Glucosamine and Chondroitin Sulfate for Knee Osteoarthritis
http://www.100md.com
《新英格兰医药杂志》
To the Editor: The study by Clegg et al. (Feb. 23 issue)1 suggests that glucosamine and chondroitin are not better than placebo for knee osteoarthritis. The most remarkable aspect of their study is perhaps the huge placebo effect — 60 percent of the patients noted at least a 20 percent decrease in the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) pain score from baseline to week 24.1 The authors comment that "elevated rates of response to placebo have been reported in other osteoarthritis trials." However, the two references provided did not show long-term placebo effects after oral medication — one used intraarticular injections2 and the other had only 13 weeks of follow-up.3 The extraordinary response to oral placebo in the trial by Clegg et al. might explain its findings. That is, if placebo is effective in 60 percent of patients, it is difficult for other treatments to surpass this mark. The use of rescue medication seems to have increased during the trial period in all groups. No data were provided on other concomitant treatments, such as physiotherapy. Was the improvement in the placebo group the result of treatment or a true placebo response?
Edzard Ernst, M.D., Ph.D.
Peninsula Medical School
Exeter EX2 4NT, United Kingdom
edzard.ernst@pms.ac.uk
References
Clegg DO, Reda DJ, Harris CL, et al. Glucosamine, chondroitin sulfate, and the two in combination for painful knee osteoarthritis. N Engl J Med 2006;354:795-808.
Altman RD, Moskowitz R. Intraarticular sodium hyaluronate (Hyalgan) in the treatment of patients with osteoarthritis of the knee: a randomized clinical trial. J Rheumatol 1998;25:2203-2212.
Lehmann R, Brzosko M, Kopsa P, et al. Efficacy and tolerability of lumiracoxib 100 mg once daily in knee osteoarthritis: a 13-week, randomized, double-blind study vs. placebo and celecoxib. Curr Med Res Opin 2005;21:517-526.
To the Editor: Patients in pain eagerly seek analgesia to relieve the pain, enabling them to continue having an adequate quality of life. In the study by Clegg and colleagues, patients with knee osteoarthritis were randomly assigned to receive glucosamine, chondroitin sulfate, the combination of the two, celecoxib, or placebo and were allowed to have "rescue" acetaminophen (up to 4000 mg daily) whenever they wanted more pain relief. Irrespective of the pain scores and the complex analyses by the authors, it is evident that patients in all categories used similar amounts of "rescue" acetaminophen (between 800 mg and 950 mg of acetaminophen daily in all five groups of the study). This similarity provides a good surrogate for the efficacy of medication (i.e., no medication was better than placebo). I would be interested to know how the authors could explain this. In addition, I was surprised to see that no patients were randomly assigned to receive regular acetaminophen and a nonsteroidal antiinflammatory drug as rescue medication, which is by far the safest strategy — and a very effective one — favored by many clinicians.
Vassilios S. Vassiliou, M.B., B.S.
University of Cambridge
Cambridge CB1 8HD, United Kingdom
vassiliou@doctors.org.uk
To the Editor: Clegg et al. suggest that glucosamine or chondroitin alone did not reduce pain effectively and that the combination of the two may be effective in the subgroup of patients with moderate-to-severe pain. It is disturbing to see that celecoxib is not effective in this subgroup of patients, which is the major subgroup of patients in our daily practice. It is even more disturbing that in patients with mild pain, celecoxib was marginally effective (P=0.04) in reducing the primary outcome and did not improve the secondary outcomes. These results contrast with those of other studies. In that context, concerns may also be extrapolated to glucosamine and chondroitin. Are these results linked to the high response rate in the placebo group? Since only a small portion of the patients (22 percent) had moderate-to-severe pain, we feel one should be very cautious about drawing any firm conclusions.
Jean-Pierre Pelletier, M.D.
University of Montreal Arthritis Centre
Montreal, QC H2L 4M1, Canada
dr@jppelletier.ca
Dr. Pelletier reports having received consulting fees and a research grant from Pfizer.
The authors reply: We agree with Ernst that the elevated placebo response may have dampened the ability to differentiate among the treatments. The study was designed to detect clinically meaningful treatment effects as compared with a placebo response rate of up to 65 percent. The cause of the high placebo response rate is not clear and is probably multifactorial. Placebo responses are often larger in studies of alternative therapies.1 This may, in part, be due to elevated expectations of patients. In addition, in the Glucosamine/chondroitin Arthritis Intervention Trial (GAIT), we recruited patients with relatively mild symptoms for which the outcome measures may be less sensitive. Traditionally, studies of osteoarthritis pain have predefined "flare" criteria that must be met in order to enter the trial. In such studies, the baseline pain level is higher than that in GAIT, and this may be associated with a diminished placebo response.
The observation that the use of rescue medication increased in all groups between baseline and the end of the study is also interesting. The mean increases were relatively small (approximately 200 to 300 mg of acetaminophen per day) and the standard deviations were large. Baseline use of rescue medication was assessed at a time when some patients were still in the washout period for other analgesics. GAIT controlled for effects of other treatments (physical therapy, muscle-conditioning programs, or other interventions for osteoarthritis) by excluding patients who would not agree to forgo any given intervention or maintain it at a stable level throughout the trial.
We agree with Vassiliou that the amounts of rescue analgesia follow changes in the primary and the other secondary outcome measures and demonstrate little change among the groups. Interestingly, in the subgroup with moderate-to-severe pain, the combination and celecoxib groups used less rescue analgesia and had more improvement in pain.
With regard to the comments of Pelletier, although pain relief with celecoxib was not found to be statistically significant in the subgroup with moderate-to-severe pain, it met the definition of clinical significance (a 15 percent improvement as compared with placebo). The borderline statistical significance is a result of reduced statistical power in this relatively small subgroup. We concur that the outcome measures did not behave as rigorously as we had expected; however, chondroitin sulfate reduced joint swelling with little suggestion of pain relief. Glucosamine, on the other hand, was nearly as effective as celecoxib in terms of pain relief in the subgroup with moderate-to-severe pain, suggesting the possibility of different biologic effects among the supplements.
Daniel O. Clegg, M.D.
University of Utah
Salt Lake City, UT 84132
gait.study@hsc.utah.edu
Domenic J. Reda, Ph.D.
Department of Veterans Affairs Cooperative Studies
Coordinating Center
Hines, IL 60141
References
Kaptchuk TJ. The placebo effect in alternative medicine: can the performance of a healing ritual have clinical significance? Ann Intern Med 2002;136:817-825.
Edzard Ernst, M.D., Ph.D.
Peninsula Medical School
Exeter EX2 4NT, United Kingdom
edzard.ernst@pms.ac.uk
References
Clegg DO, Reda DJ, Harris CL, et al. Glucosamine, chondroitin sulfate, and the two in combination for painful knee osteoarthritis. N Engl J Med 2006;354:795-808.
Altman RD, Moskowitz R. Intraarticular sodium hyaluronate (Hyalgan) in the treatment of patients with osteoarthritis of the knee: a randomized clinical trial. J Rheumatol 1998;25:2203-2212.
Lehmann R, Brzosko M, Kopsa P, et al. Efficacy and tolerability of lumiracoxib 100 mg once daily in knee osteoarthritis: a 13-week, randomized, double-blind study vs. placebo and celecoxib. Curr Med Res Opin 2005;21:517-526.
To the Editor: Patients in pain eagerly seek analgesia to relieve the pain, enabling them to continue having an adequate quality of life. In the study by Clegg and colleagues, patients with knee osteoarthritis were randomly assigned to receive glucosamine, chondroitin sulfate, the combination of the two, celecoxib, or placebo and were allowed to have "rescue" acetaminophen (up to 4000 mg daily) whenever they wanted more pain relief. Irrespective of the pain scores and the complex analyses by the authors, it is evident that patients in all categories used similar amounts of "rescue" acetaminophen (between 800 mg and 950 mg of acetaminophen daily in all five groups of the study). This similarity provides a good surrogate for the efficacy of medication (i.e., no medication was better than placebo). I would be interested to know how the authors could explain this. In addition, I was surprised to see that no patients were randomly assigned to receive regular acetaminophen and a nonsteroidal antiinflammatory drug as rescue medication, which is by far the safest strategy — and a very effective one — favored by many clinicians.
Vassilios S. Vassiliou, M.B., B.S.
University of Cambridge
Cambridge CB1 8HD, United Kingdom
vassiliou@doctors.org.uk
To the Editor: Clegg et al. suggest that glucosamine or chondroitin alone did not reduce pain effectively and that the combination of the two may be effective in the subgroup of patients with moderate-to-severe pain. It is disturbing to see that celecoxib is not effective in this subgroup of patients, which is the major subgroup of patients in our daily practice. It is even more disturbing that in patients with mild pain, celecoxib was marginally effective (P=0.04) in reducing the primary outcome and did not improve the secondary outcomes. These results contrast with those of other studies. In that context, concerns may also be extrapolated to glucosamine and chondroitin. Are these results linked to the high response rate in the placebo group? Since only a small portion of the patients (22 percent) had moderate-to-severe pain, we feel one should be very cautious about drawing any firm conclusions.
Jean-Pierre Pelletier, M.D.
University of Montreal Arthritis Centre
Montreal, QC H2L 4M1, Canada
dr@jppelletier.ca
Dr. Pelletier reports having received consulting fees and a research grant from Pfizer.
The authors reply: We agree with Ernst that the elevated placebo response may have dampened the ability to differentiate among the treatments. The study was designed to detect clinically meaningful treatment effects as compared with a placebo response rate of up to 65 percent. The cause of the high placebo response rate is not clear and is probably multifactorial. Placebo responses are often larger in studies of alternative therapies.1 This may, in part, be due to elevated expectations of patients. In addition, in the Glucosamine/chondroitin Arthritis Intervention Trial (GAIT), we recruited patients with relatively mild symptoms for which the outcome measures may be less sensitive. Traditionally, studies of osteoarthritis pain have predefined "flare" criteria that must be met in order to enter the trial. In such studies, the baseline pain level is higher than that in GAIT, and this may be associated with a diminished placebo response.
The observation that the use of rescue medication increased in all groups between baseline and the end of the study is also interesting. The mean increases were relatively small (approximately 200 to 300 mg of acetaminophen per day) and the standard deviations were large. Baseline use of rescue medication was assessed at a time when some patients were still in the washout period for other analgesics. GAIT controlled for effects of other treatments (physical therapy, muscle-conditioning programs, or other interventions for osteoarthritis) by excluding patients who would not agree to forgo any given intervention or maintain it at a stable level throughout the trial.
We agree with Vassiliou that the amounts of rescue analgesia follow changes in the primary and the other secondary outcome measures and demonstrate little change among the groups. Interestingly, in the subgroup with moderate-to-severe pain, the combination and celecoxib groups used less rescue analgesia and had more improvement in pain.
With regard to the comments of Pelletier, although pain relief with celecoxib was not found to be statistically significant in the subgroup with moderate-to-severe pain, it met the definition of clinical significance (a 15 percent improvement as compared with placebo). The borderline statistical significance is a result of reduced statistical power in this relatively small subgroup. We concur that the outcome measures did not behave as rigorously as we had expected; however, chondroitin sulfate reduced joint swelling with little suggestion of pain relief. Glucosamine, on the other hand, was nearly as effective as celecoxib in terms of pain relief in the subgroup with moderate-to-severe pain, suggesting the possibility of different biologic effects among the supplements.
Daniel O. Clegg, M.D.
University of Utah
Salt Lake City, UT 84132
gait.study@hsc.utah.edu
Domenic J. Reda, Ph.D.
Department of Veterans Affairs Cooperative Studies
Coordinating Center
Hines, IL 60141
References
Kaptchuk TJ. The placebo effect in alternative medicine: can the performance of a healing ritual have clinical significance? Ann Intern Med 2002;136:817-825.