Update: Safety Issues Related to NSAIDs and COX-2 Inhibitors
April 2002
NSAIDs remain among the most commonly used medications worldwide. Because of their more favorable GI toxicity profile compared to traditional NSAIDs in at-risk patients, the selective COX-2 inhibitors have become increasingly popular agents for the management of pain and inflammation. Several issues related to the safety of these drugs have recently been raised in published literature and in the press.
Cardiovascular
Complications
The potential association
of cardiovascular complications with the use of COX-2 inhibitors remains incompletely
defined (see ACR Hotline "Selective
COX-2 Inhibitors as Risk Factors for Cardiovascular Events,"
August 2001; and Strand V., Hochberg M. The risk of cardiovascular thrombotic
events with COX-2 selective inhibitors. Arthritis Rheum; Arthritis Care Res
2002 [in press]). Critical to the interpretation of data in this area is an
understanding of the potential harmful or protective effects of traditional
NSAIDs on cardiovascular risk. Previous studies have been inconclusive, with
studies suggesting either no effect or a protective effect Arthritis
Rheum 2001;44:S372;S230;S266; Epidemiology 2000;11:382). Ray
et al have recently reported a study of heart disease and NSAID use among 181,441
Tennessee Medicaid recipients between the age of 50-84 years (Lancet 2002;359:118)
conducted between 1987-1998 (before the introduction of COX-2 inhibitors). Use
of various non-aspirin NSAIDs (mean of 1.5 years) was not associated with either
an increased or reduced risk of serious coronary heart disease or of stroke,
as compared to controls. In another, ex vivo study, Lawson et al studied the
effect of NSAIDs on the ability of aspirin to inhibit platelet function (N Engl
J Med 2001;345:1809).
They hypothesize that ibuprofen, given before aspirin, may limit
aspirin's cardioprotective effect, as assessed by impact on serum thromboxane
B2 formation and platelet aggregation. This was not a class effect of NSAIDs,
as it was not seen with diclofenac; rofecoxib and acetaminophen also did not
have that effect. For review of the clinical data, e.g., from VIGOR and other
trials, see the references above.
Bottom Line: Despite challenges to conventional recommendations regarding the protective effects of aspirin (Lancet 2002;359:92), the weight of current evidence suggests that patients taking either traditional NSAIDs or selective COX-2 inhibitors who are at increased risk for cardiovascular disease should be offered anti-platelet agents where no contraindication to their use exists.
Congestive Heart Failure
(CHF)
It has long been appreciated that NSAIDs may be associated with exacerbations
of CHF, particularly among elderly persons with impaired renal function (Arch
Intern Med 1998;158:1108). The burden of illness associated with
CHF exacerbations may even approach that of GI adverse effects related to NSAID
use (Arch
Intern Med 2000;160:777). A recent study examined this association
in a group of 7,277 elderly Dutch patients (Arch
Intern Med 2002;162:265). In patients with CHF, the adjusted
relative risk of exacerbation associated with the use of NSAIDs was 9.9 (95%
CI 1.7-57.0). However, the relative risk of incident CHF was 1.1 (95% CI 0.7-1.7)
among current NSAID users. While these studies were of traditional NSAIDs, COX-2
inhibitors also have the potential to decrease glomerular filtration and alter
sodium and water excretion.
Bottom Line: NSAIDs and COX-2 inhibitors should be used with caution in
patients with CHF, as well as those with renal impairment, diabetes, and intravascular
volume depletion. Appropriate assessment, risk stratification, and monitoring
of patients should allow optimal use of these agents.
H. Pylori testing for
patients on NSAIDs?
The relationship between the two most important risk factors for peptic ulcer
disease, infection with Helicobacter pylori and use of NSAIDs, has been an area
of debate. It has been reported that treatment of H pylori both lowers the incidence
of NSAID-induced ulcers (Lancet 1997;350:975) or had no effect on ulcer development
and actually impaired ulcer healing (Lancet 1998;352:1016). Two recent reports
seem to favor an interaction between these two risk factors. Chan and colleagues
reported that screening for and treatment of H pylori reduced the risk of NSAID-induced
ulcers over a 6 month period from 34.4% to 12.1% among NSAID naïve patients
commencing long-term NSAID therapy (Lancet 2002;359:9). This difference was
highly significant, as was the difference in complicated ulcers (27% versus
4.2%).
Huang and colleagues performed
a meta-analysis of published literature to address this issue. They found a
synergistic relationship between the two risk factors. For patients with either
NSAID use or H pylori infection, the presence of the other risk factor increased
the risk of developing ulcer by approximately 3.5 fold. Compared with H pylori
negative NSAID non-users, those with both risk factors had a risk of ulcer of
61 (95% CI 9.98-373).
Bottom Line: Clinicians should be aware of these two important risk factors
for peptic ulcer disease. Concomitant with an understanding of the risks for
individual patients, based upon the presence of comorbidities and other factors,
it would seem prudent to consider appropriate screening and treatment of patients.
Aseptic Meningitis
Aseptic meningitis has long been recognized as a rare side effect that can occur
in association with the use of various traditional NSAIDs. A recent paper, based
upon information from the FDA Adverse Event Reporting System (AERS), reported
5 cases of aseptic meningitis associated with the use of rofecoxib (Arch
Intern Med 2002;162:713). In all cases, symptoms occurred within
1 and 12 days following exposure to rofecoxib. In one case, 2 rechallenges were
positive. Several other cases of aseptic meningitis in patients taking rofecoxib
and celecoxib have been received through AERS, but causality in those cases
cannot be clearly defined. As a result of these reports, the FDA asked Merck
to add aseptic meningitis to the list of potential adverse effects on the package
insert for rofecoxib.
Bottom Line: Rare cases of aseptic meningitis, which appear to be idiosyncratic
reactions rather than hypersensitivity reactions or sequelae of inhibition of
prostaglandins, have been reported with traditional NSAIDs and now with the
COX-2 selective agent rofecoxib. Health care providers should be aware of this
potential adverse effect in patients receiving these medications.
Aspirin-sensitive asthma
Aspirin-sensitive asthma is an important clinical concern, with an estimated
prevalence among asthmatic patients of 2-6% by history and 8-34% on oral challenge.
The association is stronger among asthmatic patients with nasal polyposis. The
cross-reactivity between aspirin and other NSAIDs suggested that alteration
in prostaglandin synthesis were relevant; this was supported by elevated concentrations
of urinary leukotrienes after aspirin challenge in affected patients. With the
introduction of the COX-2 selective inhibitors, it was hypothesized they may
be safer for patients with aspirin sensitivity, because they spare prostaglandins
synthesized via COX-1. However, patients with aspirin sensitivity were specifically
excluded from the clinical trials of these drugs. Since then, anecdotal observations,
open assessment, and small controlled trials seemed to support this contention.
Recently, several reports provide support for the lack of cross reactivity between
aspirin and COX-2 inhibitors (J
Allergy Clin Immunol 2001;108:47). Stevenson et al studied 60
patients with confirmed aspirin-sensitivity in a double-blind placebo controlled
study. On aspirin challenge, 32 of the 60 patients experienced bronchospasm
and upper airway symptoms, while 28 experienced only upper airway symptoms.
By contrast, none of the patients experienced nasal symptoms, changes in nasal
examination or significant change in FEV1 with rofecoxib (25 mg) challenge.
Bottom Line: While appropriate caution is indicated, COX-2 inhibitors
may be a therapeutic option for patients with aspirin-sensitive airway symptoms.
Hotline Editors: Eric L.Matteson,
MD, MPH; John J. Cush, MD; Arthur Kavanaugh, MD
April 5, 2002
Hotline is provided by
the ACR Communications and Marketing Committee as a service to members. This
Hotline reflects the views of the author(s) and does not represent a position
statement of the College.
