sign in sign up
<<
Home , Glucosamine

View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector

Osteoarthritis and (2002) 10, 836–837 © 2002 Research Society International. Published by Elsevier Science Ltd. All rights reserved. 1063–4584/02/$35.00/0 doi:10.1053/joca.2002.0841, available online at http://www.idealibrary.com on

International Cartilage Repair Society

Commentary

Published meta-analyses of pharmacological therapies for osteoarthritis T. E. Towheed Department of Medicine, Queen’s University, Kingston, Ontario, Canada

The objectives of this commentary are twofold: (1) to efficacy, to distinguish between equivalent recommended identify which pharmacological agents used in the manage- doses of NSAIDs ment of osteoarthritis (OA) have published evidence in the Eccles et al.6 analysed three RCTs in an attempt to rank form of systematic reviews (SR) and meta-analyses (MA) the relative efficacy of acetaminophen and NSAIDs in the which supports their efficacy and/or toxicity, and (2) to treatment of OA. Their main finding was that NSAIDs summarize the main findings from these studies. were slightly superior to acetaminophen in the outcomes of An MA can be defined as an SR that employs statistical at rest (effect size 0.35, 95% confidence interval, methods to combine and summarize the results of several 0.17–0.53) and pain on motion (effect size 0.28, 95% studies. In evidence-based medicine, MA of randomized confidence interval, 0.08–0.48). However, there was no controlled trials (RCTs) may well constitute the strongest significant difference in the time to walk 50 feet, or in the evidence for the value of an intervention. For example, a quality of life improvements produced by the two therapies. commonly used grading scheme for evaluating the quality Moore et al.7 published an MA evaluating the efficacy of the evidence is one adapted from the US Agency for and safety of topical NSAIDs vs placebo. Twenty-five RCTs Health Care Policy and Research. Grades 1A and 1B evaluating subjects with ‘chronic pain conditions’ (including evidence are the most robust and least biased sources of OA) were analysed covering the time period of 1966 to evidence1. Grade 1A evidence is evidence obtained from September 1996. The main findings were: (1) 8/13 RCTs MA of RCTs. Grade 1B evidence is evidence obtained from found that topical NSAIDs were superior to placebo with a at least one RCT. There have been a total of 10 published pooled relative risk (RR) for benefit of 2.0 (95% confidence MA evaluating the following therapies for OA: conventional interval, 1.5–2.7). The number needed to treat (NNT) was NSAIDs, simple analgesics (e.g. acetaminophen), COX-2 only 3.1 (95% confidence interval, 2.7–3.8), (2) two RCTs selective NSAIDs, topical NSAIDs, topical capsaicin, compared topical NSAIDs to oral NSAIDs and found and glucosamine2–12. equal efficacy, and (3) the safety of topical NSAIDs was The meta-analyses of pharmacological therapies for OA equivalent to placebo. of the hip analysed 43 RCTs covering the time period of Zhang et al.8 published an MA evaluating the efficacy 1966 to August 19942,3. Main findings from these reviews and safety of topical capsaicin in a number of chronic included the following: (1) the quality of the RCTs was painful conditions, including OA. Three RCTs were generally poor, (2) there was a lack of standardization of analysed covering the time period of 1980 to February case definition of OA and also a lack of standardization of 1994. The main findings were: (1) capsaicin cream was the outcome assessments, (3) NSAIDs were always better better than placebo in providing pain relief in OA (odds ratio than placebo, (4) it was rare to find differences in efficacy of 4.36, 95% confidence interval of 2.8–6.9), (2) true between NSAIDs, and (5) there was a tendency for blinding was probably difficult to conduct with capsaicin. indomethacin to be found more effective than comparator The UK’s National Institute of Clinical Excellence (NICE) NSAIDs, but also more toxic. Low dosages of published on their website an MA which evaluated the (<750 mg/day) and ibuprofen (<1600 mg/day) were less efficacy and safety of the four available COX-2 selective efficacious than other NSAIDs. NSAIDs: , , , and etodolac9. The meta-analyses of pharmacological therapies for OA Fifty-three RCTs involving 61 731 patients with OA and RA of the knee involved two publications4,5. The first analysed were analysed. The main findings of the NICE review were: 80 RCTs covering the time period of 1966 to August 19944. (1) COX-2 selective inhibitors have equivalent efficacy to The second analysed 16 RCTs covering the time period of conventional NSAIDs, (2) COX-2 selective inhibitors are 1966 to 19965. The main findings from these reviews effective in reducing the incidence of gastrointestinal (GI) included the following: (1) evidence exists for the efficacy toxicity compared to conventional NSAIDs, (3) there is no of acetaminophen, topical capsaicin, intra-articular (IA) direct evidence to suggest that any one of the four COX-2 steroids, IA hyaluronate, and NSAIDs in the treatment of selective drugs is superior to another, (4) cost-effectiveness OA of the knee, (2) the quality of the RCTs was generally of the COX-2 selective inhibitors is more likely to be poor, and (3) no substantial evidence is available related to favorable in patients at high risk for gastrointestinal toxicity, (5) COX-2 selective inhibitors are not recommended for Tel: 613-533-6896; Fax: 613-533-2189; E-mail: routine use in OA, and (6) COX-2 selective inhibitors are [email protected] recommended for use in patients at high risk for GI toxicity.

836 Osteoarthritis and Cartilage, Vol. 10, No. 11 837

Leeb et al.10 published an MA evaluating the efficacy and recommendations for the management of knee safety of chondroitin sulfate (CS) in OA. Seven RCTs osteoarthritis: report of a task force of the Standing published between 1991 and 1998 were analysed. The Committee for International Clinical Studies main findings of this review were: (1) CS was more effec- Including Therapeutic Trials (ESCISIT). Ann Rheum tive than placebo in relieving the pain of OA with an effect Dis 2000;59:936–44. size of 0.9, (2) in terms of pain reduction, 65% of patients 2. Towheed T, Shea B, Wells G, Hochberg M. Analgesia taking CS will benefit more than from taking a placebo, (3) and non-, non-steroidal anti-inflammatory the effect size for improvement in the Lequesne Index was drugs for osteoarthritis of the hip (Cochrane Review). 0.74, and (4) adverse effects were greater for the placebo In: The Cochrane Library, Issue 3. Oxford: Update treated patients vs the CS treated patients. Software 2001. McAlindon et al.11 published an MA evaluating the 3. Towheed TE, Hochberg MC. A systematic review of efficacy of both glucosamine and CS in OA. Fifteen RCTs were analysed covering the time period of 1966 to June randomized controlled trials of pharmacological 1999. The main findings of this review were: (1) there were therapy in osteoarthritis of the hip. J Rheumatol moderate effect sizes for glucosamine (0.44, with a 95% 1997;24:349–57. confidence interval of 0.24–0.64) and large effect sizes for 4. Towheed TE, Hochberg MC. A systematic review of CS (0.78, with a 95% confidence interval of 0.60–0.95), (2) randomized controlled trials of pharmacological quality scores were generally low, with scores ranging from therapy in osteoarthritis of the knee, with an empha- 12% to 55% of the maximum score, with a mean of 35%, sis on trial methodology. Semin Rheum (4) the effects of both compounds are likely to be exagger- 1997;26:755–70. ated given the methodological weaknesses inherent in the 5. Watson MC, Brookes ST, Kirwan JR, Faulkner A. RCTs (for example, lack of intention-to-treat analyses, lack Non-aspirin, non-steroidal anti-inflammatory drugs of allocation concealment and publication bias) and (3) for treating osteoarthritis of the knee (Cochrane despite the methodological weaknesses, some degree of Review). In: The Cochrane Library, Issue 3. Oxford: efficacy does appear probable for both glucosamine and Update Software 2001. CS. 6. Eccles M, Freemantle N, Mason J, for the North of Towheed et al.12 published an MA evaluating the efficacy England Non-Steroidal Anti-Inflammatory Drug and toxicity of glucosamine compounds in OA. Sixteen Guideline Development Group. North of England RCTs were analysed covering the time period of 1966 to evidence based guideline development project: sum- December 1999. The main findings of this review were: (1) mary guideline for non-steroidal anti-inflammatory Glucosamine was superior to placebo in terms of pain drugs versus basic analgesia in treating the pain of reduction (effect size 1.40, with a 95% confidence interval degenerative arthritis. BMJ 1998;317:526–30. of 0.65–2.14), (2) glucosamine was superior to placebo in 7. Moore RA, Tramer MR, Carroll D, Wiffen PJ, McQuay terms of improvements in the Lequesne Index (odds ratio of HJ. Quantitative systematic review of topically 2.04 with a 95% confidence interval of 1.38–3.02), (3) applied non-steroidal anti-inflammatory drugs. BMJ glucosamine was superior to NSAIDs in terms of pain 1998;316:333–8. reduction (effect size 0.86 with a 95% confidence interval of 8. Zhang WY, Po Wan Li A. The effectiveness of topically 0.58–1.14), and (4) glucosamine demonstrated an excel- applied capsaicin. A meta-analysis. Eur J Clin lent safety profile (only 14 of the nearly 1000 subjects Pharmacol 1994;46:517–22. randomized to glucosamine were withdrawn because of 9. National Institute for Clinical Excellence (NICE) Tech- toxicity, and only 61 subjects reported any adverse nology Appraisal Guidance No. 27. Guidance on the reactions). use of cyclo-oxygenase (COX) II selective inhibitors, In summary, SR and MA are useful techniques which celecoxib, rofecoxib, meloxicam and for can be used to efficiently summarize and document the osteoarthritis and rheumatoid arthritis (2001): relative value of pharmacological therapies in OA. This www.nice.org.uk. methodology will continue to be very useful for the future 10. Leeb BF, Schweitzer H, Montag K, Smolen J. A meta- development of evidence-based guidelines for the analysis of chondroitin sulfate in the treatment of management of OA. The Cochrane Musculoskeletal Group osteoarthritis. J Rheumatol 2000;27:205–11. is taking an active leadership role in this regard with the 11. McAlindon TE, LaValley MP, Gulin JP, Felson DT. publication of systematic reviews of OA therapies in the Glucosamine and chondroitin for treatment of Cochrane Library. osteoarthritis. A systematic quality assessment and meta-anaysis. JAMA 2000;283:1469–75. 12. Towheed TE, Anastassiades TP, Shea B, Houpt J, References Welch V, Hochberg MC. Glucosamine therapy for treating osteoarthritis (Cochrane Review). In: The 1. Pendleton A, Arden N, Dougados M, Doherty M, Cochrane Library, Issue 3. Oxford: Update Software Bannwarth B, Bijlsma JWJ, et al. EULAR 2001.