Date: 2006-06-14

Help on accessing alternative formats, such as Portable Document Format (PDF), Microsoft Word and PowerPoint (PPT) files, can be obtained in the alternate format help section.

---

Contact
MHPD

Report on the Cardiovascular Risks Associated with COX-2-Selective Non-Steroidal Anti-Inflammatory Drugs

Health Products and Food Branch
Marketed Health Products Directorate and Therapeutic Products Directorate

Summary of regulatory recommendations

Health Canada has completed an extensive review of the data available as of December 31, 2005, on the cardiovascular (including cerebrovascular) side-effects of COX-2-selective non-steroidal anti-inflammatory drugs (NSAIDs), including consultations at the COX-2 Public Forum and Expert Advisory Panel meeting of June 2005. The drugs reviewed were celecoxib (Celebrex), meloxicam (inlcuding Mobicox and generic forms of meloxicam), rofecoxib (Vioxx) and valdecoxib (Bextra)¹. These medications were approved for use in the symptomatic treatment of arthritis pain.

This report describes the available evidence and the analyses of Health Canada physicians and scientists, which are the basis of the following regulatory positions and recommendations:

• Clinical studies indicate that the long-term use of the COX-2-selective drugs from the NSAID class is associated with an increased risk of adverse cardiovascular events, such as heart attack and stroke, when compared to placebo (no drug).
  
  
• The limited available evidence suggests that cardiovascular safety concerns associated with COX-2-selective NSAIDs also apply to most traditional drugs within the NSAID class², and that the degree of risk differs among NSAIDs. A lack of comparative randomized clinical trial data prevents the ranking of these drugs with respect to cardiovascular risk.
  
  
• The Basic Product Monograph Information for all NSAIDs should be revised to describe the available data or lack of data on cardiovascular toxicity, the increased cardiovascular risk, and to recommend caution accordingly, particularly for patients with cardiovascular disease or with risk factors for cardiovascular disease. A contraindication to use in patients undergoing certain high-risk cardiovascular surgeries and a reminder to use NSAIDs at the lowest effective dose for the shortest time possible period should be added to labelling. As data on cardiovascular risk/safety becomes available for a particular NSAID drug, the labelling will be revised accordingly.
  
  
• Valdecoxib sales were voluntarily suspended by the market authorization holder in April 2005, and Health Canada has formalized this off-market status, as noted in an advisory issued on December 9, 2005. This decision was based on the risk associated with long-term exposure to most NSAIDs, the serious cardiovascular side effects reported in patients taking valdecoxib up to 2 weeks following coronary artery bypass surgery, and a higher reported rate of rare but serious and unpredictable skin reactions for valdecoxib than for other COX-2 selective NSAIDs. A New Drug Submission (NDS) would have to be submitted by the manufacturer and found to be satisfactory upon review by Health Canada before valdecoxib could be authorized for future sales in Canada.
  
  
• The available data supports the continued availability of celecoxib and meloxicam on the Canadian market, with labelling revised to reflect new cardiovascular warnings and precautions as for most NSAIDs. Additional labelling revisions describe the evidence of cardiovascular toxicity noted in a recent 3-year randomized clinical trial for celecoxib, and the lack of long-term trials for meloxicam. Celecoxib labelling revisions were completed in the fall of 2005, as noted in advisories issued on September 21, 2005.
  
  
• Rofecoxib will remain off the Canadian market, further to its voluntary worldwide withdrawal by the manufacturer in September 2004, unless the manufacturer chooses to submit a new NDS. Such a submission would be reviewed by Health Canada, in light of the new cardiovascular safety data for all COX-2 selective NSAIDs, of the specific findings from the APPROVe trial for rofecoxib and any additional information that may be available at that time to evaluate whether or not Vioxx could be re-authorized for marketing in Canada.

1. "Bextra" (valdecoxib) is the registered trademark of Pharmacia & Upjohn Company LLC, Pfizer Canada Inc., Licensee. "Vioxx" (rofecoxib) is the registered trademark of Merck & Co., Inc. Used under license. "Celebrex" (celecoxib) is the registered trademark of G.D.Searle & Co., Pfizer Canada Inc., Licensee. Meloxicam is sold under various generic names and as "Mobicox", the registered trademark of Boehringer Ingelheim KG.

2. The older ("traditional", "conventional" or "non-selective") NSAIDs include naproxen, diclofenac, ibuprofen, indomethacin, aspirin and other less commonly used drugs. The cardiovascular safety concerns associated with the traditional NSAIDs are not extended to aspirin.

Background

On September 30, 2004, new information arising from a 3-year randomized controlled clinical trial prompted Merck & Co., Inc. to announce the worldwide withdrawal of rofecoxib. An increased relative risk was reported for cardiovascular events, including heart attacks and strokes, in patients taking rofecoxib compared to those taking placebo. In October 2004, scientific, medical and epidemiological experts at Health Canada began an extensive and thorough review of the data available for related drugs (celecoxib, valdecoxib, and meloxicam) to determine whether the cardiovascular safety concerns observed with rofecoxib were applicable to other COX-2-selective NSAIDs. External clinicians, epidemiologists, patient representatives and the public were also consulted at the COX-2 Public Forum and Expert Advisory Panel meeting of June 2005.

The available evidence, the conclusions of Health Canada experts regarding the interpretation of the available evidence, and the recommendations for regulatory actions that prevent or manage the risks to patients using COX-2-selective NSAIDs are highlighted below.

Evidence of cardiovascular risk

This review considered the evidence available from pre-clinical studies, randomized clinical trials, non-randomized observational studies, spontaneously reported adverse event data, the published literature on the mechanisms of action of COX-2-selective NSAIDs, and input from external experts consulted during the COX-2 Expert Advisory Panel meeting of June 2005. This information was mainly available in the published scientific literature, presentations and submissions from Market Authorization Holders (manufacturers) to Health Canada, the Canadian Adverse Drug Reaction Monitoring Program and in the expert presentations to, and discussions of the COX-2 Expert Advisory Panel.

Preliminary information from two long-term placebo-controlled trials (the APPROVe trial studying rofecoxib, and the APC trial studying celecoxib) indicated that the risk of cardio-thromboembolic adverse events increased with increasing drug dose and exposure period. In the APPROVe trial, a difference in the increased relative risk of cardiovascular adverse events first became evident after 18 months of treatment, and became significant after 3 years of exposure. Based on information from these and other studies, the reviews of evidence for each drug focused on studies with an extended daily exposure to the medication, e.g. a clinical trial treatment duration of 12 weeks or longer.

Pre-clinical studies (all drugs)

The review of pre-clinical studies included longer-term studies of up to 2 years duration, and the results were examined for any evidence of abnormalities of the cardiovascular system. The pre-clinical studies revealed no treatment-related adverse findings in the cardiovascular system for any of celecoxib, valdecoxib and meloxicam. These findings were consistent with the pre-clinical data available for rofecoxib, which did not predict the increased cardiovascular risk observed in the APPROVe trial.

Randomized clinical trials (RCTs)

Randomized clinical trials of rofecoxib (Vioxx)

The analysis of the cardiovascular risk of COX-2-selective NSAIDs included rofecoxib as a basis for comparison with the other COX-2-selective NSAIDs under review.

Patients taking rofecoxib in sixteen short-term randomized trials (up to 6 months duration) did not experience a significantly higher incidence of serious cardiovascular events compared to those exposed to active comparators (comparison drugs) or placebo (no drug). In four placebo-controlled studies of longer duration (1 to 4 years), the incidence of serious cardiovascular events was higher than incidences reported in the shorter trials. Patients taking 25 mg daily rofecoxib in two Alzheimers studies of 12 months duration did not have a higher incidence of heart attack than those taking placebo (1720 patients total, averaging 75 years of age), though in one of these studies rofecoxib use was associated with an increased incidence of stroke compared to placebo. Among 1457 patients enrolled in an Alzheimers study for up to 4 years, those taking 25 mg daily rofecoxib experienced a higher incidence of cardiac events (3.6%) than those in the placebo group (2.6%) (including heart attack, unstable angina pectoris and cardiac arrest and death). Two trials reported a significantly higher incidence of myocardial infarction and cardiac failure in the rofecoxib group compared to placebo (APPROVe trial) or naproxen (VIGOR trial).

The APPROVe trial was a three year RCT involving 2586 patients with a history of colorectal adenomas. The trial was designed to test the effect of rofecoxib on colorectal polyp recurrence, but included an assessment of cardiovascular problems as a prespecified cardiovascular endpoint. A significantly increased risk of thrombotic (blood-clot-related) cardiovascular events was reported for rofecoxib at a dose of 25 mg daily, compared with placebo (relative risk of 1.92). This increased relative risk became apparent after 18 months, and became significant after 3 years of daily treatment. Cardiac arrest (fatal and non-fatal) and cerebrovascular accidents were also more common among those patients taking rofecoxib compared to those on placebo. The Market Authorization Holder decided to withdraw rofecoxib from the worldwide market based on the data in this clinical trial.

In the VIGOR trial (8076 patients; 9-months long), rheumatoid arthritis patients received a 50 mg daily dose of rofecoxib (this high daily dose was never approved in Canada for more than 5 days of consecutive use). Rofecoxib patients experienced a significantly higher incidence of serious cardiovascular events (2.5%) than those taking 500 mg twice daily naproxen (1.1%), a traditional NSAID; this was due in part to a significant 5-fold higher incidence of heart attacks among the rofecoxib users. This difference was attributed by the authors to a potential cardio-protective effect of naproxen, though debate over the evidence of such an effect of naproxen continues in the scientific literature. The primary endpoint of this study was confirmed clinical upper gastrointestinal events. Cardiovascular events were not pre-specified as endpoints but were assessed by an independent committee. In another 1-year RCT that compared rofecoxib with naproxen, cardiovascular adverse events were also more common among the rofecoxib group.

Randomized clinical trials of celecoxib (Celebrex)

A total of 20 randomized clinical trials (RCTs) were reviewed that examined the use of celecoxib for 12 weeks or more. Among the 12- to 26- week trials, an increased incidence of cardiovascular events was seen with increasing doses of celecoxib, though no significant difference was observed in the incidence of serious cardiovascular events among patients who received celecoxib and those who received active comparators or placebo. In the three 12-month studies, the incidence of serious cardiovascular events was higher than that reported in the 12-week trials.

A dose-related effect was reported in the long-term randomized controlled APC trial (2035 patients). After treatment for an average of 33 months, the increased risk (hazard ratio) for heart attack, stroke or death from cardiovascular causes was 2.5 among patients taking 200 mg twice daily celecoxib, and 3.4 among patients taking 400 mg twice daily celecoxib, compared to placebo. This cancer prevention trial was not designed or statistically powered to evaluate cardiovascular risk, so some events may have gone unreported.

Preliminary results from two other long-term (3-year) randomized controlled prevention trials did not show an increased risk of cardiovascular events (heart attack, stroke or death) in patients receiving 200 mg twice daily celecoxib compared to naproxen (ADAPT Alzheimers trial), or 400 mg once daily celecoxib versus placebo (PreSAP colorectal cancer prevention trial). Conclusions regarding these data are reserved pending the availability of final data.

Randomized clinical trials of valdecoxib (Bextra)

The review of RCTs involving treatment with valdecoxib for 12 weeks or longer showed no significant differences in the incidence of heart attack, stroke, or cardiac failure among patients who received valdecoxib compared with patients who received another NSAID or placebo. It is important to note that none of these trials was more than 1 year in duration, so it remains unknown if there is an increased risk of cardiovascular events with longer-term use of valdecoxib.

Two short-term studies (10-14 days) were conducted in a total of 2,098 patients who were undergoing coronary artery bypass graft ("CABG") surgery and therefore were already susceptible to further cardiovascular problems. In one CABG study, an increased risk of heart attack and cerebrovascular accident was reported in patients who received 40 mg valdecoxib twice daily compared with patients who received standard pain medication and/or placebo. In the second CABG study, patients receiving 20 mg valdecoxib twice daily had a higher frequency of cardiovascular events (2.0%) than those given placebo (0.5%). These studies suggest that the short-term use of valdecoxib may predispose susceptible patients to cardiovascular events.

Randomized clinical trials of meloxicam (Mobicox and generic meloxicam)

A total of 10 randomized clinical trials were reviewed that examined the use of meloxicam for 12 weeks or more. The incidence of cardiovascular events was rare and there were no significant differences between meloxicam-treated groups and active comparator- or placebo-treated groups. None of these trials were longer than 1 year in duration so there is a lack of data regarding possible long-term cardiovascular risks associated with meloxicam.

Non-randomized (observational) studies (all drugs)

The majority of randomized clinical trials were designed to evaluate efficacy in the treatment of chronic pain. While the cardiovascular safety of these drugs was also assessed in these studies, they were generally not powered (designed) to evaluate cardiovascular events. Furthermore, the long-term effects of these drugs may not become readily apparent until well after trial completion. In addition, the setting of and the patients enrolled in the RCTs may differ from the typical setting/patient environment of a clinical practice. Results from well-designed, non-randomized (observational) studies can provide results that are more applicable to "real-world" experience of clinical practice.

A systematic review of published non-randomized studies was completed, identifying high-quality observational studies (8 case-control; 8 cohort studies). The results reported in these studies suggest that there are differences in the cardiovascular risks associated with exposure to celecoxib vs. rofecoxib. Only one of 14 studies on celecoxib reported an increased risk of cerebrovascular events in patients prescribed this medication, while eight of the 14 studies reported an increased risk associated with exposure to either all doses or high dose (>25 mg daily) rofecoxib (heart attack, death from any cause, cerebrovascular events, congestive heart failure and new cases of hypertension). Six studies showed no increased risk associated with rofecoxib use. The two studies evaluating meloxicam reported no increased risk of heart attack among patients prescribed this drug, and a lower risk of thromboembolic (blood-clot related) events compared to rofecoxib and celecoxib.

It is difficult to determine what factors contributed to the differences in results among these observational studies. Most of the studies (13 of 16) used information from electronic databases to determine outcome and exposure measures. Only five of these studies reported that the quality of the information in their databases had been validated. The fact that the studies reporting an increased cardiovascular risk associated with rofecoxib use came from validated databases supports a potential greater cardiovascular risk for rofecoxib than for celecoxib.

While most of the non-randomized studies evaluated in this systematic review may have good external validity (i.e. are high-quality studies), there are still problems inherent in non-randomized studies. Non-randomized treatment allocation may result in potential bias and thus differences in the cardiovascular risks between groups cannot necessarily be attributed solely to differences in drug treatment. For instance, celecoxib patients in an osteoarthritis study were more likely to be older and on high doses of the drug compared to rofecoxib patients. In another study, rofecoxib patients were more likely to be women and to have rheumatoid arthritis compared to the patients on other NSAIDs. As the scientific literature reports an association between rheumatoid arthritis and cardiovascular disease, it becomes difficult to determine if the increased risk of a heart attack among patients prescribed a high dose of rofecoxib (>25 mg daily) is due to the drug treatment or to one or more confounding factors.

Spontaneous Adverse Event Reports (all drugs)

Post-market evidence was reviewed to compare the number of adverse cardiovascular events spontaneously reported to Canadian and international event monitoring systems for celecoxib, valdecoxib, meloxicam and rofecoxib. Like observational studies, and in contrast with controlled randomized clinical trials, such data are considered a reflection of the "real world" experience.

During any given period since the first marketing of each drug, international exposure (worldwide distribution) was generally slightly higher for celecoxib and rofecoxib than for valdecoxib, which was in turn higher than for meloxicam. The international and Canadian data generally describe a similar rate of reported heart attacks for celecoxib and valdecoxib, and a slightly higher reporting rate for rofecoxib. International evidence suggests that cerebrovascular accidents (e.g. stroke) were also generally reported more frequently for rofecoxib throughout the marketing of this drug.

There were limited spontaneous Canadian reports of pre-specified adverse cardiovascular event for a valid comparison among drugs. For example, there were no spontaneous reports of any serious thrombotic adverse events reported for meloxicam in Canada (heart attack, stroke, death from cardiac problems). Under-reporting of adverse events is common, and events such as heart attacks that occur relatively frequently among the normal population are more likely to go un-reported as they may be assumed to be unrelated to the drug treatment.

Interpretation and Recommendations

Rofecoxib (Vioxx)

The analysis of the cardiovascular risk of COX-2-selective NSAIDs included rofecoxib as a basis for comparison with the other COX-2-selective NSAIDs under review. Strong evidence of cardiovascular toxicity comes from the 3-year APPROVe trial finding of a relative risk of 2 for serious cardiovascular events for rofecoxib compared to placebo.

Consistent with the cardiovascular findings of the APPROVe trial, a higher incidence of serious cardiovascular events was found among rofecoxib users in a number of smaller, shorter placebo-controlled trials, and in the 9 month VIGOR trial comparing high dose rofecoxib to naproxen, a traditional NSAID. A systematic review of non-randomized (observational) studies also suggests that rofecoxib has increased risks of cardiovascular events compared with celecoxib, and the available post-market adverse event data is consistent with such a difference.

The opinion of Expert Advisory Panel members was that the available information justifies the potential future marketing of rofecoxib in Canada (12 in favour, 1 against). The panel noted that most NSAIDs appear to carry cardiovascular risk, that rofecoxib (and celecoxib and valdecoxib) are associated with a decreased frequency of both gastrointestinal intolerance and clinically important peptic ulcer disease compared with traditional 'non-selective' NSAIDs, and that patients benefit from having a variety of drugs to choose from.

Rofecoxib was voluntarily withdrawn from the worldwide market in September 2004 by the Market Authorization Holder (manufacturer). Subsequent regulatory action ensured that rofecoxib remained off the Canadian market. Unless the manufacturer chooses to submit a New Drug Submission (a request for authorization to market the drug and supporting scientific information), and this submission is deemed satisfactory upon review by Health Canada, rofecoxib will not be authorized for use in Canada.

Celecoxib (Celebrex)

Strong evidence of cardiovascular toxicity associated with celecoxib comes from the APC study, where a hazard ratio of 2.5 indicates that patients have a higher risk of cardiovascular problems if they are taking 200 mg of celebrex twice daily than if they are taking a placebo. The preliminary results of two other randomized trials of similar duration (the PreSAP and ADAPT trials) did not find a significantly increased risk.

Trends noted in shorter clinical studies (12-52 weeks duration) include an increased incidence of cardiovascular adverse events with increasing celecoxib dose and, in three 52 week studies, a higher incidence of cardiovascular events compared to 12 week studies. Although the clinical trial data is not consistent, there is evidence available to date associating the long-term use of celecoxib with an increased risk of serious cardiovascular adverse events.

Only one observational study found an increased cardiovascular risk associated with celecoxib use, while a number of studies indicated that rofecoxib is associated with a higher risk of heart attack than celecoxib. This result is echoed in the descriptive findings of the post-market spontaneous reporting rate data.

Members of the Expert Advisory Panel consulted regarding COX-2-selective NSAIDs unanimously recommended that celecoxib should remain on the market in Canada. The debate focused on the fact that the increased risk of cardiovascular events associated with celecoxib appears to be shared by most NSAIDs, the risk of gastrointestinal harm appears to be less than for most NSAIDs, that patients benefit from having a variety of drugs to choose from for pain relief, and the option to individually assess and accept potential risks.

Based on the available evidence, Health Canada has recommended that celecoxib remain on the Canadian market with revised labeling, given that:

1. The cardiovascular risks associated with celecoxib appear to be lower than those associated with rofecoxib.
   
   
2. There is no evidence that the cardiovascular risks associated with celecoxib are greater than the risks associated with traditional NSAIDs, which are currently the most likely alternatives to celecoxib use.
   
   
3. Revised labelling can reduce the risk of serious side effects by informing health care professionals and patients of the possible risks of serious cardiovascular events.
   
   
4. The removal of celecoxib from the Canadian market would further limit the choices of Canadians with respect to selective COX-2 NSAIDs given that rofecoxib and valdecoxib are presently not marketed in Canada.

The suggested labelling changes for celecoxib are consistent with those recommended for the NSAID Guidance document (see below), and include a description of the available evidence regarding the cardiovascular safety and toxicity of celecoxib.

Pending the availability of clearer, more consistent data, this regulatory approach will minimize the exposure of susceptible patients to this drug, and inform the prescribing and medication decisions of healthcare professionals and patients allowing their consideration of the benefit-risk balance on a case-by-case basis.

This recommendation was implemented, as noted in the Health Canada-endorsed Public Advisory and Letter to Health Professionals, issued by the manufacturer on September 21, 2005. These advisories are available on the Health Canada website at the following addresses (accessed January 1, 2006):

http://www.hc-sc.gc.ca/dhp-mps/medeff/advisories-avis/public/2005/celebrex_pa-ap-eng.php
http://www.hc-sc.gc.ca/dhp-mps/medeff/advisories-avis/prof/2005/celebrex_3_hpc-cps-eng.php

Valdecoxib (Bextra)

Unlike rofecoxib and celecoxib, there are no studies that assess daily valdecoxib treatment for more than 1 year. A cardiovascular safety signal was seen following the acute use of valdecoxib in high risk populations (two CABG studies). This indicated the existence of underlying mechanisms of acute cardiovascular toxicity that could contribute to a potential increased risk of serious cardiovascular events when used long-term in patients with cardiovascular disease or risk factors for cardiovascular disease. Pending the availability of long-term controlled clinical trial data demonstrating cardiovascular safety, the available evidence for valdecoxib is consistent with an NSAID class effect.

In a separate review of post-market data, spontaneous reports of adverse drug reactions were analyzed for severe cutaneous adverse reactions (SCARs) such as Stevens-Johnson syndrome, toxic epidermal necrolysis and erythema multiforme. Such skin reactions are typically drug-induced, however they are rare events and are not typically detected in controlled clinical trials due to the relatively small size of clinical trials. Though these adverse skin reactions occur only rarely, they are reported at a higher rate for patients taking valdecoxib than for those treated with other COX-2-selective NSAIDs. Currently there are no data available that allow the clinician to predict which patients might be susceptible.

Eight of 13 members of the Expert Advisory Panel on selective COX-2 NSAIDs considered the benefit-risk profile of valdecoxib unfavourable and voted against future market access to the drug. The discussion highlighted the evidence of acute cardiovascular toxicity in high-risk patients, the lack of long-term controlled clinical trials and the increased risk of SCAR with valdecoxib (vs other COX-2-selective drugs).

Further to review of this evidence, Health Canada considers the benefit-risk profile unfavourable and has recommended that valdecoxib not be permitted back onto the Canadian market, given:

1. evidence of increased cardiovascular risk with long term use of other Cox-2 selective inhibitors (i.e. rofecoxib in the APPROVe trial, and celecoxib in the APC trial);
   
   
2. the lack of long-term studies examining the cardiovascular safety profile of valdecoxib, and the cardiovascular safety signal seen with the acute use of valdecoxib in high risk populations (CABG studies);
   
   
3. the risk of severe and potentially fatal skin reactions that occur rarely but more often with valdecoxib than with other Cox-2 selective inhibitors;
   
   
4. the absence of evidence identifying populations at risk of the severe skin reactions; and
   
   
5. the availability of other therapeutic alternatives, including anti-inflammatory medications.

This recommendation was implemented, as noted in the advisory issued by Health Canada on December 16, 2005. This advisory is available on the Health Canada website at the following address: http://www.hc-sc.gc.ca/ahc-asc/media/advisories-avis/_2005/2005_134-eng.php .

Should the manufacturer seek to resume the marketing of valdecoxib in Canada, a New Drug Submission (NDS) with additional safety data and/or revised labelling will be required, which would have to be found satisfactory upon review by Health Canada.

Meloxicam (Mobicox and generic meloxicam)

The limited available evidence indicates that the occurrence of cardiovascular adverse events are rare for meloxicam and no significant differences have been documented at different dosages, or between meloxicam and comparator or placebo groups. However, there are no studies that assess daily meloxicam treatment for more than 1 year.

Two non-randomized studies reported no increased risk of myocardial infarction among meloxicam patients and a lower risk of cerebrovascular events compared to rofecoxib and celecoxib users. Post-market spontaneous adverse cardiovascular events are too few to be informative. These low event numbers may be due, at least in part, to a lower use of meloxicam in comparison to celecoxib and rofecoxib.

The Expert Advisory Panel of June 2005 was asked to focus on the COX-2 selective NSAIDs for which more data was available. Consequently no guidance was provided specifically for meloxicam, though it is encompassed by panel discussions that addressed COX-2 selective inhibitors and NSAIDs in general.

Based on the available evidence, Health Canada has recommended that meloxicam remain on the Canadian market with revised labeling, given that:

1. the cardiovascular risks associated with meloxicam appear to be lower than those associated with rofecoxib;
   
   
2. the cardiovascular risks associated with meloxicam use do not appear to be greater than the risks associated with 'non-selective' NSAIDs, the most likely alternative to meloxicam use; and
   
   
3. the removal of meloxicam from the Canadian market would further limit the choices of Canadians with respect to anti-inflammatory medications.

The suggested labelling changes for meloxicam are consistent with those recommended for the NSAID Guidance document (see below), and include an indication of the lack of available evidence regarding the cardiovascular safety and toxicity of meloxicam.

Pending the availability of longer-term cardiovascular safety data, this regulatory approach will minimize the exposure of susceptible patients to meloxicam, and inform the prescribing and medication decisions of healthcare professionals and patients allowing their consideration of the benefit-risk balance on a case-by-case basis.

NSAIDs as a class

It is recognized scientifically that the term "selective" is a relative term. While the newer, COX-2-selective NSAIDs are so-named because they have been designed to act more selectively at the COX-2 enzyme (vs the COX-1 enzyme) and/or have been marketed with this emphasis, some traditional NSAIDs can also act on the COX-2 enzyme. Traditional (so-called 'non-selective') NSAIDs³ were included in this review of COX-2-selective NSAIDs to the extent that the RCTs and non-randomized studies under review provided data for both traditional and COX-2-selective NSAIDs. As there are few other studies available that provide data following long-term exposure to traditional NSAIDs, this evidence was considered in formulating recommendations for NSAIDs as a class.

The scientific literature on the mechanism(s) of action of COX-2-selective drugs proposes that the increased risk of adverse cardiovascular events is due, in part, to the selectivity of these drugs towards the COX-2 enzyme and is therefore a characteristic of all COX-2 selective NSAIDs. However, it is also clear from the literature that this property of the drug is unlikely to be the only determinant of its cardiovascular risk; each particular drug acting on the COX-2 enzyme might additionally possess distinct COX-2 enzyme-independent actions that could contribute to the likelihood of an adverse cardiovascular event. The mechanistic evidence also suggests that the dose and duration of exposure to a drug, and individual patient characteristics, will influence the cardiovascular effect of the drug. This information supports the concept of a common underlying mechanism of cardiovascular toxicity for all non-aspirin NSAIDs, including COX-2 selective drugs, but also describes alternative additional effects and factors that may come into play for each drug.

The view of the Expert Advisory Panel was that, as a group, selective Cox-2 inhibitors are associated with an increased risk of clinically important cardiovascular events compared with placebo, and that this increased risk is similar to the risk associated with most NSAIDs. This increased risk is present for all patients taking anti-inflammatory agents (with the possible exception of naproxen), but the absolute risk likely increases with longer-term use and in the presence of risk factors for, or a history of, cardiovascular disease.

The limited available evidence from clinical studies and the literature is similarly interpreted by Health Canada experts to indicate that cardiovascular safety concerns associated with COX-2-selective NSAIDs also encompass traditional ('non-selective') drugs within the NSAID class, with the exception of aspirin. While it is recognized that the degree of cardiovascular risk differs between NSAIDs, a lack of comparative randomized clinical trial data prevents the ranking of these drugs with respect to cardiovascular risk.

To address these concerns regarding class-wide cardiovascular toxicity, Health Canada's NSAID guidance document, entitled "Draft Guidance for Industry: Basic Product Monograph Information for Nonsteroidal Anti-Inflammatory Drugs (NSAIDs)", was updated to inform both prescribers and consumers of the cardiovascular risks now associated with most NSAIDs . A draft version of the NSAID Guidance Document was provided to the members of the Expert Advisory Panel of the COX-2 Public Forum for consultation. The draft dated August 25, 2005 was posted for external review by stakeholders, with the following proposed revisions regarding cardiovascular safety issues:

Part I - Healthcare Professional Information

• A contraindication to the use of the NSAID in patients in the peri-operative period of CABG surgery period (i.e directly before, during and after surgery),
  
  
• A warning concerning the use of the NSAID in patients with ischemic heart disease, cerebrovascular disease, or congestive heart failure, or in patients with risk factors for these conditions.
  
  
• A recommendation that NSAIDs should not be used in patients with an increased risk of developing such cardiovascular problems if other effective treatments are available.
  
  
• A reminder to use the drug as recommended at the lowest effective dose for the shortest possible duration of time.

Part III - Consumer Information

• Revisions in everyday language for patients, reflecting the revisions directed to healthcare professionals in Part I.

This NSAID guidance document will be finalized further to consideration of the comments from stakeholders.

3. The older ("traditional", "conventional" or "non-selective") NSAIDs include naproxen, diclofenac, ibuprofen, indomethacin, aspirin and other less commonly used drugs. The cardiovascular safety concerns associated with the traditional NSAIDs are not extended to aspirin. There are no long-term data comparing naproxen to placebo, and the limited available data on the cardiovascular safety of naproxen are inconsistent. Ibuprofen is the only non-aspirin NSAID that is available in Canada without a prescription (over-the-counter), but the doses recommended for non-prescription ibuprofen use are below those available by prescription. Acetaminophen is also an analgesic (pain-reliever), and is available without a prescription, but is not a member of the NSAID class of drugs and is not associated with the cardiovascular safety concerns described above.