sign in sign up
<<
Home , Lumiracoxib

Number 4 Effective Health Care

Comparative Effectiveness and Safety of for Osteoarthritis Executive Summary

Background Effective Health Care Program Osteoarthritis is a chronic condition involving degeneration of cartilage within The Effective Health Care Program the joints. It is the most common form of was initiated in 2005 to provide valid arthritis and is associated with pain, evidence about the comparative substantial disability, and reduced quality effectiveness of different medical of life. About 6 percent of U.S. adults aged interventions. The object is to help 30 years or older have symptomatic consumers, health care providers, and osteoarthritis of the knee, and 3 percent others in making informed choices have symptomatic osteoarthritis of the hip. among treatment alternatives. Through Osteoarthritis increases with age: the its Comparative Effectiveness Reviews, incidence and prevalence increase two- to the program supports systematic tenfold from age 30 to 65 and continue to appraisals of existing scientific increase after age 65. The total costs for evidence regarding treatments for arthritis, including osteoarthritis, may be high-priority health conditions. It also greater than 2 percent of the gross promotes and generates new scientific domestic product, with more than half of evidence by identifying gaps in these costs related to work loss. existing scientific evidence and supporting new research. The program Common oral medications for puts special emphasis on translating osteoarthritis include nonsteroidal findings into a variety of useful antiinflammatory drugs (NSAIDs) and formats for different stakeholders, acetaminophen. Patients with osteoarthritis including consumers. also use over-the-counter supplements not regulated by the U.S. Food and Drug The full report and this summary are Administration (FDA) as pharmaceuticals, available at www.effectivehealthcare. including and chondroitin, as ahrq.gov/reports/final.cfm well as topical agents. medications are also used for selected patients with efficacy and safety may vary for individual refractory, chronic pain but are not drugs within a class. Nonpharmacologic recommended for first-line treatment of interventions (such as physical therapy, osteoarthritis and therefore not included in weight reduction, and exercise) also help this review. Each class of medication or improve pain and functional status in supplement is associated with a unique patients with osteoarthritis. balance of risks and benefits. In addition,

Effective Health Care

A challenge in treating osteoarthritis is deciding which Oral agents include: medications will provide the greatest symptom relief • Acetaminophen with the fewest serious adverse effects. NSAIDs decrease pain, inflammation, and fever by blocking • cyclo-oxygenase (COX) . Understanding of the • pharmacology of NSAIDs continues to evolve, but it is • Choline magnesium trisalicylate now thought that most NSAIDs block three different • Chondroitin COX isoenzymes, known as COX-1, COX-2, and COX-3. COX-1 protects the lining of the stomach from • acid. COX-2 is found in joint and muscle, and • mediates effects on pain and inflammation. By blocking • COX-2, NSAIDs reduce pain compared to placebo in 1 patients with arthritis, low back pain, minor injuries, • and soft tissue rheumatism. However, NSAIDs that also • block the COX-1 (also called “nonselective • NSAIDs”) can cause gastrointestinal bleeding. In the • Glucosamine United States, there are an estimated 16,500 annual deaths due to NSAID-induced gastrointestinal • complications, a higher death rate than that for cervical • Indomethacin cancer or malignant melanoma. Theoretically, NSAIDs • that block only the COX-2 enzyme (also called “coxibs,” “COX-2 selective NSAIDs,” or “selective • Ketoprofen ER NSAIDs”) should be safer with regard to • gastrointestinal bleeding, but they also appear to be • Lumiracoxib1 associated with increased rates of serious cardiovascular • Meclofenamate sodium and other adverse effects. Less is known about COX-3, which is found in the cerebral cortex and cardiac tissue • and appears to be involved in centrally mediated pain. • For this report, we defined the terms “selective • NSAIDs” or “COX-2 selective NSAIDs” as drugs in • the “coxib” class (celecoxib, , , • etoricoxib, lumiracoxib). We defined “partially selective NSAIDs” as other drugs shown to have partial in vitro • COX-2 selectivity (etodolac, nabumetone, meloxicam). • Rofecoxib1 Aspirin differs from other NSAIDs because it • irreversibly inhibits platelet aggregation, and the • derivatives (aspirin and salsalate) were considered a separate subgroup. We defined • Tenoxicam1 “nonaspirin, nonselective NSAIDs” or simply • Tiaprofenic acid1 “nonselective NSAIDs” as “all other NSAIDs.” • This report summarizes the available evidence • Valdecoxib1 comparing the benefits and harms of analgesics in the treatment of osteoarthritis.

1 These drugs are currently not approved by the FDA for use in the United States (etoricoxib, lumiracoxib, , ) or have been withdrawn from the market (rofecoxib and valdecoxib).

2

Questions addressed in this report are: nabumetone, etodolac) in efficacy for pain 1. What are the comparative benefits and harms of relief or improvement in function. treating osteoarthritis with oral medications or • In one short-term trial, salsalate and aspirin did supplements? How do these benefits and harms not differ significantly in efficacy for pain change with dosage and duration of treatment, and relief or symptom improvement. what is the evidence that alternative dosage • No studies evaluated the comparative efficacy strategies, such as intermittent dosing and drug of salsalate or aspirin vs. a nonaspirin NSAID. holidays, affect the benefits and harms of oral medication use? (Note: The only benefits considered • COX-2 selective NSAID vs. nonselective NSAID under this question are improvements in • COX-2 selective NSAIDs and nonselective osteoarthritis symptoms from long-term use. NSAIDs did not clearly differ in efficacy for Evidence of harms associated with NSAID use pain relief, based on many good-quality, include long-term studies of these drugs for treating published trials. osteoarthritis or rheumatoid arthritis and for cancer prevention.) • COX-2 selective NSAID vs. different COX-2 selective NSAID 2. Do the comparative benefits and harms of oral treatments for osteoarthritis vary for certain • Celecoxib and rofecoxib did not differ demographic and clinical subgroups of patients? significantly in efficacy for pain relief at commonly used and comparable doses, based • Demographic subgroups include age, sex, and on consistent evidence from six good-quality race. trials. • Coexisting diseases include hypertension, • No studies compared efficacy of COX-2s other edema, ischemic heart disease, heart failure; than celecoxib and rofecoxib. peptic ulcer disease; history of previous bleeding due to NSAIDs. Harms: gastrointestinal (GI) and cardiovascular (CV) • Concomitant medication use includes • Rofecoxib vs. nonselective NSAID anticoagulants. • In the only large, long-term trial (VIGOR), 3. What are the comparative effects of coprescribing rofecoxib 50 mg daily caused fewer serious of H2-antagonists, , or proton pump ulcer complications than naproxen 1,000 mg inhibitors (PPIs) on the gastrointestinal harms daily in patients with rheumatoid arthritis but associated with NSAID use? also significantly increased the risk of 4. What are the comparative benefits and harms of myocardial infarction. The overall rate of treating osteoarthritis with oral medications as serious adverse events was higher with compared with topical preparations? Topical rofecoxib than with naproxen. preparations include: , diclofenac, • There were about 16 fewer symptomatic ibuprofen, ketoprofen, and salicylate. ulcers, including 5.2 fewer serious GI A summary of the findings is shown in Table A. complications, for every 1,000 patients treated with rofecoxib vs. naproxen after a Conclusions median of 9 months of treatment. • There were 3.0 additional myocardial Oral NSAIDS infarctions for every 1,000 patients treated Benefits: improvements in osteoarthritis symptoms with rofecoxib compared to naproxen in • Nonselective NSAID vs. another nonselective VIGOR. NSAID • Rofecoxib was associated with an increased • Many trials found no clear differences between risk of myocardial infarction relative to placebo various nonaspirin, nonselective NSAIDs or in the most comprehensive systematic review partially selective NSAIDs (meloxicam, of randomized controlled trials (RCTs).

3

• About 3.5 additional myocardial diclofenac, ibuprofen, or naproxen in two fair- infarctions occurred for every 1,000 quality meta-analyses of published and patients treated for 1 year with rofecoxib unpublished trials. compared to placebo in the systematic • There have been too few events reported in review. RCTs of patients with chronic conditions to • Rofecoxib was withdrawn from the market in accurately assess CV risk associated with September 2004, primarily because of CV valdecoxib. risks. • Two short-term trials in a high-risk post- • Celecoxib vs. nonselective NSAID or placebo coronary-artery-surgery setting found that • It is not clear whether celecoxib has fewer valdecoxib was associated with a two- to potential harms than nonselective NSAIDs threefold higher risk of CV events compared when used longer than 3-6 months. In the only with placebo. large, published trial (CLASS), celecoxib at • Valdecoxib was withdrawn from the market 800 mg daily did not decrease predefined due to life-threatening skin reactions and serious ulcer complications overall compared increased CV risk. with diclofenac and ibuprofen; the risk of • Etoricoxib vs. nonselective NSAID serious GI events was lower than with ibuprofen, but not diclofenac, at 6 months in • Etoricoxib was associated with fewer GI patients who did not use aspirin; and there was adverse events (perforations, symptomatic no reduction in serious GI events at the end of ulcers, and bleeds) than nonselective NSAIDs followup. The overall rate of serious adverse in a fair-quality meta-analysis of 10 trials. events with celecoxib was similar to the rate • In primarily short-term trials, systematic with ibuprofen and diclofenac. reviews of RCTs suggest that etoricoxib has a • In fair-quality meta-analyses of arthritis trials, similar CV safety profile compared to other most of which evaluated short-term use, NSAIDs, with the possible exception of celecoxib caused fewer ulcer complications naproxen. Definitive conclusions are not than nonselective NSAIDs and did not increase possible because of small numbers of CV the risk of myocardial infarction. events. • Celecoxib 400 mg twice daily was associated • Lumiracoxib vs. nonselective NSAID with an increased risk of serious CV events • Results from one large trial (TARGET) found (CV death or myocardial infarction) relative to fewer adverse GI events with lumiracoxib than placebo in a long-term trial of polyp with naproxen and ibuprofen. prevention. • There was no statistically significant difference • Celecoxib was associated with an increased in rates of serious CV events between risk of myocardial infarction relative to placebo lumiracoxib relative to naproxen or ibuprofen in the most comprehensive systematic review in TARGET. of RCTs. Most of the CV events with celecoxib were reported in two large polyp-prevention • Too few events have been reported in RCTs to trials evaluating 200 mg or 400 mg twice daily, accurately assess CV risk associated with or 800 mg once daily. lumiracoxib. • About 3.5 additional myocardial • Partially selective NSAID vs. nonselective infarctions occurred for every 1,000 NSAID patients treated for 1 year with celecoxib • Meloxicam: There were no significant compared to placebo. differences in risks of serious GI events in • Valdecoxib vs. nonselective NSAID or placebo several meta-analyses of up to 28 primarily short-term clinical trials, and no difference in • Valdecoxib was associated with a lower risk of CV risk in three observational studies. upper GI complications compared with

4

• Nabumetone or etodolac: There was • One meta-analysis of celecoxib found no insufficient evidence to make reliable difference between celecoxib and nonselective judgments about relative GI safety and no NSAIDs, but there were few events. evidence on CV safety. • In one fair-quality cohort study, nabumetone was • Nonselective NSAID vs. nonselective NSAID or associated with a lower risk of all-cause mortality any COX-2 selective NSAID compared with diclofenac and naproxen, but this • No clear difference in GI safety was found finding has not been replicated. among nonselective NSAIDs at commonly Harms: hypertension, congestive heart failure used doses. (CHF), edema, and impaired renal function • The CV safety of naproxen was moderately • All NSAIDs and COX-2 inhibitors can cause or superior to that of any COX-2 selective NSAID aggravate these conditions. in a large systematic review of RCTs. • There is good evidence from short-term trials that, • There were 3.3 additional myocardial on average, nonselective NSAIDs raise mean infarctions for every 1,000 patients treated blood pressure by about 5.0 mm Hg (95-percent with any COX-2 inhibitor instead of confidence interval [CI] 1.2 to 8.7). However, naproxen for 1 year. similar average blood pressure changes may not necessarily correspond with similar likelihoods of • The CV safety of nonselective NSAIDs other an event requiring withdrawal, medication change, than naproxen (data primarily on ibuprofen and or other clinical consequences. diclofenac) was similar to that of COX-2 selective NSAIDs in a large systematic review. • Evidence from good-quality observational studies suggests that rofecoxib is associated with greater • In indirect analyses, naproxen was the only risks of hypertension, CHF, and edema than nonselective NSAID associated with neutral celecoxib. Indirect evidence from various meta- CV risk relative to placebo. analyses of either rofecoxib or celecoxib vs. • Aspirin nonselective NSAIDs are consistent with these • Aspirin is associated with a lower risk of findings. Direct randomized trial evidence, thromboembolic events and a higher risk of GI however, is limited in quantity and difficult to bleeds compared to placebo or nonuse when interpret because of possible non-equivalent given in long-term prophylactic doses. dosing of drugs. Evidence regarding the comparative risk of renal dysfunction for celecoxib • There is insufficient evidence to assess the and rofecoxib is sparse. balance of GI and CV safety of higher dose aspirin as used for pain relief compared with • There was weak evidence that aspirin and sulindac nonaspirin NSAIDs. have less hypertensive effect than other nonselective NSAIDs. • Salsalate • There were no clear differences among other • Salsalate was associated with a lower risk of selective or nonselective NSAIDs for these adverse adverse events than other selective and events. nonselective NSAIDs using broad composite endpoints in older, poor-quality observational Harms: hepatotoxicity studies. In a more recent observational study, • Clinically significant hepatotoxicity was rare. salsalate had a similar rate of complications • Among currently marketed NSAIDs, only compared with other NSAIDs. diclofenac was associated with a significantly • Almost no data are available on CV safety. higher rate of liver-related discontinuations compared with placebo (1 additional case for Harms: mortality every 53 patients treated with diclofenac). • Individual trials were not large enough to detect differences in mortality between the included Tolerability drugs. • Relative to nonselective NSAIDs, COX-2 selective and partially selective NSAIDs were better or

5

similarly tolerated and aspirin was less well baseline pain, there was a modest benefit for tolerated. pain relief, but this did not appear to be a • There were no clear differences in tolerability preplanned analysis. among COX-2 selective or nonselective NSAIDs. • Systematic reviews of older trials found • Uncertainty remains regarding the comparative glucosamine modestly superior to oral NSAIDs tolerability of salsalate and nonselective NSAIDs. and placebo in most trials, but there was some Available evidence is somewhat sparse and mixed, inconsistency between trials, most trials had with two of three short-term trials suggesting some flaws and results may not be directly salsalate is less well tolerated than nonselective applicable to the United States because the NSAIDs and older, flawed observational studies positive trials primarily evaluated suggesting that salsalate is less toxic than pharmaceutical-grade glucosamine available in nonselective NSAIDs. Europe. • Only 2 of 20 placebo-controlled trials assessed Other oral agents: benefits and harms effects of glucosamine on radiologic disease • Acetaminophen progression. One fair- and one good-quality trial found pharmaceutical-grade glucosamine • Acetaminophen was modestly inferior to superior to placebo for progression of knee NSAIDs for pain and function in four joint space narrowing over 3 years. systematic reviews. • Glucosamine and chondroitin were generally • Pain severity ratings averaged less than 10 well tolerated and no serious adverse events points higher for acetaminophen were reported in clinical trials. compared to NSAIDs on 100-point visual analog scales. Effect of dosage and duration of treatment on • Compared with NSAIDs, acetaminophen had the benefits and harms of oral medication use fewer GI side effects (clinical trials data) and • We found no studies evaluating the GI or CV serious GI complications (observational safety of alternative dosing strategies (such as studies). alternate day dosing, once daily versus twice daily • Acetaminophen may be associated with modest dosing, or periodic drug holidays). increases in blood pressure and renal • The risk of GI bleeding increases with higher dysfunction (observational studies). doses of nonselective NSAIDs. • One good-quality, prospective observational • The most comprehensive systematic review of study found an increased risk of CV events RCTs found no clear association between duration with heavy use of acetaminophen that was of exposure and CV risk of COX-2 inhibitors. similar to the risk associated with heavy use of However, estimates of CV risk with shorter NSAIDs. duration of exposure are imprecise due to low • Acetaminophen at therapeutic doses does not numbers of events. appear to be associated with an increased risk • The most comprehensive systematic review of of hepatotoxicity compared to nonuse in RCTs found higher doses of celecoxib associated patients without underlying liver disease. with increased CV risk, but could not determine • Glucosamine and chondroitin the effects of dose on CV risk associated with rofecoxib due to low numbers of events at lower • In one large, good-quality trial the combination doses. Most trials of nonselective NSAIDs of pharmaceutical-grade glucosamine involved high doses. hydrochloride plus chondroitin (not currently available in the United States) was not superior Differences in demographic and clinical to placebo among all patients studied. Neither subgroups glucosamine nor chondroitin alone was • GI and CV complication rates are higher among superior to placebo. In an analysis of a small older patients and those with predisposing subgroup of patients with at least moderate comorbid conditions, but there is no evidence that

6

the relative safety of different NSAIDs varies Effects of coprescribing H2-antagonists, according to baseline risk. misoprostol, or PPIs • Compared to nonuse of NSAIDs, one • Consistent evidence from good-quality systematic additional death per 1 year of use occurred for reviews and numerous clinical trials found every 13 patients treated with rofecoxib, 14 coprescribing of PPIs to be associated with the with celecoxib, 45 with ibuprofen, and 24 with lowest rates of endoscopically detected duodenal diclofenac in one large, population-based ulcers relative to gastroprotective agents. observational study of high-risk patients with • Coprescribing of misoprostol is associated with acute myocardial infarction. similar rates of endoscopically detected gastric • There is no evidence that the comparative safety or ulcers as coprescribing of PPIs. efficacy of specific selective or nonselective • While misoprostol offers the advantage of being NSAIDs varies depending on age, gender, or racial the only gastroprotective agent to reduce rates of group, although data are sparse. perforation, obstruction, or bleeding, there is a • Among patients who had a recent episode of upper high rate of withdrawals due to adverse GI GI bleeding, there is good evidence that rates of symptoms. recurrent ulcer bleeding are high (around 5 percent • The risk of endoscopic duodenal ulcers for after 6 months) in patients prescribed celecoxib or standard-dose H2 blockers was lower than a nonselective NSAID plus a PPI. placebo, similar to misoprostol, and higher than Concomitant anticoagulant use omeprazole. Standard dosages of H2 blockers were associated with no reduction of risk for • Concomitant use of anticoagulants (e.g., warfarin) gastric ulcers relative to placebo. and any nonselective NSAID increases the risk of GI bleeding three- to sixfold compared to • Double (full) dose H2 blockers were associated anticoagulants alone. with a lower risk of endoscopic gastric and duodenal ulcers relative to placebo. It is unknown • Reliable conclusions about the safety of selective how full-dose H2 blockers compare to other NSAIDs used with anticoagulants are not possible antiulcer medications because head-to-head trials due to flaws in existing observational studies, are lacking. although there are case reports of serious bleeding events, primarily in the elderly. Comparison of oral medications with topical preparations Concomitant aspirin use • Topical NSAIDs: efficacy • In the CLASS studies, there was no difference in • Studies of topical NSAIDs typically evaluated rates of ulcer complications between celecoxib and proprietary formulations not approved by the nonselective NSAIDs in the subgroup of patients FDA. who took aspirin. • Topical NSAIDs were similar to oral NSAIDs • Concomitant low-dose aspirin use increased the for pain relief in trials primarily of patients rate of endoscopic ulcers by about 6 percent in with osteoarthritis of the knee, with topical both patients on celecoxib and those on diclofenac (often with dimethyl sulphoxide nonselective NSAIDs in one meta-analysis. [DMSO], a drug not approved for use in • Rofecoxib plus low-dose aspirin or ibuprofen humans in the United States) best studied. alone were associated with similar risks of • Topical ibuprofen was superior to placebo in endoscopic ulcers (16-17 percent), which were several trials. significantly higher than those for placebo (6 percent) or aspirin alone (7 percent). • Topical NSAIDs: safety • The most comprehensive systematic review of • Consistent evidence from good-quality trials, RCTs found that compared to nonuse of aspirin, systematic reviews, and observational studies concomitant aspirin use did not ameliorate the found topical NSAIDs to be associated with increased risk of vascular events associated with increased local adverse events compared with COX-2 selective NSAIDs. oral NSAIDs.

7

• Total adverse events and withdrawal due to Remaining Issues adverse events were similar. • The CV safety of nonselective NSAIDs has not • Data from one good-quality trial found topical been well studied in large, long-term clinical trials. NSAIDs superior to oral NSAIDs for GI Naproxen, in particular, may be associated with events, including severe events, and changes in fewer CV risks than other NSAIDs and should be hemoglobin. investigated in long-term, appropriately powered • Topical salicylates and capsaicin trials. • Topical salicylates were no better than placebo • Large observational studies assessing the safety of in higher quality placebo-controlled trials. NSAIDs have been helpful for assessing comparative benefits and harms but have generally • Compared to placebo, one additional patient had a narrow focus on single adverse events. achieved pain relief for every eight that used Observational studies that take a broader view of topical capsaicin in a good-quality meta- all serious adverse events would be substantially analysis, but capsaicin was associated with more helpful for assessing the overall tradeoffs increased local adverse events and withdrawals between benefits and harms. due to adverse events. • The CV risks and GI benefits associated with Balance of evidence and harms different COX-2 selective NSAIDs may vary. Each of the analgesics evaluated in this report was Large, long-term trials with active and placebo- associated with a unique set of benefits and risks. Each controlled arms would be needed to assess the was also associated with gaps in the evidence necessary safety and benefits of any new COX-2 selective to determine the true balance of benefits vs. harms. The . role of selective and nonselective oral NSAIDs and • Meta-analyses of the risks associated with alternative agents will continue to evolve as additional selective COX-2 inhibitors need to continue to information emerges. At this time, although the assess the effects of dose and duration as more amount and quality of evidence vary, no currently data become available; current estimates of risks at available analgesic reviewed in this report was lower doses and with shorter duration of exposure identified as offering a clear overall advantage are less precise than estimates at higher doses and compared with the others. This is not surprising, given longer duration of exposure because of small the complex tradeoffs between the many benefits (pain numbers of events. relief, improved function, improved tolerability, and • Large, long-term trials of the GI and CV safety others) and harms (CV, renal, GI, and others) involved. associated with full-dose aspirin, salsalate, or Individuals are likely to differ in how they prioritize the acetaminophen compared with nonaspirin NSAIDs importance of the various benefits and harms of or placebo are lacking. Recent observational data treatment. Adequate pain relief at the expense of an suggesting an increased CV risk with heavy use of increase in CV risk, for example, could be an acetaminophen highlight the need for long-term, acceptable tradeoff for some patients. Others may appropriately powered clinical trials. consider even a marginal increase in CV risk • Given the large number of patients who meet unacceptable. Factors that should be considered when criteria for aspirin prophylaxis for CV events, weighing the potential effects of an analgesic include more trials evaluating the dose-related effects of age (older age being associated with increased risks for aspirin 50-1500 mg on GI benefits and CV safety bleeding and CV events), comorbid conditions, and are needed. concomitant medication use (such as aspirin and anticoagulation medications). As in other medical • The effects of alternative dosing strategies such as decisions, choosing the optimal analgesic for an intermittent dosing or drug holidays have not been individual with osteoarthritis should always involve assessed. Studies evaluating the benefits and risks careful consideration and thorough discussion of the associated with such strategies compared with relevant tradeoffs. conventional dosing could help clarify the effects of these alternative dosing strategies. In addition, although there is speculation that once daily versus

8

twice daily dosing of certain COX-2 inhibitors there was no difference for the prepecified, could reduce CV risk, this hypothesis has not yet primary composite renal outcome of peripheral been tested in a clinical trial. edema, hypertension, renal dysfunction, or • Most trials showing therapeutic benefits from arrhythmia. There was no clear association glucosamine were conducted using between other COX-2 inhibitors pharmaceutical-grade glucosamine not available in (valdecoxib/, etoricoxib, or lumiracoxib) the United States and may not be applicable to and either arrhythmia or renal events (no currently available over-the-counter preparations. arrhythmia events reported with lumiracoxib). Large trials comparing currently available over- • A good-quality meta-analysis of cardiovascular the-counter preparations of glucosamine and risk (primarily myocardial infarction) from 23 chondroitin with oral NSAIDs are needed, as these observational studies was largely consistent with are likely to remain available even if the FDA our qualitative assessment of the observational approves pharmaceutical-grade formulations. literature. It found rofecoxib associated with a • No topical NSAIDs are FDA approved in the dose-dependent, increased risk of cardiovascular United States, yet compounding of NSAIDs is events that was detectable during the first month widely available. Although recent trials of topical of treatment. Of the other NSAIDs, diclofenac NSAIDs are promising, most have been conducted was associated with the highest risk, followed by using a proprietary formulation of diclofenac with indomethacin and meloxicam. Celecoxib, DMSO, which is not approved in the United States naproxen, piroxicam, and ibuprofen were not for use in humans. Cohort studies using large associated with increased risks. Assessments of observational databases may be required to increased risk were modest (relative risks all <2.0), adequately assess CV risk. and all of the main analyses were associated with substantial between-study heterogeneity. Addendum Full Report As this report was going to press, two relevant meta- analyses on risks associated with NSAIDs were This executive summary is part of the following published. We were unable to fully incorporate these document: Chou R, Helfand M, Peterson K, Dana T, studies into this report, but found their results generally Roberts C. Comparative Effectiveness and Safety of consistent with our conclusions: Analgesics for Osteoarthritis. Comparative Effectiveness Review No. 4. (Prepared by the Oregon • A fair-quality meta-analysis of arrhythmia and Evidence-based Practice Center under Contract No. renal event (peripheral edema, hypertension, or 290-02-0024.) Rockville, MD: Agency for Healthcare renal dysfunction) risk from 114 randomized trials Research and Quality. September 2006. Available at: of COX-2 selective NSAIDs found rofecoxib www.effectivehealthcare.ahrq.gov/reports/final.cfm. associated with increased risks of arrhythmia (primarily ventricular fibrillation, cardiac arrest, or For More Copies sudden cardiac death) and renal dysfunction (peripheral edema, hypertension, or renal For more copies of Comparative Effectiveness and dysfunction) relative to control treatments Safety of Analgesics for Osteoarthritis: Executive (placebo, other NSAIDs, or mixed/other) . The Summary. No. 4 (AHRQ Pub. No. 06-EHC009-1), increased risk was equivalent to approximately 1.1 please call the AHRQ Clearinghouse at 1-800-358-9295 additional arrhythmia events per 1,000 patients or e-mail [email protected]. treated with rofecoxib. Celecoxib was associated with lower risks of renal dysfunction and hypertension than control treatments, although

9

ents are reater in v leeding is leeding ents increases v vidence that risk of vidence that vidence that vidence that risk of leeding episodes. vidence that vidence that risks of oups leeding and CV leeding is g ents are higher in ents increases with age. reater in patients with air e air e leeding episodes. v v with age. GI b patients with prior b Good e risk of GI b g prior b F e Good e risk of GI b and CV e patients with cardiac and renal comorbidities. Good e GI b e Good e F risks of CV and renal CV and renal e higher in patients with cardiac and renal comorbidities. • • Special considerations in • • • • subgr . xib xib, ents) ents. v v ents ter v reater risk e X-2 v X-2 e NSAIDs. e vs. v v ents than celeco xib and celeco e NSAIDs, v , and other y selecti v ents with v tiall . ut misoprostol is et due to life-threatening et in September 2004, xib X-2 selecti vidence that CO e NSAIDs are y to rofeco v e to nonselecti e NSAIDs are associated with er doses and with shor v ents compared to CO v e and par w xib is associated with g uprofen and diclofenac air e v arction) compared to placebo. v xicam and poor for etodolac e NSAIDs are sparse, er serious GI e xceptions: v y selecti w). F w y primaril reater dosages and durations of vidence that coprescription of w e tiall wn from the mark wn from the mark ocardial inf xib, and lumiraco y selecti , edema, and cardiorenal e y air for melo tiall y m vidence of fe e data on CV risks of CO ointestinal, cardiovascular as withdra e NSAIDs compared to nonselecti v as withdra irst 6 months of treatment. v xib, etorico y increase with g vidence (f xceptions (see belo ut estimates of risks at lo vidence that all nonselecti e and par tension, CHF xib w v y because of CV risks. umetone) that par le, dose-dependent increases in risk of serious GI e e NSAIDs. xib w w e vidence suggests that rofeco v vidence that high doses of ib e NSAIDs are associated with increased risks of serious aldeco yper ents (primaril thritis, with Strength of Evidence v ell tolerated. y similar risks of serious CV e air to good e v r Comparati Data on CV risks of nonselecti F Good e air e aldeco air e : : X-2 selecti skin reactions and increased CV risk. of h V associated with decreased risk relati and nab primaril as CV safety data are less precise (due to small numbers of e for v car selecti • No clear e CV e CV risks ma with a fe • Rofeco GI: CO CV nonselecti • Cautions about CV risk appl Other • GI: comparab compared to nonuse. Good e CV • F misoprostol or PPIs can attenuate this risk, b less w selecti durations of treatment are imprecise due to small numbers e treatment, b at least in the f NSAIDs are sparse, with a fe • F Harms: gastr • • • • • Osteoar e v e e v y v le in y of Findings on Comparative Effectiveness and Safety Analgesics for y . icac e NSAIDs f v X-2 selecti X-2 icac f le in ef . y selecti le in ef tiall e NSAIDs are vidence CO vidence CO vidence nonselecti y to each other v icac f ef Good e Good e selecti (pain relief) to nonselecti NSAIDs. are comparab and par Good e to each other NSAIDs are comparab comparab able A. Summar Benefits: symptom relief T • • • e e v v e v y selecti en), x X-2 selecti tiall reatment NSAIDs CO nonselecti (including NSAIDs : napro par T

10

Special considerations in Treatment Benefits: symptom relief Harms: gastrointestinal, cardiovascular, and other subgroups

• Fair evidence that naproxen is associated with a lower risk • Fair evidence that using of CV events than COX-2 selective NSAIDs and no excess NSAIDs concomitantly risk compared to placebo. with anticoagulants • Other: Fair evidence that diclofenac is associated with higher increases GI bleeding risk rates of aminotransferase elevations than other NSAIDs. three- to sixfold.

Aspirin/ • No evidence comparing • Good evidence that aspirin 50-1500 mg (for thrombotic event • Good evidence that salsalate efficacy of aspirin or prophylaxis) is associated with greater risks of serious GI concomitant use of aspirin salsalate to COX-2s or . events compared to placebo or when added to warfarin. attenuates or eliminates the NSAIDs • Good evidence that low-dose aspirin is effective for preventing GI benefits of COX-2 CV events. selective NSAIDs. • Insufficient evidence to assess GI and CV risks associated • Fair evidence that with higher doses of aspirin for pain control or with salsalate. concomitant use of low- dose aspirin does not eliminate CV risks when added to NSAIDs . 11 Acetaminophen • Good evidence that • Good evidence of lower risk of GI complications with None acetaminophen is modestly acetaminophen compared to NSAIDs. inferior in efficacy compared • Fair evidence of increased risk of blood pressure and renal to NSAIDs. dysfunction with acetaminophen compared to nonuse. • Poor evidence (a single observational study) that heavy use of acetaminophen carries a similar CV risk compared to heavy use of NSAIDs. Glucosamine • Fair evidence (some • Good evidence that glucosamine and chondroitin are well None (pharmaceutical inconsistency between tolerated and do not appear to be associated with serious grade)/ clinical trials) clinical adverse events. that pharmaceutical-grade glucosamine and chondroitin are not more effective than placebo in unselected patients, including one recent, large, good- quality trial finding no beneficial effects from glucosamine or chondroitin alone or in combination.

ug; y dr September 2006 . No. 06-EHC009-1 oups Special considerations in None subgr AHRQ Pub er le w globin , and other als due als due to w w astrointestinal; NSAID = nonsteroidal antiinflammator ents, and changes in hemo v ents and withdra ents compared with oral NSAIDs. ents and withdra ascular; GI = g v v v v ere e v erse e erse e erse e ointestinal, cardiovascular ents. v ents compared to placebo. v vidence that topical NSAIDs are associated with vidence that topical capsaicin is associated with vidence that topical and oral NSAIDs are comparab vidence that topical NSAIDs are associated with fe erse e ents, including se v erse e xygenase; CV = cardio in rates of total adv to adv Good e increased local adv compared to oral NSAIDs. Good e Good e GI e Good e increased local adv adv • • • • Harms: gastr yclo o y X = c le ailab ere y v ylates are v y are ailure; CO icac ysis. t f it for pain f y se maceutical- aluate proprietar thritis. v ysis of a small le to oral NSAIDs e hear v y of patients with vidence the vidence of no clear vidence that capsaicin, roup of patients with at een phar mulations not a . ference in ef in the United States. NSAIDs e for rade glucosamine or ut this did not apear to be a ut not topical salic air e air e F In an anal subg least moderatel baseline pain in the latter trial, there appeared to be a modest benef relief from the combination, b preplanned anal F dif betw g chondroitin and NSAIDs. No studies compared glucosamine or chondroitin to acetaminophen. Good e comparab compared to placebo. for pain relief in trials primaril knee osteoar b superior for pain relief • Most trials of topical • • • • Benefits: symptom relief CHF = congesti ylates viations: e opical opical reatment br T T NSAIDs T salic and capsaicin Ab PPI = proton pump inhibitor

12