bypass risks and COX2-mediated effects on ischemic preconditioning could predispose patients to thrombotic events.

Concerns have been present since the VIGOR trial suggested rofecoxib might increase myocardial infarction risk.³ The APPROVe⁴ and APC⁵ trials with rofecoxib and celecoxib, respectively, also showed increased cardiovascular complications; however, they were inadequately powered and not designed to examine cardiovascular outcomes.

The CLASS study and other major cohort analyses suggest that celecoxib could be less hazardous than other COX2 inhibitors. This theory is supported by the PreSAP and ADAPT trials, although the CIs for these trials are high enough not to exclude a hazard of a similar size to that seen by Nussmeier et al., or in the the APC or APPROVE trials.

Controversy about the mechanisms of potential cardiovascular side effects of COX inhibitors and NSAIDs persists. Intriguingly, the sulfone COX2 inhibitors rofecoxib and etoricoxib might increase blood pressure to a greater extent than celecoxib and NSAIDs. Blood-pressure and cardiovascular-safety data on valdecoxib are, however, limited. While some of the cardiovascular effects of COX2-selective inhibitors might be caused by a prostacyclin and thromboxane imbalance, multiple opposing cardiovascular influences (particularly on inflammation, oxidation and the c-Jun N-terminal kinase pathway) might be at work. If the prostacyclin–thromboxane imbalance theory holds, the addition of aspirin should eliminate risk; however, patients in the present study received aspirin and the addition of this COX1 inhibitor did not prevent cardiovascular side effects. Moreover, if the prostacyclin–thromboxane hypothesis represented the only explanation, nonselective NSAIDs should be associated with fewer cardiovascular side effects. The increased cardiovascular risk with naproxen compared with placebo and celecoxib seen in the ADAPT trial, combined with observational studies of several million patients, highlight the need to scrutinize these agents as rigorously as COX2-selective inhibitors. Hence the FDA moved to place warnings concerning cardiovascular side effects on all COX inhibitors. In addition, they have been contraindicated in post-CABG surgery pain management.

Despite the FDA suspension of valdecoxib sales and marketing, several regulatory authorities have decided that parecoxib still shows a positive risk-to-benefit ratio. Importantly, the lack of a safe intravenous NSAID will impact the decision to keep parecoxib available. To address the uncertainty, our institution is planning a multicenter trial of celecoxib with a head-to-head design, involving more than 20,000 patients with osteoarthritis to examine cardiovascular outcomes against traditional NSAIDs. Many patients are receiving these agents instead of COX2-selective inhibitors on the misleading presumption of improved cardiovascular safety. Until we have these results, analysis of cardiovascular risk versus gastrointestinal benefit needs to be performed for each agent.

Acknowledgments

The synopsis was written by Hannah Camm, Associate Editor, Nature Clinical Practice.

References

1. Ott E et al. (2003) Efficacy and safety of the cyclooxygenase 2 inhibitors parecoxib and valdecoxib in patients undergoing coronary artery bypass surgery. J Thorac Cardiovasc Surg 125: 1481–1492 | Article | PubMed | ISI | ChemPort |
2. Whelton A (online 7 March 2005) Safety of parecoxib and valdecoxib in the treatment of postoperative pain following coronary artery bypass graft surgery or major general surgery [http://www.acc05online.acc.org/highlights/keyLectures.aspx?sessionId=8030&date=7] (accessed 15 August 20005)
3. Mukherjee D et al. (2001) Risk of cardiovascular events associated with selective COX-2 inhibitors. JAMA 286: 954–959 | Article | PubMed | ISI | ChemPort |
4. Bresalier RS et al. (2005) Cardiovascular events associated with rofecoxib in a colorectal adenoma chemoprevention trial. N Engl J Med 352: 1092–1102 | Article | PubMed | ISI | ChemPort |
5. Solomon SD et al. (2005) Cardiovascular risk associated with Celecoxib in a clinical trial for colorectal adenoma prevention. N Engl J Med 352: 1071–1080 | Article | PubMed | ISI | ChemPort |

Contact the journal about this article

Subject areas under which this article appears: Intervention | Therapy