Reports

This report is based on medical evidence presented at sanctioned medical congress, from peer reviewed literature or opinion provided by a qualified healthcare practitioner. The consumption of the information contained within this report is intended for qualified Canadian healthcare practitioners only.

MEDICAL OPTIONS - Pain Management

September 2009

SELECTING NSAIDs IN OSTEOARTHRITIS: A PRACTICAL APPROACH

Editorial Review:

Algis V. Jovaisas, MD, FRCPC

Assistant Professor of Medicine, Division of Rheumatology, University of Ottawa, Ottawa, Ontario

Many factors conspire to emphasize the need for a rational approach to the treatment of osteoarthritis (OA). Although non-pharmacological lifestyle modifications should underlie all OA management, the goal of pain control in moderate to severe OA is likely to demand a nonsteroidal anti-inflammatory drug (NSAID). NSAIDs are effective but accompanied by an increased potential for gastrointestinal (GI) and cardiovascular (CV) adverse events, particularly in aging patients. Importantly, the risk of these events is strongly influenced by relative susceptibility. An appreciation of the determinants of GI and CV risk is critical when determining whether to offer a conventional non-selective NSAID, an NSAID selective for the cyclooxygenase-2 (COX-2) enzyme, an adjunctive gastroprotective agent to combine with a conventional NSAID, or an alternative analgesic in order to avoid NSAIDs. Progress in quantifying relative risks has improved guidance for appropriate clinical choices.

Increase in Age-related Risks

OA is reasonably characterized as a disease of aging. Although OA can develop in relatively young individuals, there is a linear increase in prevalence starting at about age 50 and extending until at least age 80.¹ By the age of 70, approximately onethird of men and 40% of women have OA (Figure 1). The parallel age-related increases in risk of GI bleeding and CV disease are relevant to the management of OA because the mainstay of therapy, NSAIDs, influence both. The risk of GI bleeding, a potentially fatal condition in elderly individuals (73% of deaths due to an upper GI bleed occur in individuals over the age of 60²), is increased by more than fourfold in current users of non-selective NSAIDs, relative to placebo.³ CV events, which also increase in a linear fashion relative to age,⁴ have had a more variable relationship with NSAIDs, whether non-selective or COX-2-selective, but labelling for all NSAIDs now acknowledges CV risk and guidelines suggest CV risk should be considered in OA management.⁵

Figure 1.

In the treatment of OA, NSAIDs are dominant therapies, but it is important to recognize that the first line of therapy should be lifestyle modifications, particularly in individuals with mild disease. Often overlooked, exercise has been associated with improvement in symptoms of OA in multiple studies, including two meta-analyses.^6,7 Similar data, including a meta-analysis, support the benefit of moderate weight loss in overweight individuals (Table 1).⁸ Educating patients about their disease and the benefits of even moderate increases in exercise and reductions in weight are important steps toward better disease management even among tho
erapies.

Table 1.

<img3451|center>

Managing GI Risk

In OA, the goals of therapy are pain relief and improvement in function, both of which are essential to an adequate quality of life. In patients with mild OA, acetaminophen, which has been found more effective for pain relief relative to placebo in numerous studies,⁹ may be sufficient to reach these goals. Although the risk of significant toxicity from acetaminophen is low, conventional regimens should be employed due to the dose-related risk of renal injury.¹⁰

In patients with OA not controlled with acetaminophen, NSAIDs are effective relative to acetaminophen and placebo.¹¹ In a placebo-controlled study that directly compared celecoxib to acetaminophen in patients with OA of the hip or the knee, acetaminophen was statistically superior to placebo for pain relief, but celecoxib was significantly superior to acetaminophen as measured with a symptom index or by patient preference (Figure 2).¹² The choice between a non-selective and selective NSAID is dictated by relative risks rather than relative efficacy. Although there are many studies comparing non-selective to COX-2-selective NSAIDs on the basis of efficacy, large studies undertaken to compare relative GI safety in OA or rheumatoid arthritis (RA), such as CLASS,¹³ VIGOR,¹⁴ and SUCCESS,¹⁵ have associated COX-2 inhibitors with a reduced
ut a similar efficacy for control of joint symptoms.

Figure 2.

<img3452|center>

Despite the removal from the market of several COX-2-selective NSAIDs, the relative risk of GI toxicity has been and remains the primary distinction between non-selective NSAIDs and celecoxib, the only COX-2-selective NSAID currently available in Canada. While COX-2-selective inhibitors were developed specifically to reduce inhibition of COX-1, which is an enzyme important to production of gastroprotective prostaglandins,¹⁶ the efficacy of this strategy deserves emphasis. In the SUCCESS-1 study, for example, the risk of upper GI complications was reduced by 86% with celecoxib relative to naproxen or diclofenac. In the VIGOR study with rofecoxib, which was subsequently removed from the market because of an association with increased CV risk, there was a 54% reduction in upper GI complications relative to naproxen.

Treatment Choices with Increased CV Risk

An unanticipated CV risk led to removal of several COX-2-selective NSAIDs, such as rofecoxib, but CV risk has not proven to be COX-specific.¹⁷ Rather, in comparing currently available NSAIDs, including nonselective NSAIDs and celecoxib, for such CV events as myocardial infarction (MI), no clear pattern for risk has yet emerged.¹⁸ As a result, similar labelling in regard to CV risk is carried by all NSAIDs, including naproxen, celecoxib and diclofenac.¹⁹ As a practical approach for patients who are candidates for NSAIDs, it is therefore appropriate to consider CV risk in all patients but to concentrate choice between selective and non-selective NSAIDs by relative risk of GI complications.

The risk factors for GI bleeds include age older than 60, a previous history of upper GI bleeding and/or a previous history of peptic ulcer disease.²⁰ Smoking and Helicobacter pylori infections also increase susceptibility for GI bleeds. Screening and eradication of H. pylori is particularly important in patients who have upper GI symptoms. Patients should also be screened for medications that increase risk of GI bleeding, such as corticosteroids. In patients on low-dose ASA, non-selective NSAIDs can more than double the risk of bleeding.²¹

In patients at high risk of a GI bleed, both a non-selective NSAID combined with a proton pump inhibitor (PPI) or a COX-2- selective NSAID have been associated with a reduced risk of symptomatic peptic ulcers, which is likely to be a surrogate for bleeding risk.²² However, celecoxib is a reasonable first choice because it provides relative protection to both the upper GI tract and the small bowel, which is unaffected by PPI therapy. In addition, PPIs have been associated with an increased risk of osteoporotic fracture,²³ which is a concern in an elderly population. Except for pantoprazole, there is also a potential interaction between PPIs and clopidogrel.²⁴ These drugs have also been associated with an increased incidence of hospitalacquired pneumonia and Clostridium difficile infections.²⁵

In patients at high CV risk, Health Canada and the U.S. Food and Drug Administration have both adopted the position that nonselective NSAIDs and celecoxib cannot be differentiated for relative safety. In a retrospective analysis that compared relative CV incidence rates among patients at particularly high risk (those hospitalized for coronary heart disease) on NSAIDs, celecoxib and naproxen were associated with the lowest relative incidence rates on a numerical basis, but most differences between selective and non-selective NSAIDs did not reach statistical significance.¹⁹ The exception was diclofenac, which was associated with a significantly increased incidence rate relative to those not taking an NSAID (Table 2). A prospective, randomized study comparing celecoxib to traditional NSAIDs for CV events is needed for definitive data on relative risk, and such a trial is now underway. However, in patients with elevated CV risk who require an NSAID and are taking low-dose ASA, celecoxib may be one of the reasonable choices. Unlike ibuprofen and naproxen, celecoxib does not appear to negate the antiplatelet effects of ASA.²⁶ The risk of interactions has not been studied with all NS
ective NSAIDs, such as diclofenac, may also preserve the activity of ASA.²⁷

Table 2.

<img3453|center>

Balancing Treatment and Risk

In patients at very high risk of both GI and CV events, celecoxib plus a PPI may be a reasonable choice. Although the CV safety of this combination has not been demonstrated in a prospective randomized trial, this combination has been associated with a low risk of recurrent upper GI bleeding relative to alternative strategies.²⁸ In any patient with elevated CV risk, aggressive management of modifiable risk factors should be a routine component of therapy.

Although it may be reasonable to employ the lowest effective doses of NSAIDs in patients at high CV risk, it is not clear whether CV risk is dose-related. The decision to forego NSAIDs altogether in patients at elevated CV risk may be complicated by a limited selection of alternative therapies. Joint replacement is one such strategy. A low-dose narcotic regimen for pain control is another, but analgesics other than NSAIDs often introduce new clinical concerns, particularly in the elderly for whom there is a delicate balance between risks and benefits. For example, tramadol, a synthetic opioid, may be well tolerated by some individuals, but this agent, like most narcotics, is associated with constipation. As it may create a sense of euphoria, it is also associated with some abuse potential.²⁹

Due to an aging population and increasing rates of obesity, a risk factor for OA, the prevalence of this disease is expected to climb precipitously over the coming decades. The increasing availability of objective data with which to replace expert opinion provides an expectation for more useful guidelines as more is understood about how to balance GI and CV risk in the management of joint pain. However, it is expected that the important differences in relative risks for these complications between patients will continue to require individualization of therapy.

Summary

Treatment strategies for OA have evolved and may continue to evolve with new information about relative risks and benefits. Although withdrawal of several COX-2 inhibitors generated concern of a class effect in regard to CV risk, subsequent data suggest that the currently available COX-2 inhibitor cannot be differentiated from non-selective NSAIDs for CV outcomes. Rather, the decision to employ a selective vs. a non-selective NSAID should be directed primarily by relative GI risk. While specific regimens should be modified according to both GI and CV risks, NSAIDs are typically required in many OA patients with moderate to severe disease when the goal is sufficient symptom control to maintain an adequate quality of life.

References

1. Kopec et al. Descriptive epidemiology of osteoarthritis in British Columbia, Canada. J Rheumatol 2007; 34:386-93.

2. Yavorski RT, Wong RK, Maydonovitch C. Analysis of 3,294 cases of upper gastrointestinal bleeding in military medical facilities. Am J Gastroenterol 1995;90:568-73.

3. Offman et al. A meta-analysis of severe upper gastrointestinal complications of non-steroidal antiinflammatory drugs. J Rheumatol 2002;29:804-12.

4. Canadian Heart and Stroke Foundation. Statistics. Website: www.heartandstroke.com/site/c.ikIQLcMWJtE/ b.3483991/k.34A8/Statistics.htm#deaths. Accessed May 18, 2009.

5. American College of Rheumatology Ad Hoc Group on use of selective and non-selective nonsteroidal anti-inflammatory drugs. Arthritis Rheum 2008;59: 1058-73.

6. Fransen M, McConnell S, Bell M. Therapeutic exercise for people with osteoarthritis of the hip or knee. A systematic review. J Rheumatol 2002;29:1737-45.

7. Pelland et al. Efficacy of strengthening exercises for osteoarthritis (Part I): A meta-analysis. Phys Ther Rev 2004;9:77-108.

8. Christensen et al. Effect of weight reduction in obese patients diagnosed with knee osteoarthritis: a systematic review and meta-analysis. Ann Rheum Dis 2007;66:433-9.

9. Towheed et al. Acetaminophen for osteoarthritis. Cochrane Database Syst Rev 2006;1:CD004257. 10. Perneger TV, Whelton PK, Klag MJ. Risk of kidney failure associated with the use of acetaminophen, aspirin, and nonsteroidal antiinflammatory drugs. N Engl J Med 1994;331:1675-9.

11. American College of Rheumatology Subcommittee on Osteoarthritis Guidelines. Recommendations for the medical management of osteoarthritis of the hip and knee: 2000 update. Arthritis Rheum 2000;43: 1905-15.

12. Pincus et al. Patient preference for placebo, acetaminophen, (paracetamol), or celecoxib efficacy studies (PACES): two randomized, double blind, placebo controlled crossover clinical trials in patients with knee or hip osteoarthritis. Ann Rheum Dis 2004;63:931-9.

13. Silverstein et al. Gastrointestinal toxicity with celecoxib vs. nonsteroidal antiinflammatory drugs for osteoarthritis and rheumatoid arthritis. The CLASS study: a randomized controlled trial. JAMA 2000;284:1247-55.

14. Bombardier et al. Comparison of upper gastrointestinal toxicity of rofecoxib and naproxen in patients with rheumatoid arthritis. N Engl J Med 2000;343:1520-8. 15. Singh et al. Celecoxib versus naproxen and diclofenac in osteoarthritis patients: SUCCESS-1 study. Am J Med 2006;119:255-66.

16. Vane JR, Bakhle YS, Botting RM. Cyclooxygenase 1 and 2. Annu Rev Pharmacol Toxicol 1998;38:97-120. 17. Dajani EZ, Islam K. Cardiovascular and gastrointestinal toxicity of selective cyclooxygenase-2 inhibitors in man. J Physiol Pharmacol 2008;59(suppl 2):117-33.

18. Hernández-Díaz S, Varas-Lorenzo C, García Rodriguez LA. Non-steroidal anti-inflammatory drugs and the risk of acute myocardial infarction. Basic Clin Pharmacol Toxicol 2006;98:266-74.

19. Ray et al. Cardiovascular risks of nonsteroidal antiinflammatory drugs in patients after hospitalization for serious coronary heart disease. Circ Cardiovasc Qual Outcomes 2009;2:155-63.

20. Dincer et al. NSAID-related upper gastrointestinal bleeding: are risk factors considered during prophylaxis? Int J Clin Pract 2006;60:546-8.

21. Lanas et al. Nitrovasodilators, low-dose aspirin, other nonsteroidal antiinflammatory drugs, and the risk of upper gastrointestinal bleeding. N Engl J Med 2000; 343:834-9.

22. Hooper et al. The effectiveness of five strategies for the prevention of gastrointestinal toxicity induced by non-steroidal anti-inflammatory drugs: systematic review. BMJ 2004;329:948-53.

23. Targownik et al. Use of proton pump inhibitors and risk of osteoporosis-related fractures. CMAJ 2008;179:319-26.

24. Ho et al. Risk of adverse outcomes associated with concomitant use of clopidogrel and proton pump inhibitors following acute coronary syndrome. JAMA 2009;301:937-44.

25. Herzig et al. Acid-suppressive medication use and the risk for hospital-acquired pneumonia. JAMA 2009;301:2120-8.

26. Renda et al. Celecoxib, ibuprofen, and the antiplatelet effect of aspirin in patients with osteoarthritis and ischemic heart disease. Clin Pharmacol Ther 2006;80:264-74.

27. Bird et al. A study to determine the clinical relevance of the pharmacokinetic interaction between aspirin and diclofenac. Inflamm Res 1986;18:447-9.

28. Chan et al. Combination of a cyclo-oxygenase-2 inhibitor and a proton pump-inhibitor for prevention of recurrent ulcer bleeding in patients at very high risk: a double-blind, randomized trial. Lancet 2007;369(9573):1621-6.

29. Adams et al. A comparison of the abuse liability of tramadol, NSAIDs, and hydrocodone in patients with chronic pain. J Pain Symptom Manage 2005; 31:465-76.

CONSIDERING GASTROINTESTINAL RISK IN CANDIDATES REQUIRING ANTI-INFLAMMATORY DRUGS

Editorial Review:

Richard Hunt, MD, FRCP, FRCPC, FACG, AGAF

Farncombe Family Digestive Disease Research Institute, McMaster University Health Science Centre, Professor of Medicine, Division of Gastroenterology, McMaster University, Hamilton, Ontario

Gastrointestinal (GI) complications are a common limitation of non-steroidal antiinflammatory drugs (NSAIDs). A variety of strategies have been developed to reduce the GI risk. These include co-therapy with a gastroprotective agent, including proton pump inhibitors (PPIs), H₂-receptor antagonists and the selection of NSAIDs with a greater selectivity for the cyclooxygenase-2 (COX-2) isoenzyme (coxibs). In the aging adult population, in whom chronic NSAIDs are most frequently prescribed and in whom the GI risks are greatest, the currently available coxib and the traditional NSAIDs (t-NSAIDs) pose comparable cardiovascular (CV) and renovascular risks. However, a coxib may be a more reasonable choice than a t-NSAID in combination with a PPI in patients in several specific situations.¹ This includes patients taking low-dose ASA because of the additional protection offered by the coxib but not a PPI to the small bowel.

NSAIDs, many of which are available without a prescription, are commonly used for control of pain and inflammation. The risk of serious GI complications from even short courses of these agents, although low in relatively young and otherwise healthy individuals, increases steeply in those with a history of a previous gastric or duodenal ulcer, and particularly those with a previous history of GI bleeding, those taking other medications that adversely influence gastric integrity, and the elderly.² Rational strategies for GI risk reduction have improved with more information and a better understanding about the characteristics of GI risk, the impact of current options for risk reduction on the CV system, and how to define benefit in the context of risk.

Aging Population and GI Complications

Aging adults, who are at particularly high risk for serious GI complications, are among the most frequent consumers of NSAIDs due primarily to the high prevalence of osteoarthritis (OA), a painful, progressive and age-related condition that can be detected radiographically in the majority of individuals over the age of 55.³ The concentration of deaths from upper GI bleeding in older individuals has been attributed to chronic NSAID therapy for control of OA, but the assertion that short courses of NSAIDs circumvent the risk of major complications has not been supported by clinical data. Rather, some studies indicate that complications may be highest during the early period of exposure.⁴ Although other studies demonstrate a relatively constant risk over time, the preponderance of data indicate that no period of NSAID therapy can be considered safe.⁵

The introduction of NSAIDs with greater selectivity for the COX-2 enzyme relative to the COX-1 enzyme was initially embraced as an important alternative to t-NSAIDs for GI risk reduction. The greater affinity for COX-2 preserved the anti-inflammatory and analgesic effects and the clinical benefits of t-NSAIDs while significantly reducing the risks of clinically important upper GI ulcers and their complications. They are also associated with lower rates of dyspepsia.^4,6 Compared to t-NSAIDs, coxibs reduce the risk of serious GI events by approximately 50%.⁷ An association between the coxibs and an increased risk of CV events including myocardial infarction led to the market withdrawal of all but one coxib in Canada. This has naturally led to confusion about the relative benefit-to-risk ratio of these agents. Subsequent evidence indicated that the risk of CV adverse events is also seen with t-NSAIDs and now all anti-inflammatory drugs carry a similar warning. The current evidence shows that the only coxib available in Canada, celecoxib, is associated with no different CV risk than t-NSAIDs but has consistently demonstrated a lower risk of GI damage as establi
scopy (Table 1).⁸ This provides a basis for re-evaluating current options for GI risk management.⁹

Table 1.

<img3454|center>

Gastroprotective Combination Treatment

The major alternative to a coxib for patients who require NSAIDs for pain control has been a gastroprotective co-therapy. This concept, which preceded the development of the coxibs, is now primarily achieved with proton pump inhibitors (PPIs) which have overtaken alternatives, such as the prostaglandin analogue misoprostol, because of their greater efficacy and tolerability.¹⁰ PPIs reduce NSAID-induced damage in the upper GI tract by lowering gastric acidity which is an important factor in gastroduodenal damage. However, acid does not play a role in NSAID-induced damage in the lower GI tract, which is the site of up to one-third of all NSAID-related GI complications.^{11</sup
is not shared by the coxibs, which have been shown to be associated with a substantially lower propensity than t-NSAIDs to induce damage to the small bowel or large intestine1 (Figure 1).}

Figure 1.

<img3455|center>

These relative differences between the two major strategies for reducing the risk of GI complications from t-NSAIDs are potentially important to several risk groups, particularly patients who are also taking low-dose ASA for CV prophylaxis. Commonly prescribed to older adults, who represent the largest population with OA, low-dose ASA poses similar risks to the GI tract, including injury to both the upper and lower GI tracts, as t-NSAIDs and particularly as a cause of bleeding. When taken in combination with t-NSAIDs, the rate of GI complications is substantially increased,¹² further emphasizing the need for some form of prophylaxis. When low-dose ASA is taken with a coxib, the risk of GI complications is also increased relative to the use of the coxib alone, but there is lower risk for ASA plus a coxib when compared to ASA plus a t-NSAID. In a metaanalysis exploring this question, the relative risk of an endoscopic ulcer with celecoxib plus ASA was 50% lower when compared to t-NSAIDs plus ASA.¹³

Strategies for Reducing GI Complications

The practical strategies for reducing GI complications in high-risk patients who require control of inflammation and acute or chronic pain involve a careful analysis of the benefits and risks. Accepting the premise that some risk is warranted to control pain, NSAIDs are often the most appropriate choice despite their potential complications. On the spectrum of analgesics, acetaminophen is a good firstline choice that poses a low risk of GI adverse events, but it offers relatively modest pain control and its association with hepatotoxicity at daily doses greater than 4 g indicates that it is not risk-free.^14,15 At the other end of this same spectrum, opioids offer potent analgesia but less anti-inflammatory effect and are often impractical, particularly in the elderly, because of the extensive list of side effects that include constipation, depressed cognitive function and other adverse central nervous system (CNS) effects.¹⁶

On the basis of efficacy and safety, NSAIDs are often the most practical and readily available option for pain control. The substantial, if not uniform, evidence that the renovascular and CV risks of celecoxib and t-NSAIDs are comparable has been an important step toward renewing attention on the GI tract as the major focus of NSAIDinduced risks for strategies to control pain.⁷ In any patient with multiple concomitant medical issues, a common situation in elderly patients who are the most frequent candidates for NSAIDs, therapy must be individualized, but there are several practical concepts that are useful for guiding treatment choice.

The focus of GI risk management has been reasonably directed at reducing the risk of GI bleeding due to the potential for life-threatening complications, but dyspepsia may not be a trivial issue when the goal of treatment is to improve quality of life. Dyspepsia often generates concern because of its role as an early sign of a serious GI event. However, these symptoms correlate poorly with the presence of erosions or ulceration.¹⁷ Rather, the more important clinical significance of dyspepsia, as well as other upper GI symptoms, such as heartburn or abdominal bloating, is the potential adverse effect on compliance with NSAIDs and the subsequent potential to increase office visits with a negative cumulative effect on well-being.

Although coxibs do not eliminate dyspepsia, a meta-analysis associated coxibs with a 12% reduction in risk of dyspepsia relative to NSAIDs, an advantage reinforced by lower rates of PPI prescriptions in patients taking a coxib vs. a t-NSAID.¹⁸

While one of the advantages of a coxib is the potential reduction in the need for gastroprotective co-therapy such as a PPI, all of the relative advantages of a coxib to a t-NSAID can be increased by employing a coxib and a PPI together for patients at high risk (Figure 2).¹⁹ This includes a reduction in the risk of serious adverse events, such as bleeding, as well as dyspepsia. The single most important disadvantage of this combination, which is well tolerated and would be expected to provide better protection against both upper and lower GI adverse events relative to a coxib alone or a t-NSAID plus PPI combination, is cost. However, a recent analysis u
ound that adding a PPI to either a t-NSAID or a coxib was cost effective for all patient groups considered, producing an incremental cost effectiveness to quality adjusted life-years gain ratio of $1650 (US).²⁰

Figure 2.

<img3456|center>

At the time of their introduction, the coxibs were widely embraced as a safe alternative to t-NSAIDs. Subsequent studies have indicated that the safety is relative rather than absolute, but these agents should be retained among the alternative strategies which can be individualized to provide analgesia and control of inflammation, including long-term pain control as required in patients on chronic NSAID therapy. Although all NSAIDs should be employed cautiously with a full understanding of the relative CV and GI safety issues, coxibs can offer advantages over t-NSAIDs in specific groups who require an NSAID to maintain an acceptable quality of life.

Summary

GI complications are the most common serious adverse events associated with NSAID therapy. Although NSAIDs have an important role in the control of pain and inflammation, particularly in older patients with OA, it is important to consider strategies to provide the optimal ratio of benefit to risk. As outlined in evidence-based guidelines for treatment choice,¹ coxibs reduce the risk of upper and lower GI complications relative to t-NSAIDs and should be considered among choices in NSAID risk management. The current data indicate that coxibs provide greater GI safety than t-NSAIDs without increasing the risk of complications to organ systems outside of the GI tract.

References

1. Lanas A, Hunt R. Prevention of anti-inflammatory drug-induced gastrointestinal damage: benefits and risks of therapeutic strategies. Ann Med 2006;38: 415-28.

2. Wolfe et al. Gastrointestinal toxicity of nonsteroidal antiinflammatory drugs. N Engl J Med 1999;340: 1888-99.

3. van Saase et al. Epidemiology of osteoarthritis: Zoetermeer Survey. Comparison of radiological osteoarthritis in a Dutch population with that in 10 other populations. Ann Rheum Dis 1989;48:271-80.

4. Bombardier et al. Comparison of upper gastrointestinal toxicity of rofecoxib and naproxen in patients with rheumatoid arthritis. N Engl J Med 2000;343:1520-8.

5. MacDonald et al. Association of upper gastrointestinal toxicity of non-steroidal anti-inflammatory drugs with continued exposure: cohort study. BMJ 1997;315:1333-7.

6. Silverstein et al. Gastrointestinal toxicity with celecoxib vs. nonsteroidal antiinflammatory drugs for osteoarthritis and rheumatoid arthritis: the CLASS study: a randomized controlled trial. JAMA 2000;284:1247-55.

7. Moore et al. Non-steroidal anti-inflammatory drugs (NSAIDs), cyclooxygenase-2 selective inhibitors (coxibs) and gastrointestinal harm: review of clinical trials and clinical practice. BMC Musculoskelet Disord 2006;7:79-86.

8. Deeks et al. Efficacy, tolerability, and upper gastrointestinal safety of celecoxib for treatment of osteoarthritis and rheumatoid arthritis: systematic review of randomised controlled trials. BMJ 2002;325(7365):619-26.

9. Hunt et al. Myths and facts in the use of antiinflammatory drugs. Ann Med 2009;41(6):423-37.

10. Hawkey et al. Omeprazole compared with misoprostol for ulcers associated with nonsteroidal antiinflammatory drugs. N Engl J Med 1998;338:727-34.

11. Laine et al. Systematic review: the lower gastrointestinal adverse effects of non-steroidal anti-inflammatory drugs. Aliment Pharmacol Ther 2006;24(5):751-67.

12. Sørensen et al. Risk of upper gastrointestinal bleeding associated with use of low-dose aspirin. Am J Gastroenterol 2000;95(9):2218-24.

13. Moore et al. Tolerability and adverse events in clinical trials of celecoxib in osteoarthritis and rheumatoid arthritis: systematic review and meta-analysis from company clinical trial reports. Arthritis Res Ther 2005;7(3):R644-65.

14. U.S. Food and Drug Administration. June 29-30, 2009: Joint Meeting of the Drug Safety and Risk Management Advisory Committee with the Anesthetic Life Support Drugs Advisory Committee and the Nonprescription Drugs Advisory Committee. Available at: http://www. fda.gov/AdvisoryCommittees/Calendar/ucm143083. htm. Accessed August 20, 2009.

15. Larson et al. Acetaminophen-induced acute liver failure: results of a United States, multicenter, prospective study. Hepatology 2005;42(6):1364-72.

16. Swegle JM, Logemann C. Management of common opioid-induced adverse effects. Am Fam Physician 2006;74:1347-54.

17. Scarpignato et al. Working team report: towards a GI safer antiinflammatory therapy. Gastroenterol Int 1999;53:185-97.

18. Spiegel et al. Comparing rates of dyspepsia with coxibs vs. NSAIDs+ PPI. A meta-analysis. Am J Med 2006;119(5):448e27-36.

19. Chan et al. Combination of a cyclo-oxygenase-2 inhibitor and a proton-pump inhibitor for prevention of recurrent ulcer bleeding in patients at very high risk: A double-blind, randomised trial. Lancet 2007;369(9573):1621-6.

20. Latimer et al. Cost effectiveness of COX2 selective inhibitors and traditional NSAIDs alone or in combination with a proton pump inhibitor for people with osteoarthritis. BMJ 2009;339:b2538.

CONSIDERING CARDIOVASCULAR RISK IN CANDIDATES FOR NSAIDs

Editorial Review:

George Honos, MD, FACC, ABIM, CSPQ, FRCPC

Head, Division of Cardiology, Centre Hospitalier de l’Université de Montréal (CHUM), Associate Professor of Medicine, Université de Montréal, Montreal, Quebec

Relative cyclooxygenase-2 (COX-2) selectivity does not differentiate nonsteroidal anti-inflammatory drugs (NSAIDs) for cardiovascular (CV) risk at commonly prescribed doses. In the retrospective studies on which CV risk for NSAIDs is based, there is variability in both the non-selective and COX-2-selective (coxib) subclasses of NSAIDs, but the hazard ratios (HRs) for CV events are not higher for the currently available coxib relative to nonselective NSAIDs on the basis of the current evidence. In practical terms for patient care, the baseline CV risk of an individual patient may be relevant to the decision to offer NSAID therapy, but it is essential to consider the balance of risks and benefits in each candidate. For example, patients rendered more active by the ability of NSAID therapy to control joint symptoms may achieve a net advantage for CV risk modification. Simple formulas may not be broadly applicable, particularly in elderly patients with multiple medical issues who derive significant quality of life benefits from effective control of inflammatory conditions.

The COX Selectivity Continuum

The anti-inflammatory effects of NSAIDs are mediated by their ability to inhibit COX which has at least two isoforms. The evidence that inhibition of the COX-1 isoenzyme is largely responsible for the gastritis associated with NSAIDs led to the development of inhibitors that are more selective for COX-2, which is the target for anti-inflammatory effects.¹ However, the selectivity of these two types of NSAIDs is relative. Inhibition of both COX-1 and COX-2 by agents within either NSAID subclass can be measured on a continuum. There is a greater but not absolute selectivity of coxibs for COX-2 so that the risk of COX-1-mediated gastritis is reduced but not eliminated.

The initial association between NSAIDs and CV risk was drawn from randomized trials with the COX-2 inhibitor rofecoxib, which was eventually withdrawn from the market for this reason.² The same CV risk association, not previously observed with nonselective NSAIDs, was unexpected. Subsequent associations with several other coxibs created an impression that this was a risk confined to COX-2-selective inhibitors.³ When similar analyses were conducted with non-selective NSAIDs, it became apparent that COX-2 selectivity was not an essential characteristic of increased CV risk.⁴ Rather, the cumulative retrospective data indicate that there is a range of risk among both selective and non-selective NSAID subclasses.

The evidence that increased CV risk may be shared has resulted in comparable product labelling for selective and non-selective NSAIDs despite the substantial variability observed within both classes. For example, when data are collated from independent studies, the HR estimations relative to no exposure to NSAIDs have been higher for diclofenac (HR 1.40; 95% CI, 1.16-1.70) than they have been for ibuprofen (HR 1.07; 95% CI, 0.97-1.18) although the 95% CIs overlap.⁵
or rofecoxib (HR 1.35; 95% CI, 1.35-1.59) is lower than that of diclofenac, although the CIs again overlap. Celecoxib (HR 1.06; 95% CI, 0.91-1.23) is associated with a comparable risk to ibuprofen and has a lower risk, without overlap of the CIs, relative to rofecoxib (Table 1).

Table 1.

<img3457|center>

Exploring Relative Risk

However, an important selection bias is one of the major limitations of these kinds of retrospective analyses, which have gathered data from studies with different study designs, different entry criteria, and different periods of exposure. For example, more of the studies with coxibs may have been conducted in elderly populations at high risk for gastrointestinal (GI) events. This selection is also likely to isolate a population at high risk for CV events. Overall, these data are not well suited to demonstrate relative risk even though they do indicate that a modest CV risk is shared among both selective and non-selective NSAIDs.

The chief mechanism by which NSAIDs increase CV risk remains uncertain. Although all NSAIDs raise blood pressure when administered in sufficient doses, celecoxib does not sha
to rofecoxib at commonly used doses. In a double-blind study comparing these two drugs in osteoarthritis (OA) patients 65 years or older, only 6.9% of celecoxib patients vs. 14.9% of rofecoxib patients (P<0.01) had a 20 mm Hg or greater rise in systolic blood pressure at any time over the six-week study6 (Figure 1).

Figure 1.

<img3458|center>

Similarly, in a comparison that included diclofenac, ibuprofen was associated with a significantly greater risk of systolic blood pressure increases of >20 mm Hg than celecoxib.7 It has also been speculated that NSAIDs increase thrombotic risk by altering platelet activity.⁸ This hypothesis was particularly attractive when the adverse effect on CV risk was thought to be confined largely to coxibs due to the importance of COX-2 to production of prostacyclin, a vasodilator that downregulates platelet aggregability. The importance of this mechanism is less clear with evidence that some non-selective NSAIDs may produce a greater CV risk than some coxibs. It now appears possible that more than one pathway may mediate the association between NSAIDs and CV risk. The potential for significant differences between NSAIDs on kidney function, which is also relevant to CV risk, is also a focus of efforts to understand the relative risks of these agents⁹.

Few pharmaceutical therapies can be described as safe in an absolute sense. In general, some set of potential risks is balanced against some set of anticipated benefits. While the CV risk of NSAIDs is negligible in otherwise healthy individuals without CV risk factors, particularly if treatment is of short duration, CV risk should be among considerations when considering an NSAID to control arthritis in the older individuals who most commonly require chronic dosing for adequate analgesia. This depends on recognizing the baseline CV risk of the patient in the context of other health issues, including the relative need for NSAIDs and their risk for a GI adverse event. Although it is important to recognize that the risk of a CV event is common to all individuals over the age of 60, many of whom die of a CV event without any significant known risk factors other than age, the contribution of NSAIDs to risk is small relative to modifiable risk factors such as smoking, hypertension or hypercholesterolemia.

Evaluating the Benefit:Risk Ratio

Patients at high risk of a CV event should already be on aggressive management, which will typically include a low-dose ASA. In such patients, a 10-day course of acetaminophen is a reasonable first step for control of joint pain. While an improvement of quality of life may be the most important criterion of benefit for the patient, an increase in activity level in those rendered sedentary by their joint pain should be an outcome of interest to physicians due to expected health benefits. In patients not controlled on acetaminophen, the goal is to provide an NSAID in the lowest dose and for the shortest duration needed to control pain and increase activities of daily living. In patients taking a low-dose ASA, the GI risks imposed by NSAIDs are increased.¹⁰ This includes the risk of GI symptoms as well as GI bleeds. A coxib may be approp
lling symptoms that might impair compliance with the NSAID and in reducing the risk of GI bleeds. There is no evidence that the currently available coxib increases CV risk relative to non-selective NSAIDs. When prescribed to appropriate patients, the coxib may provide a more attractive benefit-to-risk ratio when both GI symptoms and CV risk are considered (Table 2).

Table 2.

<img3459|center>

The retrospective evidence that celecoxib offers comparable CV safety to non-selective NSAIDs is now being definitively tested in a large prospective study called PRECISION, which is expected to be completed in 2011. It is randomizing approximately 20,000 patients to celecoxib, ibuprofen, or naproxen with CV events as a primary end point. While this trial has the potential to resolve the current questions regarding relative CV risk of these three agents from a population perspective, it will not eliminate the importance of individualized therapy. In a patient population with multiple medical issues, it is particularly important to consider the relative risks of treatment in the context of all of the concomitant conditions. Moreover, it is appropriate to remain sensitive to the substantial variability among patients in their therapeutic response and ability to tolerate any given treatment. An empiric trial-and-error approach may be required to identify a therapy that offers an acceptable benefit-to-risk ratio.

The withdrawal of several widely prescribed coxibs from the pharmaceutical market may have been an inevitable response to the safety concerns generated by the clinical trials with those agents, but it is important to recognize that arthritis patients at a high GI risk may have no alternative to a coxib that offers comparable safety. Celecoxib, the available coxib, lowers GI risk while offering a level of CV safety similar to non-selective NSAIDs on the lowest end of the spectrum of associated CV risk. These relative risks are important to understand in the context of each patient for whom therapeutic strategies must be individualized.

Summary

The available data do not substantiate an association between COX specificity and increased CV risk. In the retrospective data from which the relative CV risk of the coxib and non-selective subclasses of NSAIDs has been drawn, agents within each class have been associated with elevated CV risk but with considerable variability within these classes. Although an ongoing prospective trial will provide definitive data, the retrospective evidence indicates that celecoxib, the only COX-2-specific NSAID currently available in Canada, is at least as safe from a CV perspective as non-selective NSAIDs. In many elderly patients, particularly those with an elevated GI risk, a coxib may be the most reasonable choice for pain control.

References

1. FitzGerald GA, Patrono C. The coxibs, selective inhibitors of cyclooxygenase-2. N Engl J Med 2001;345:433-42.

2. Topol EJ. Failing the public health—rofecoxib, Merck, and the FDA. N Engl J Med 2004;351:1707-9.

3. Nussmeier et al. Complications of the COX-2 inhibitors parecoxib and valdecoxib after cardiac surgery. N Engl J Med 2005;352:1081-91.

4. 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.

5. Rahme E, Nedjar H. Risks and benefits of COX-2 inhibitors vs. non-selective NSAIDS: does their cardiovascular risk exceed their gastrointestinal benefits? A retrospective cohort study. Rheumatology 2007;46:435-8.

6. Whelton et al. Effect of celecoxib and rofecoxib on blood pressure and edema in patients =65 years of age with systemic hypertension and osteoarthritis. Am J Cardiol 2002;90:959-63.

7. Whelton et al. Cardiorenal effects of celecoxib as compared with the non-steroidal anti-inflammatory drugs diclofenac and ibuprofen. Kidney Int 2006;70:1495-502.

8. Joshi GP, Gertler R, Fricker R. Cardiovascular thromboembolic adverse effects associated with cyclooxygenase selective inhibitors and nonselective antiinflammatory drugs. Anesth Analg 2007;105:1793-804.

9. Dajani EZ, Islam K. Cardiovascular and gastrointestinal toxicity of selective cyclooxygenase-2 inhibitors in man. J Physiol Pharmacol 2008;59(suppl 2):117-33.

10. Laine L. Gastrointestinal bleeding with low-dose ASA, what’s the risk? Aliment Pharmacol Ther 2006;24:897-908.

MAKING PATIENTS AWARE OF THE BENEFIT TO RISK RATIO OF NSAIDs

Editorial Review:

Andrew E. Thompson, MD, FRCPC

Director of Post-graduate Education, Assistant Professor of Medicine, Division of Rheumatology, Schulich School of Medicine, University of Western Ontario, London, Ontario

In medicine, such terms as “safe” or “effective” must always be qualified because there are few courses of action without risk or for which a benefit can be guaranteed. The effort to communicate relative risk to patients who are evaluating one option over others is an important clinical skill not typically included in formal medical training. Risks and benefits need to be presented in a sensible order and a useful context, differentiating common from rare events. The exercise of weighing the relative risks and benefits of cyclooxygenase-2 (COX-2)- selective and non-selective non-steroidal anti-inflammatory drugs (NSAIDs) illustrates the issues of communication and the importance of individualizing risk assessment. The benefits and the risks of these agents for controlling symptoms of osteoarthritis (OA) are always relative and not fully predictable. In guiding decisions, it is important to recognize the level at which the patient grasps relative risk relationships and wishes to exercise control of a clinical choice.

Product monographs for pharmaceuticals provide an exhaustive list of potential risks based on potential associations from clinical trials. Due to a lack of context and insufficient distinctions between potential and likely associations, these lists are nearly uninterpretable for patients as well as physicians. In counselling patients, the most difficult hurdle is making the concept of relative risk useful from a practical perspective. When comparing one course of action to another, the higher or lower risk from one may be counterbalanced by increased efficacy. Moreover, even when the absolute risk from one option is greater than the other, the risk may be very low for both, making more convenient dosing or other minor advantages relevant.

Relative risks are typically derived from controlled trials, but trials may or may not be reflective of routine practice. Restrictions imposed by trial design may affect the pool of patients evaluated and participation in a clinical trial may generate different behaviour than typical of routine practice such as closer adherence when patients are aware they are being monitored. In addition, even relatively large clinical trials will not expose a sufficient number of patients to detect rare events. In general, phase IV post-marketing data may provide a more accurate reflection of realworld experience, but phase IV data are not formally or robustly collected for all agents and these data are not necessarily included in product monographs or otherwise readily accessible.

Considerations in Individualizing Risks

Due at least in part to the large market for NSAIDs, both the efficacy and the safety of these agents have been closely scrutinized. When data with non-selective NSAIDs or selective COX-2 inhibitors (coxibs) are combined, more than 145,000 patients have participated in controlled trials.¹ In observational studies conducted to assess risks for cardiovascular (CV), gastrointestinal (GI), or other outcomes, the data may extend to several million patients.

Although this provides an uncommonly deep pool of safety data, the relative risks derived from these data must still be individualized to patients who may or may not closely fit any of the subgroups for which there are specific safety and efficacy information.

While the consideration of how to individualize GI and CV risks has been evaluated in detail by others, the task of communicating risk involves principles of patient care that are based on practical sense and experience. Clearly, the first step is to provide patients with a positive orientation about the goals of therapy and the potential to resolve problems. Even when there is legitimate concern about potential risks, a negative emphasis at the outset can be counterproductive for a patient seeking a resolution of health problems. Although it is important to provide a fair and thorough assessment of potential risks, it is also helpful to end the discussion with a positive message for the same purpose of encouraging confidence in the course of treatment.

Effective communication of risks requires a layering approach in which physicians can help patients differentiate the risk of common from uncommon adverse events. Context is important. Many risks from medicine are far lower than risks patients face on a daily basis when driving their car or stepping out of the bathtub (Figure 1). Patients need to understand that the risk is being considered in the context of a much greater likelihood of benefit. One reasonable approach is to lis
st common complaints brought by patients derived from published data as well as from the clinician’s own experience. Personal anecdotes about the frequency with which a physician has heard a given complaint provide a useful context that may allow patients to relate more personally to the information they are hearing, increasing their vigilance for signs of the specific events and reducing alarm if they occur.

Figure 1.

<img3460|center>

Less common events should be listed next, again with an orientation that permits patients to distinguish infrequent from rare. For example, a clinician might provide some context from his or her own experience regarding the number of times that the specific event has been encountered. However, patients should not be given the impression that uncommon side effects, even those never encountered by the physician providing the counselling, are sufficiently unlikely that the risk can be dismissed. Most physicians have had experiences with very rare side effects. Recounting an anecdote about such an experience may be a useful tool for allowing patients to remain alert for symptoms even if they are reassured that their absolute risk is low.

Conveying Accurate and Thorough Information on Risks and Benefits

Despite the importance of a general orientation toward reassurance, honesty is critical. The failure to produce counselling that is accurate and thorough may render irreparable damage to a patient relationship if the information is suspected of being incomplete. Most patients can accept even significant risks to gain significant benefits if they are comfortable that they have been fully informed. To some degree, the ability to convey information that the patient recognizes as honest and complete is a learned art. Body language, eye contact and tone of voice are not unimportant in communication and should not be overlooked.

In addition to a stepwise approach to a discussion of risks by the frequency with which they occur, it may also be helpful to take a stepwise approach to the level of detail, providing a simple and basic review of relative risks and benefits to all patients and then moving to more technical information as suitable to the individual patient. The sophistication of the discussion may depend less on education level than on level of interest. While some patients seek a maximum amount of information in order to more fully participate in decisions about their care, a substantial proportion of patients are not comfortable with a decision-making role and seek information about risks primarily so that adverse events can be readily identified when they occur. These important differences should guide goals of patient education.

Putting Risk in Proper Perspective

There is a well documented phenomenon for individuals to overestimate rare risks and Table 1. Additional GI Bleeding Events and Myocardial Infarction Associated with Using NSAID and Coxib Use The calculations used a mortality rate of 10% for GI bleeding and 30% for CV events. Adapted from Moore et al. Arthritis Res Ther 2008;10(1):R20. underestimate common risks.² In counselling patients about drug therapies, the concept of an increased risk is often provided in the absence of a denominator. For example, the background rate for upper GI bleeding in the absence of NSAIDs is an estimated 2.2 per 1,000 patient-years on therapy.³ Although the relative risk is quadrupled on nons
a substantial range among NSAIDs), this adds only 5 events per 1,000 patient-years, so the absolute event rate remains low.4 In high-risk patients, such as those with a previous GI bleed or who are taking an anticoagulant, the increase, however, can be many times greater and highly clinically relevant, necessitating either a selective NSAID, such as celecoxib, which is not associated with a meaningful increase in GI bleeds, or NSAID avoidance (Table 1).

Table 1.

<img3461|center>

Similar data are generated by an analysis of CV risk, which is increased 10% overall on NSAIDs (again with a substantial range), but with a low expected clinical impact outside of high-risk groups.¹ In many patients, the importance of NSAIDs for CV risk is very small relative to other modifiable factors, such as smoking or hypercholesterolemia, that are ignored by patients even when informed of the health benefits of lifestyle changes. By hazard ratio (HR), the range for CV risk includes a slight non-significant reduction in risk for celecoxib (HR 0.96; 95% CI, 0.90 – 1.02) and naproxen (HR 0.98; 95% CI, 0.92 – 1.05) to more than a 40% increase on diclofenac (HR 1.44; 95% CI, 1.32 – 1.56).¹

Patient perception of risk is not irrelevant. Although objective information is useful for an informed decision, patients who fear specific events, such as a CV event or a GI bleed, out of proportion with their absolute risk should be permitted to develop their own sense of relative benefit-to-risk. Unlike therapies that have the potential to alter the natural history of disease, the primary goals of an NSAID prescription for OA are to relieve symptoms and improve quality of life. In patients not convinced of a net benefit relative to their own safety concerns, therapy can be stopped. However, for the large number of OA patients who do require NSAIDs to achieve adequate daily function, accurate risk counselling is essential so that there is a clear understanding of the relative gains to the relative risks. Although risks of therapy can be modified but not eliminated by selecting an appropriate treatment strategy, for most patients, the relative improvement in quality of life on therapy will correlate with the degree of acceptable relative risk.

Summary

Accurate data are essential but not sufficient for providing patients with meaningful information about the relative benefits and risks of treatments. Risk data require a context within which the patient can organize common and uncommon events. Physicians can greatly improve their ability to describe the risks and benefits of NSAIDs by following simple principles of communication, which include an organized presentation and disclosing detail in quantities appropriate to patient interest. It is useful to provide a thorough and forthright description of risks within a framework that emphasizes the goal of resolving health issues.

References

1. Hernández-Díaz S, Varas-Lorenzo C, García Rodríguez LA. Non-steroidal anti-inflammatory drugs and risk of acute myocardial infarction. Basic Clin Pharmacol Toxicol 2006;98:266-74.

2. Hakes JK, Viscusi WK. Dead reckoning: demographic determinants of the accuracy of mortality risk perceptions. Risk Anal 2004;24:651-64.

3. Mamdani et al. Observational study of upper gastrointestinal haemorrhage in elderly patients given selective cyclo-oxygenase-2 inhibitors or conventional non-steroidal anti-inflammatory drugs. BMJ 2002;325:624-9.

4. Moore et al. What do we know about communicating risk? A brief review and suggestion for contextualizing serious but rare risk and the example of cox-2 selective and non-selective NSAIDs. Arthritis Res Ther 2008;10:R20.