Download Appropriate use of NSAIDs: Considering cardiovascular risk in the

Appropriate use of NSAIDs: Considering
cardiovascular risk in the elderly
Publish date: Mar 1, 2007
By: Martin J. Stillman, MD, JD, FCLM, M. Thomas Stillman, MD, FACP
Geriatrics. 2007 Mar;62(3):16-21.
Abstract
In the first of this two-part article, we reviewed essential gastrointestinal (GI) data necessary
for choosing selective COX-2 inhibitors (coxibs) versus nonselective nonsteroidal antiinflammatory drugs (NSAIDs), as well as other NSAID-related GI issues. Although GI
considerations are critical to appropriate NSAID selection, the worldwide withdrawal of
rofecoxib because of adverse cardiovascular (CV) events has changed the focus of
appropriate NSAID selection. In part 2, we discuss relevant CV adverse effects related to
NSAID use. Based upon data reviewed, we believe there are differences between coxibs and
that all NSAIDs, including nonselective agents, have some degree of CV risk. Their use should
be based upon patients' risks and benefits. Our clinical use pathway or algorithm will continue
to frame the ongoing discussion and guide clinicians along what has become a difficult
decision in daily practice.
Stillman MJ, Stillman MT. Appropriate use of NSAIDs. Considering cardiovascular risk in the
elderly. GERIATRICS 2007; 62(Mar):16-21.
Key words: nonsteroidal anti-inflammatory drugs (NSAIDs) • COX-2 inhibitors (coxibs) •
hypertension • cardiovascular toxicity adverse effects • aspirin • clinical use pathway
Drugs discussed: rofecoxib • lumiracoxib • ibuprofen • celecoxib aspirin • naproxen •
meloxicam • diclofenac • acetaminophen valdecoxib
Nonsteroidal anti-inflammatory drugs (NSAIDs), both cyclooxygenase-2 (COX-2) selective and
nonselective, play a valuable role in treating certain types of pain and inflammation. Deciding
when to use more costly pharmacologic therapies is often difficult, and making the appropriate
choice to use selective COX-2 inhibitors (coxibs) is no exception. Since all types of NSAIDs
are essentially equally efficacious to each other, selection of one over another generally
depends on cost, side effect profiles, and ease of patient compliance.
As we discussed in part 1 of this two-part article,* we believe that data support the improved
safety of coxibs over nonselective NSAIDs regarding fewer GI adverse events. However, this
improved safety has more clinical meaning and impact in the patient at higher risk for GI
complications. The degree of GI benefit may be reduced in certain clinical situations, possibly
in patients taking low-dose aspirin for cardioprotection.
Figure 1 NSAID clinical use pathway for managing pain
The clinical use pathway presented in part 1, as well as in this paper (figure), helps guide
physicians in deciding who should receive an NSAID and considers factors affecting the
choice between specific agents in the NSAID class. In part 2, we navigate decision making in
light of the cardiovascular (CV) effects of NSAIDs, and review study findings while comparing
risk across the NSAID group.
Does the evidence support an association of coxibs with CV events?
This question has moved to the forefront of discussions involving coxibs since rofecoxib was
globally withdrawn from the market because of its increased association with CV events. Since
the short answer to this question is "yes," physicians should consider whether this CV side
effect is unique to rofecoxib or if it is a class effect shared by all coxibs. Our belief is that there
are differences among the coxibs, and that not all coxibs have the same CV risk.
As support for this opinion, it is important to note that several differences in the side effect
profiles of coxibs are well established. Specifically, rofecoxib has been shown to have a higher
incidence of edema and destabilization of treated hypertension than celecoxib. In patients with
concomitant osteoarthritis (OA) and stable treated hypertension, rofecoxib, 25 mg/d, was
associated with a significant increase in systolic blood pressure and peripheral edema,
compared with celecoxib, 200 mg/d.1 Similar results were noted in patients with OA and type II
diabetes mellitus who had stable hypertension after treatment with an angiotensin-converting
enzyme (ACE) inhibitor.2
A retrospective study examined the incidence of new-onset hypertension following the initiation
of coxib therapy in 17,844 patients aged 65 or older.3 Rofecoxib users had an overall
increased relative risk of 1.6 compared with patients taking celecoxib. This relative risk of newonset hypertension increased to 2.1 in patients with a history of chronic renal disease, liver
disease, or congestive heart failure (CHF).
In a 1-year cardiovascular outcome trial, patients taking lumiracoxib had a significantly smaller
mean change from baseline of systolic blood pressure than those taking ibuprofen or
naproxen.4
The association between NSAIDs and CHF has also been examined. Although prospective
randomized controlled studies are lacking, retrospective analyses have been completed. In a
study of elderly patients, it was estimated that non-selective NSAIDs were responsible for
close to 20% of first hospital admissions with CHF.5 In a large retrospective cohort analysis,
general rates of admission for CHF were examined. In this study, patients on rofecoxib and
nonselective NSAIDs had an increased risk of admission for CHF compared with non-NSAID
users. This difference was not seen with users of celecoxib.6
These differences in edema, hypertension, and CHF outcomes suggest that side effect profiles
may vary among coxibs. Theoretically, endothelial COX-2 produces a prostacyclin-induced
vasodilatory effect, as well as an inhibitory effect on platelet thrombosis.7 It has been proposed
that the use of a coxib therefore causes unopposed COX-1-directed platelet formation of
thromboxane and thus a prothrombotic effect in certain at-risk patients. This effect was
considered on reviewing cardiovascular data seen in the Vioxx Gastrointestinal Outcome
Research (VIGOR) trial.8 In this trial, patients taking rofecoxib, 50 mg/d, had a 5 times greater
incidence of nonfatal myocardial infarction (MI) than patients taking naproxen, 500 mg/bid.
Three possible explanations have been proposed for this finding.7 First, since the VIGOR trial
was powered to look at GI outcome and not CV outcome, this study of 8,000 patients was
underpowered to evaluate the incidence of CV events, and the increased CV events in the
rofecoxib arm may have occurred entirely on the basis of chance.
Second, because naproxen blocks platelet formation of thromboxane via COX-1 inhibition,
patients who were taking naproxen had a cardioprotective effect (similar to aspirin) not present
in the rofecoxib group. Based on subsequent retrospective observational studies, mixed
results have been reported on the cardioprotective effect of naproxen.9,10 Although there may
have been some naproxen-related protection in the VIGOR data, it is unlikely to be the entire
explanation, since naproxen would have then been associated with a 75% reduction in nonfatal MI in the VIGOR trial. This finding would far exceed that of any primary prevention trial
using the established choice of aspirin, which has been shown to reduce the risk of a first MI
by about one-third.11
The third possibility is that there was an increase in nonfatal MI in the VIGOR trial because of
an increased CV risk associated with rofecoxib. Since the VIGOR trial, the concern about
coxibs and CV risk has prompted many investigations. These have primarily been casecontrolled retro-spective analyses looking for such associations. Although studies have
supported that rofecoxib and celecoxib have differences in their effects on CV events,12 there
is growing evidence that patients taking any NSAID, selective or nonselective, have a higher
incidence of CV events compared with nonusers.
In an FDA-contracted, retrospective case-controlled study in collaboration with the Kaiser
Permanente integrated managed care organization, investigators looked for an NSAIDassociated risk of acute MI or sudden cardiac death (SCD) in 1.4 million patients taking
selective and nonselective NSAIDs over a 3-year period.13 Analysis was carried out on a total
of 8,143 patients who had serious coronary heart disease that resulted in an MI. Of these
patients, 2,210 died of cardiac causes. The matched control group was made up of 31,496
patients who had not been taking any NSAID for at least 60 days before the cardiac event
index date.
Study results revealed that rofecoxib use at a dose greater than 25 mg/d increased the risk of
acute MI and SCD by 3.15. Rofecoxib use at all doses increased the risk compared with
celecoxib. Of additional interest was the finding of a statistically significant increase in cardiac
events in patients taking naproxen or other nonselective NSAIDs, including meloxicam. This
finding provided additional evidence for a lack of a cardioprotective effect with nonselective
NSAIDs (eg, naproxen) and, in fact, supported the concept that CV risk may be associated
with all NSAIDS.
An additional retrospective study14 examined a cohort of 651,000 adults with physiciandiagnosed arthritis and involved nearly 2.4 million patient-years of follow-up. In this
retrospective analysis, 15,343 patients with acute MI had been taking selective or nonselective
NSAIDs. Although the group taking celecoxib, 200 mg/d, had a lower incidence of MI than the
group taking rofecoxib, 50 mg/d, almost all NSAIDs, including celecoxib, were associated with
some degree of CV risk for acute MI. Additionally, increased dosing of the NSAIDs was
associated with higher increased MI risk.
Although these studies focused on comparing the CV effects among coxibs and between
coxibs and nonselective NSAIDs, other studies have more directly examined the CV effects of
coxibs compared with placebo. This comparison emerged through the review of studies
involving coxibs and prevention of colorectal polyps. Specifically, the studies were designed to
examine the effect of coxibs versus placebo in preventing the recurrence of adenomatous
colorectal polyps.
In the prospective, randomized placebo-controlled Adenomatous Polyp Prevention on Vioxx
(APPROVe) trial, 2,600 patients were to be observed for 3 years to evaluate the efficacy of
rofecoxib, 25 mg/d, in preventing recurrence of colorectal polyps in patients with a history of
colorectal adenomas.15 This trial was stopped early because of a twofold increase in relative
risk for confirmed CV events (eg, MI, stroke), for patients taking rofecoxib compared with
placebo.
Notably in this study, the rofecoxib group had a statistically significant higher percentage of
hypertension and edema than the placebo group. This hypertension and edema difference
began to occur at approximately 5 months into the study. When the study was terminated early
because of these CV events, Merck and Co. announced a voluntary worldwide withdrawal of
rofecoxib.
A similar prospective, randomized placebo-controlled clinical trial involving approximately
2,100 patients was designed to evaluate the efficacy of celecoxib, 200 mg/bid and 400 mg/bid,
in preventing recurrence of colorectal polyps in patients with a history of colorectal
adenomas.16The Data Safety Monitoring Board discontinued this study after 33 months
because of a statistically significant dose-related increase in CV risk seen in the patients taking
celecoxib compared with placebo. Specifically, the incidence of CV events was 2.3% and 3.4%
for the 200 mg/bid and 400 mg/bid doses, respectively, compared with 1% for the placebo
population.
In a nearly identical polyp prevention study comparing celecoxib, 400 mg/d, to placebo, there
was no statistically significant difference in CV events between the two treatment groups.17
Further reports of these studies suggest that the drug-related CV risk may be of more concern
in patients with preexisting CV risk. CV adverse events were four to five times higher in both
the placebo and celecoxib arms if the patient had a prior history of CV events.18,19 Additionally,
Gislason et al reported that coxibs in all dosages and nonselective NSAIDs in high dosages
increase mortality in patients with previous MI.20
Finally, an extensive meta-analysis reviewed 23 observational studies regarding NSAIDs and
CV risk.21 Analysis of the 17 case-controlled portion included more than 86,000 cases of CV
events. This robust review further substantiated the increased risk of CV events associated
with rofecoxib and also concluded that celecoxib, 200 mg/d, was not associated with an
increased CV risk. As noted in other studies, the authors felt that data did not exclude an
increased risk with celecoxib at higher dosages. The meta-analysis reviewed nonselective
NSAIDs as well. Most had an associated CV risk similar to celecoxib, although the study found
that diclofenac was associated with increased CV risk. While this study seems to substantiate
the current relationship of CV events and NSAIDs, especially with increased dosing, to best
assess the CV concern regarding all NSAIDs, additional prospective, double-blind, placebocontrolled studies are needed.
Is there an effect of NSAIDs, nonselective or selective, on aspirin's cardioprotective
effect?
Many patients who take NSAIDs for pain control also take low-dose aspirin for
cardioprotection. Concern has been raised regarding the possibility that the NSAID use limits
the cardioprotective effect of aspirin.
In a pharmacologic prospective trial measuring thromboxane B2 formation,22 investigators
concluded that ibuprofen administered 2 hours before aspirin may interfere with the
cardioprotective effect of aspirin. This was explained by ibuprofen's initial competitive inhibition
at the blocking site of serum thromboxane B2 formation and consequent interference of
aspirin-induced platelet inhibition. This inhibition was not seen when aspirin was given 2 hours
before ibuprofen. The same effect was not demonstrated by simultaneous use of aspirin and
rofecoxib, diclofenac, or acetaminophen.
In an attempt to support or refute this antagonistic relationship, clinical studies have reviewed
the incidence of CV events in patients taking aspirin and non-aspirin NSAIDs. Results have
been mixed. In a study of more than 7,000 patients discharged with a low-dose aspirin
regimen following a CV event,23 those taking ibuprofen concurrently had an increased risk of
CV mortality of 1.93. Another study24 suggested that regular, but not intermittent, use of
NSAIDs inhibited the clinical benefits of aspirin. Finally, in a large, nested, case-controlled
analysis,25several different nonselective NSAIDs appeared to have no effect on the risk of
acute CV events, irrespective of aspirin use. At this point, the question of whether there may
be a mitigating effect on CV protection in patients taking con-current aspirin and NSAIDs
remains unanswered.
What were the recommendations of the FDA review?
In February 2005, all available CV data were reviewed at a joint meeting of the Arthritis
Advisory Committee and the Drug Safety and Risk Management Advisory Committee
convened by the FDA.26 This combined committee set out to discuss the overall risk-benefit
considerations involving CV and GI safety concerns for coxibs and related agents. They
reaffirmed that coxibs are important treatment options for arthritis patients in the United States,
and voted that the overall risk-benefit profile of celecoxib, valdecoxib, and rofecoxib supported
their being marketed in the United States.
Somewhat contrary to these recommendations, in April 2005, the FDA announced a series of
changes to the entire class of NSAIDs.27 This series included a boxed warning for the potential
risk of CV events and GI bleeding associated with all prescription NSAIDs, including celecoxib.
In light of this increased CV risk, as well as the increased reporting rate of rare skin reactions
with valdecoxib, the FDA requested that valdecoxib be withdrawn from the market.28 For all
prescription NSAIDs, the FDA requested a contraindication for use in patients who have
recently undergone coronary artery bypass surgery. Additionally, manufacturers of over-thecounter NSAIDs have been asked to include potential CV, GI, and skin reaction risks on their
labels.
Summary
Recently, greater attention has been given to the CV side effect profile that led to the
withdrawal of rofecoxib and, to some degree, the suspension of valdecoxib. We believe that
data describing the differing incidence of edema, aggravated hypertension, and CHF, as well
as variation in the incidence of other CV events, support that there are differences among the
coxibs and that CV events are not entirely related to a class effect of the coxibs. Rather,
retrospective studies have shown a varying degree of CV risk, especially with increased
dosing, with both coxibs and nonselective NSAIDs. Therefore, all NSAIDs should be used in
the lowest effective dosage and with caution in high-risk CV patients, particularly in our elderly
patients. Further well-designed prospective studies are needed to better clarify both the
characteristics that define "high risk" and the degree of CV differences among available
NSAIDs.
While one school of thought has been to avoid the use of coxibs altogether, we believe this is
clinically inappropriate. Many elderly patients need treatment of pain, and NSAIDs provide a
significant degree of relief, especially when inflammation is present. As our first article
discussed, GI risk can be a significant barrier to using nonselective NSAIDs, and the coxibs'
general risk reduction of GI events by 50% compared with nonselective NSAIDs is clinically
meaningful. This is especially so in the elderly, since age alone represents a GI risk factor in
NSAID use. While a link between NSAIDs and CV events appears to exist, the risk of NSAID
use cannot outweigh the benefit in every patient. For many elderly patients, improving quality
of life by relieving pain at least warrants an informed consent discussion with their physician
regarding NSAID use. The clinical use pathway reviewed in this paper can assist with this
informed discussion. In addition, it lends itself to accomplishing certain goals in clinical
practice: delivering medical treatment that accounts for appropriate evidence-based decisions
and medical costs, while still providing thorough, thoughtful, and effective patient care.
Dr. Martin Stillman is an internist and attorney. He is a faculty member in the department of
medicine, Hennepin County Medical Center, and assistant professor, University of Minnesota
Medical School.
Dr. M. Thomas Stillman is a rheumatologist. He is director, undergraduate medical education,
former director, division of rheumatology, department of medicine, Hennepin County Medical
Center, and clinical professor, University of Minnesota Medical School.
Disclosures: Dr. Martin Stillman is a study design consultant and medical lecturer for Pfizer,
Inc., and medical consultant for Novartis Pharmaceuticals Corp. Dr. M. Thomas Stillman is a
medical consultant and lecturer for Pfizer, Inc., medical consultant for Merck & Co., and
medical consultant for Novartis Pharmaceuticals Corp. No grant, research support,
consulting/advising fees, or any financial remuneration was received for preparation of this
manuscript.
Acknowledgments
The authors would like to thank Colleen Sauber for her assistance with final editing, Shannon
Benson and Angie Knauss for their administrative assistance, and David Howzer for his
ongoing support.
References
1. Whelton A, White WB, Bello AE, Puma JA, Fort JG; SUCCESS-VII Investigators. Effects of
celecoxib and rofecoxib on blood pressure and edema in patients > or = 65 years of age with
systemic hypertension and osteoarthritis. Am J Cardiol 2002; 90(9):959-63.
2. Sowers JR, White WB, Pitt B, et al. The effects of cyclooxygenase-2 inhibitors and
nonsteroidal anti-inflammatory therapy on 24-hour blood pressure in patients with
hypertension, osteoarthritis, and type 2 diabetes mellitus. Arch Intern Med 2005; 165(2):61-8.
3. Solomon DH, Schneeweiss S, Levin R, Avorn J. Relationship between COX-2 specific
inhibitors and hypertension. Hypertension 2004; 44(1):1-5.
4. Schnitzer TJ, Burmester GR, Mysler E, et al. Comparison of lumiracoxib with naproxen and
ibuprofen in the Therapeutic Arthritis Research and Gastrointestinal Event Trial (TARGET),
reduction in ulcer complications: randomised controlled trial. Lancet 2004;364(9435):665-74.
5. Page J, Henry D. Consumption of NSAIDs and the development of congestive heart failure
in elderly patients. An underrecognized public health problem. Arch Intern Med 2000; 160:
777-84.
6. Mamdani M, et al. Cyclo-oxygenase-2 inhibitors versus non-selective non-steroidal antiinflammatory drugs and congestive heart failure outcomes in elderly patients: a populationbased cohort study. Lancet 2004; 363:1751-56.
7. FitzGerald GA, Patrono C. The coxibs, selective inhibitors of cyclooxygenase-2. N Engl J
Med 2001; 345(6):433-42.
8. Bombardier C, Laine L, Reicin A, et al. Comparison of upper gastrointestinal toxicity of
rofecoxib and naproxen in patients with rheumatoid arthritis. VIGOR Study Group. N Engl J
Med 2000; 343(21):1520-8, 2 p following 1528.
9. Ray WA, Stein CM, Hall K, Daugherty JR, Griffin MR. Non-steroidal antiinflammatory drugs
and risk of serious coronary heart disease: an observational cohort study. Lancet 2002;
359(9301):118-23.
10. Solomon DH, Glynn RJ, Levin R, Avorn J. Nonsteroidal anti-inflammatory drug use and
acute myocardial infarction. Arch Intern Med 2002; 162(10):1099-104.
11. Eidelman RS, Hebert PR, Weisman SM, Hennekens CH. An update on aspirin in the
primary prevention of cardiovascular disease. Arch Intern Med 2003; 163(17):2006-10.
12. Ray WA, Stein CM, Daugherty JR, Hall K, Arbogast PG, Griffin MR. COX-2 selective nonsteroidal anti-inflammatory drugs and risk of serious coronary heart disease. Lancet 2002;
360(9339):1071-3.
13. Graham DJ, Campen D, Hui R, et al. Risk of acute myocardial infarction and sudden
cardiac death in patients treated with cyclo-oxygenase 2 selective and nonselective nonsteroidal anti-inflammatory drugs: nested case-control study. Lancet 2005; 365 (9458):475-81.
14. Singh G, Mithal A, Triadafilopoulos G. Both selective COX-2 inhibitor and nonselective
NSAIDs increase the risk of acute myocardial infarction in patients with arthritis: selectivity is
with the patient, not the drug class. Ann Rheum Dis. 2005; 64(suppl III):85.
15. Bresalier RS, Sandler RS, Quan H, et al. Cardiovascular events associated with rofecoxib
in a colorectal adenoma chemoprevention trial. N Engl J Med 2005; 352(11):1092-1102.
16. Solomon SD, McMurray JJ, Pfeffer MA, et al. Cardiovascular risk associated with celecoxib
in a clinical trial for colorectal adenoma prevention. N Engl J Med 2005; 352(11):1071-80.
17. Food and Drug Administration. Advisory committee briefing document. January 12, 2005.
Available at: http://www.fda.gov/ohrms/dockets/ac/05/briefing/2005-4090±_03_pfizer-celebrex-bextra.pdf.
Accessed February 21, 2007.
18. Bertagnoli MM, Eagle CJ, Zauber AG, et al. Celecoxib for the prevention of sporadic
colorectal adenomas. N Engl J Med 2006; 355(9):873-84.
19. Arber N, Eagle CJ, Spicak J, et al. Celecoxib for the prevention of colorectal adenomatous
polyps. N Engl J Med 2006; 355(9):885-95.
20. Gislason GH, Jacobsen S, Rasmussen JN, et al. Risk of death or reinfarction associated
with the use of selective cyclooxygenase-2 inhibitors and nonselective nonsteroidal antiinflammatory drugs after acute myocardial infarction. Circulation 2006; 113(25):2906-13.
21. McGettigan P, Henry, D. Cardiovascular risk and inhibition of cyclooxygenase. A
systematic review of the observational studies of selective and nonselective inhibitors of
cyclooxygenase 2. JAMA 2006; 296:1633-44.
22. Catella-Lawson F, Reilly MP, Kapoor SC, et al. Cyclooxygenase inhibitors and the
antiplatelet effects of aspirin. N Engl J Med 2001; 345(25):1809-17.
23. MacDonald TM, Wei L. Effect of ibuprofen on cardioprotective effect of aspirin. Lancet
2003;361(9357):573-4.
24. Kurth T, Glynn RJ, Walker AM, Chan KA, Buring JE, Hennekens CH, Gaziano JM.
Inhibition of clinical benefits of aspirin on first myocardial infarction by nonsteroidal
antiinflammatory drugs. Circulation 2003; 108(10): 1191-5.
25. Garcia Rodriguez LA, Varas C, Patrono C. Differential effects of aspirin and non- aspirin
nonsteroidal antiinflammatory drugs in the primary prevention of myocardial infarction in
postmenopausal women. Epidemiology 2000; 11(4):382-7.
26. Food and Drug Administration. Minutes of the joint meeting of the Arthritis Advisory
Committee and the Drug Safety and Risk Management Advisory Committee, February 16-18,
2005. Available at: http://www.fda.gov/ohrms/dockets/ac/05/minutes/2005-4090m1_Final.htm. Accessed
February 21, 2007.
27. Food and Drug Administration. COX-2 selective (includes Bextra, Celebrex, and Vioxx)
and nonselective non-steroidal anti-inflammatory drugs (NSAIDs). April 7, 2005; updated July
18, 2005. Available at: http://www.fda.gov/cder/drug/infopage/cox2/default.htm. Accessed February
21, 2007.
28. Food and Drug Administration. Alert for healthcare professionals: vadecoxib (marketed as
Bextra). April 7, 2005. Available at: http://www.fda.gov/cder/drug/infosheets/hcp/valdecoxibhcp.htm .
Accessed February 21, 2007.