On September 30, 2004, Merck & Co. (Whitehouse Station, New Jersey)
withdrew its blockbuster drug rofecoxib from the worldwide market. After just
five years on the market, the annual sales of rofecoxib (Vioxx) had grown to
$2.5 billion. Rofecoxib and celecoxib (Celebrex; Pfizer, New York, NY) belong
to a relatively new class of nonsteroidal anti-inflammatory drugs that were
designed to selectively inhibit the COX-2 (cyclooxygenase-2) enzyme and thus
reduce gastrointestinal side effects. While the coxibs have been proven to be
effective for reducing arthritis symptoms, their safety has been
controversial. This review updates orthopaedic surgeons on the present status
of the coxib controversy.
The COX-1 enzyme metabolizes arachidonic acid to prostaglandins. The COX-1
enzyme is constitutively expressed in most tissues, including the gastric
mucosa, whereas expression of COX-2 is thought to be induced in response to
inflammation. Thus, COX-2-selective inhibitors were designed to produce the
analgesic effects of nonsteroidal anti-inflammatory drugs while minimizing
gastrointestinal side
effects1.
Nonselective nonsteroidal anti-inflammatory drugs reduce
thromboxane-A2 production in platelets and may decrease platelet
aggregation, leading to bleeding. Selective coxibs are particularly attractive
in the perioperative period because they do not inhibit
thromboxane-A2 production and thus do not promote
bleeding2.
FitzGerald3
proposed that COX-2-selective inhibitors have a curious side effect: they
reduce production of prostaglandin I2 (prostacyclin). Prostaglandin
I2 is produced by COX-2 in the vascular endothelium, and it causes
vasodilatation and inhibits platelet aggregation. Selective inhibition of
COX-2 thus may reduce prostacyclin production, resulting in vasoconstriction
(which may cause hypertension) and platelet clumping (which could accelerate
formation of atherosclerotic
plaques)3.
The coxibs were approved by the United States Food and Drug Administration
(FDA) on the basis of two large prospective, randomized clinical studies. The
CLASS (Celecoxib Long-Term Arthritis Safety) study of 8059 patients documented
a significant reduction in the prevalence of gastrointestinal ulcers in those
taking celecoxib compared with the prevalences in those taking ibuprofen or
diclofenac (p <
0.02)4. The VIGOR
study, involving 8076 patients, showed a similar reduction in gastrointestinal
side effects when rofecoxib was compared with
naproxen5.
The orthopaedic community embraced the coxibs as a result of their
effectiveness and favorable side-effect profile. Coxibs have proved valuable
in the treatment of postoperative pain as well as pain associated with knee
and hip
osteoarthritis2.
Studies of animals, however, have raised concern that coxibs may delay
fracture-healing and
tendon-healing6,7.
The early clinical studies of the coxibs raised concern about their
cardiovascular safety. Although the VIGOR study was not designed to address
the issue of cardiovascular side effects, it did show an increased risk of
myocardial infarction in patients treated with rofecoxib (p < 0.01). A
meta-analysis of large clinical trials of celecoxib (including the CLASS
study4) did not show
any increased risk of cardiovascular
events8. Two
pharmacoepidemiologic analyses demonstrated that rofecoxib was associated with
an increase in the risk of myocardial
infarction9,10.
Merck's decision to withdraw rofecoxib was based on interim results from
the APPROVe study
(),
a randomized, prospective, placebo-controlled multicenter study designed to
investigate the role of rofecoxib in the progression of colorectal adenomas.
APPROVe was the first study performed to investigate the long-term use of
coxibs. There were 0.75 cardiovascular thrombotic events per patient-year in
the placebo group of that study and 1.5 events per patient-year in the
rofecoxib group. Thus, 25 mg of rofecoxib per day was associated with a
twofold increase in cardiovascular thrombotic events (95% confidence interval,
1.20 to 3.19; p < 0.007). The cardiovascular risk was not observed until
after eighteen months of treatment.
With rofecoxib removed from the market, orthopaedists are left with two
FDA-approved coxibs for the treatment of musculoskeletal pain: celecoxib
(Celebrex) and valdecoxib (Bextra; Pfizer). But is the increased cardiac risk
unique to rofe-coxib or is it a class effect of all coxibs? Valdecoxib
(Bextra) itself has not been shown in prospective studies to have cardiac side
effects11. However,
a study of the intravenous use of parecoxib, a pro-drug that turns into
valdecoxib in vivo, showed a significant increase in cardiac events and wound
infections (p <
0.015)12. It would
be reasonable to predict that valdecoxib has the same cardiac effects in vivo
as does its pro-drug, parecoxib. FitzGerald recently presented meta-analysis
data showing a slightly increased risk of thrombotic events in patients taking
valdecoxib13. The
FDA recently added a warning label to Bextra (valdecoxib) noting an increased
risk of Stevens-Johnson syndrome and an elevated thrombotic risk in patients
who had undergone a cardiac artery bypass grafting procedure
().
The available published data seem to indicate that cele-coxib does not
increase the risk of cardiac
problems6,8,14.
However, on December 17, 2004, Pfizer informed the FDA that the Adenoma
Prevention with Celecoxib (APC) trial had been halted. The APC trial was a
thirty-three month trial of a 400 or 800-mg daily dose of celecoxib for
prevention of colonic polyps. Analyses showed that patients taking these
supratherapeutic doses were 2.5 times more likely to have a cardiovascular
event. At the same time, results from a similar trial of a 400-mg daily dose
of celecoxib to prevent colorectal adenomas did not show an increased
cardiovascular risk over a placebo
().
Thus, the possibility of a class effect has not been eliminated. Pfizer has
proposed a full-scale trial specifically designed to address the cardiac
safety of celecoxib.
In the face of rapidly changing data on the coxibs, what course should the
prudent orthopaedic surgeon take? Acetaminophen and nonselective nonsteroidal
anti-inflammatory drugs should be the first-line treatments for
musculoskeletal pain, particularly when chronic therapy is anticipated.
Patients who are more than sixty-five years old, have a history of gastric
ulcers, or have a history of gastrointestinal bleeding could benefit from a
coxib. The surgeon should probe for a history of hypertension or
cardiovascular risk. Until definitive data on cardiovascular safety are
available, surgeons should avoid prescribing coxibs for patients with or at
high risk for coronary artery disease.