After the withdrawal from the market of the COX-2 inhibitor rofecoxib (VIOXX) because of cardiac and circulatory damage, the question arises of whether the other COX-2 inhibitors also possess an increased cardiovascular risk and which anti-inflammatory drugs or analgesics should be recommended.

In Germany there are currently four COX-2 inhibitors available on the market: celecoxib (CELEBREX), valdecoxib (BEXTRA) and parecoxib (DYNASTAT) from Pfizer as well as the latest, etoricoxib (ARCOXIA; a-t 2004; 35: 103-4) from MSD. Parecoxib is a prodrug of valdecoxib and available only in parenteral form, which is intended for short-term treatment of postoperative pain. In view of the danger of acute renal failure, however, we regard this as a strict contraindication for both conventional non-steroidal anti-inflammatory drugs (NSAIDs) and for COX-2 inhibitors. Opioids are indicated in these situations (a-t 2002; 33: 62-3).

Celecoxib, etoricoxib and valdecoxib are mainly licensed for the treatment of osteoarthritis and rheumatoid arthritis. As a result of the mechanism of action, there is a strong suspicion that the cardiovascular events are evidence of a class effect (see a-t 1999; No. 12: 123-4). The 12-month EDGE* study involving more than 7,000 patients and now presented by MSD at a conference (MSD: "welcome results" (1)) is seemingly intended to dispel doubts about etoricoxib. In our opinion, however, the preliminary results (2) tell a different story: cardiovascular complications are more frequent with 90 mg etoricoxib daily than with 150 mg diclofenac daily (VOLTAREN, etc.): not statistically significantly (p = 0.115), but twice as frequent at 0.4% vs. 0.2%. As with rofecoxib in the VIGOR study*, significantly more patients taking etoricoxib drop out of the study because of hypertension (2.3% vs. 0.7%; see a-t 2001; 32: 87-8).

In response to the crash of rofecoxib, Pfizer started an extensive publicity campaign to convince professionals of the supposed "cardiac safety" of celecoxib and valdecoxib. This includes supplying doctors and pharmacists with an unsolicited promotional letter (3) detailing three studies. These are two case control studies on celecoxib and rofecoxib (4,5), only one of which has been fully published and can thus be assessed (4), and a meta-analysis partly financed by Pfizer themselves (6), which covers small-scale and in some cases unpublished Phase II and III trials with valdecoxib.

According to the published case control study based on US insurance data, an increased risk of myocardial infarction is associated with rofecoxib, but not celecoxib, in patients aged 65 and older (4). The crucial weakness of the study is that some of the most important risk factors or markers for myocardial infarction - smoker status, use of aspirin (acetylsalicylic acid; ASA, etc.), socio-economic status and body mass index - are not known and thus cannot be included in the risk assessment. Therefore, this study is not suitable as proof of the cardiovascular safety of celecoxib.

Actually, an adviser of the US medicines agency, the FDA, sees evidence of a prothrombotic effect of celecoxib in the largest randomised, controlled, long-term trial with celecoxib, the CLASS* study. The rate of infarctions is higher with the COX-2 inhibitor than with diclofenac (VOLTAREN, etc.) or ibuprofen (BRUFEN, etc.; 0.2% versus 0.1%) in patients not taking aspirin concurrently (7,8).

The EU medicines committee (CPMP) also states in its risk assessment report completed in 2004 that, according to clinical trial results, coxibs such as rofecoxib or etoricoxib may be associated with an increased risk of suffering a myocardial infarction and "it can be considered that there is a trend towards a higher MI risk associated with the use of celecoxib" (9,10). After rofecoxib has been withdrawn from the market, the European authority has now announced a re-assessment (11).

According to the meta-analysis presented in Pfizer's publicity material, no cardiovascular-damaging potential can apparently be found for the COX-2 inhibitor in comparison with standard NSAIDs or placebo in Phase II or III trials involving 10 mg to 20 mg valdecoxib daily in patients with joint disease or rheumatoid arthritis (6). Meta-analyses are susceptible to distortion. An analysis such as the one presented, which is financed by and originates in close cooperation with the manufacturer, is of questionable value. None of the studies included in the analysis lasts longer than six months. In the APPROVe* study, however, it took 18 months for the increased risk from rofecoxib to become significant (a-t 2004; 35: 116; see page 130: Sept. 2004). Previous meta-analyses of studies from the rofecoxib development program, which originated with involvement (12) or sole authorship (13) of the manufacturer MSD, also did not reveal any increase in cardiovascular safety risks as compared with other NSAIDs or placebo - apart from a difference from naproxen (PROXEN, etc.) These findings have since been disproved by the APPROVe trial.

The FDA review of valdecoxib reports an increased rate of oedema and elevated blood pressure with daily dosages above 20 mg (14). In two studies on analgesia in the context of bypass surgery, valdecoxib alone or in combination with parecoxib showed an increase in serious cardiovascular thromboembolic complications (15). Only one of the trials has been fully published (16). As in the VIGOR study (17), the various cardiovascular complications with the coxib are listed separately, so that no significant differences emerge. However, the combined rates for myocardial infarction, stroke and death are significantly and four times higher on valdecoxib/parecoxib than in the control group, according to the calculations of an FDA adviser for drug safety (15).

There is no confirmation for any clinically relevant advantage in terms of gastrointestinal toxicity for any of the available COX-2 inhibitors. According to the complete results from the CLASS study, celecoxib is no better tolerated by the gastrointestinal system than the conventional reference NSAIDs (a-t 2001; 32: 87-8). No randomised studies with clinically significant end-points have at all been published for the other available COX-2 inhibitors.

	As a result of their mechanism of action, there is a strong suspicion that the increase in cardiovascular complications with COX-2 inhibitors is evidence of a class effect.
	The cardiovascular safety of etoricoxib (ARCOXIA), celecoxib (CELEBREX), valdecoxib (BEXTRA) or parecoxib (DYNASTAT) cannot be validated by the results put forward by the manufacturers MSD and Pfizer.
	According evaluations of US and European regulatory authorities, there are signals in clinical trials that the cardiovascular tolerability of celecoxib, valdecoxib/parecoxib and etoricoxib is worse than that of conventional non-steroidal anti-inflammatory drugs (NSAIDs) or placebo.
	There is no adequate proof of better gastrointestinal tolerability for any of the marketed COX-2 inhibitors.
	In view of the serious safety concerns and lack of advantages, we advise against COX-2 inhibitors and recommend the use of established NSAIDs such as ibuprofen (BRUFEN, etc.) or diclofenac (VOLTAREN, etc.) given at the lowest possible dosage for as short a time as possible. Paracetamol (BENURON, etc.) should first be tried for pain caused by osteoarthritis. Paracetamol plus codeine (in TALVOSILEN FORTE, etc.) or other opioid analgesics may, however, be considered as an alternative to NSAIDs for patients with rheumatoid arthritis.

(R = randomised study, M = meta-analysis)

1

Süddeutsche Zeitung dated 15 Oct. 2004

2

Scrip 2004; No. 2998: 22

3

Pfizer GmbH: "Marktrücknahme von VIOXX und VIOXX DOLOR" [Market withdrawal of VIOXX and VIOXX DOLOR], circular letter dated 7 Oct. 2004

4

SOLOMON, D.H. et al.: Circulation 2004; 109: 2068-73

5

GRAHAM, D.J. et al.: www.oroalliance.com/CHARLES/ISPE2004.ppt

6

EDWARDS, J.E. et al.: Pain 2004; 111: 286-96

7

WOLFE, S.M.: www.citizen.org/pressroom/release.cfm?ID=1796

8

WITTER, I. (FDA): Medical Review(s) Celecoxib; 20 Sept. 2000
www.fda.gov/cder/foi/nda/2002/20-998s009_celebrex.htm

9

Pharm. Ztg. 2004; 149: 3211 and 3290

10

EMEA: "Overall summary of the scientific evaluation of medicinal products containing celecoxib, etoricoxib, parecoxib, rofecoxib and valdecoxib",
www.emea.eu.int/pdfs/human/referral/celecoxib/EN%20Celecoxib.pdf

11

EMEA: Press release, 6 Oct. 2004

12

KONSTAM, M.A. et al.: Circulation 2001; 104: 2280-8

13

REICIN, A.S.: Am. J. Cardiol. 2002; 89: 204-9

14

FARRELL, A. (FDA):
www.fda.gov/cder/foi/nda/2001/21-341_BEXTRA.htm

15

LENZER, J.: BMJ 2004; 329: 935

16

OTT, E. et al.: J. Thorac. Cardiovasc. Surg. 2003; 125: 1481-92

17

BOMBARDIER, C. et al.: N. Engl. J. Med. 2000; 343: 1520-8

CLASS = Celecoxib Long-term Arthritis Safety Study;
EDGE = Etoricoxib Diclofenac Gastrointestinal Evaluation;
VIGOR = VIOXX Gastrointestinal Outcomes Research
APPROVe = Adenomatous Polyp Prevention on VIOXX