The cardiovascular risks of selective cyclooxygenase (COX)-2–blocking
nonsteroidal anti-inflammatory drugs (NSAIDs) are well recognized (JW
Gen Med Feb 25 2005). Current guidelines also discourage use of
nonselective NSAIDs in patients with known cardiovascular disease (CVD),
but some clinicians still prescribe these medications, because they
believe that short-term use is unlikely to raise risk significantly. To
address this issue, Danish investigators took advantage of that
country's highly integrated and digitized medical databases, which
accurately capture discharge diagnoses and pharmaceutical prescriptions
for the entire country.
In a prospective observational study that
involved 84,000 patients with first myocardial infarctions (MIs)
between 1997 and 2006, 40% of patients suffered recurrent MIs or death,
and 42% received NSAIDs at some time. NSAID use was associated with a
significant excess risk for recurrent MI or death (hazard ratio,
1.5–1.7, depending on timing and duration of NSAID use after MI). Not
surprisingly, risk was high for two COX-2 blockers, rofecoxib (HR,
1.4–2.3) and celecoxib (HR, 1.3–1.9). But, unexpectedly, risk was even
higher for the nonselective NSAID diclofenac (HR, 1.7–3.3). Ibuprofen
and other nonselective NSAIDs (except naproxen) also were associated
with excess risk. This risk became apparent immediately after use for
some NSAIDs and within several weeks for most (JW Cardiol Jun 8 2011).
In
a case-control study, more than 32,000 patients with initial diagnoses
of atrial fibrillation (AF) or flutter between 1999 and 2008 were
compared with nearly 326,000 age- and sex-matched controls without AF or
flutter. After adjustment for several confounders, risk for AF or
flutter was clearly higher with current use of COX-2 blockers (incidence
rate ratio, 1.27) and nonselective NSAIDs (IRR, 1.17). Those rates were
even higher for NSAID users who had started the medications within 60
days of AF or flutter onset (JW Gen Med Jul 26 2011).
In neither
of these studies were investigators able to capture fully and control
for all risk factors that could confound the analyses. Nevertheless, the
large size of the studies, the inclusion of essentially all cases in
the country (i.e., no selection bias), and the quality of the data
provide confidence that the results are accurate. The first study did
not include patients with known CVD who had not experienced MIs.
Nevertheless, the data seem to support current guidelines that
discourage NSAID use in patients with CVD. In particular, the data
indicate that even short-term use (e.g., for transient musculoskeletal
ailments) leads to excess cardiovascular risk.
[APPROVe Trial: Medline® abstract | Free full-text NEJM article PDF]
[APC trial: Medline® abstract | Free full-text NEJM article PDF]
[Nussmeier, et al study: Medline® abstract | Free full-text NEJM article PDF]
[Schjerning Olsen, et al study: Medline® abstract | Free full-text Circulation article PDF]
[Schmidt, et al study: Medline® abstract | Free full-text BMJ article PDF]
— Anthony L. Komaroff, MD
Published in Journal Watch General Medicine December 29, 2011
Copyright © 2011. Massachusetts Medical Society. All rights reserved.
No comments:
Post a Comment