Patients' welfare should be primary concern

EDITOR—Chard and Dieppe discussed the use of glucosamine in treating osteoarthritis.1 Osteoarthritis is traditionally treated with non-steroidal anti-inflammatory drugs and analgesics, just masking symptoms but doing nothing to improve the disease. Worst of all, non-steroidal anti-inflammatory drugs alone cause over 16 500 deaths and over 103 000 admissions to hospital each year in the United States.2

With regard to evidence based medicine, the authors of the Cochrane review on non-steroidal anti-inflammatory drugs for osteoarthritis of the knee conclude that, despite the large number of publications in this area, there are few randomised controlled trials.3 Furthermore, most trials comparing two or more such drugs have substantial design errors. The reviewers conclude that no substantial evidence is available related to efficacy, to distinguish between equivalent recommended doses of non-steroidal anti-inflammatory drugs. Had studies used appropriate doses of the drug that was being compared, most would have been sufficiently powerful to detect clinically important differences in efficacy. As differences in efficacy between non-steroidal anti-inflammatory drugs have not been recorded, the prescription of such a drug for osteoarthritis of the knee should be based on relative safety, acceptability to the patient, and cost.

A Cochrane review on glucosamine gave the following main results.4 Collectively, the reviewers identified 16 randomised controlled trials that provided evidence that glucosamine is both effective and safe in patients with osteoarthritis. In the 13 randomised controlled trials in which glucosamine was compared with placebo, glucosamine was found to be superior in all but one. In the four randomised controlled trials in which glucosamine was compared with a non-steroidal anti-inflammatory drug, glucosamine was superior in two and equivalent in two.

Additionally, new evidence was reported in a clinical trial published in January 2001, which showed that glucosamine sulphate can beneficially modify the disease process of osteoarthritis, from the point of view of both symptoms and structure.5 These results have already been confirmed in an independent trial, presented in November 2000 at the 64th annual meeting of the American College of Rheumatology in Philadelphia.

Glucosamine is currently the only substance known that can both help with the symptoms of osteoarthritis and improve the course of the disease, without showing any harmful effect for the patient. Despite its supplement status in certain countries, the evidence tells us that glucosamine must be taken seriously.

Sound science might have helped avoid confusion

EDITOR—In their editorial Chard and Dieppe are right to conclude that there is more confusion and hype than magic about glucosamine.1 The confusion is largely the result of poor knowledge of the chemical, pharmacokinetic, and therapeutic properties of glucosamine. The editorial itself is biased by several examples of this confusion.

Firstly, glucosamine is not a sulphated (ester with sulphuric acid) amino-monosaccharide, as defined in the editorial. There is confusion with the sulphated disaccharide units of some glycosaminoglycans such as chondroitin sulphate.

Secondly, whether these large molecules can reach chondrocytes in vivo by the oral route is unclear. For glucosamine the preferential uptake by articular cartilage after oral administration is proved. 2 3

Thirdly, their statement is strange that osteoarthritic cartilage should not be cured because it has no nerve endings and is therefore not the origin of pain. Have Chard and Dieppe never heard about cytokines released from damaged tissues that produce pain and inflammation?

Fourthly, the statement that the most appropriate dose and route of administration is still unknown is outdated. The efficacy of glucosamine sulphate in daily oral doses of 1500 mg on the symptoms of osteoarthritis is documented by several clinical trials and recently by two three year studies, in which a structure modifying effect was also shown. 4 5

Fifthly, Chard and Dieppe do not know how to classify glucosamine—is it a drug, a food supplement, a nutriceutical, or a complementary treatment? Glucosamine is classified in one or another of the mentioned categories in different countries according to the local regulations. All pivotal clinical trials were made with glucosamine sulphate regulated as a drug. In several countries, including the United Kingdom and the United States, glucosamine as its sulphate or hydrochloride is freely available also as a food supplement. As such it is not subject to control for quality and for promotional claims, especially the kind found in the lay press.

As scientists, Chard and Dieppe should have been able to discern the therapeutic merits of glucosamine on the basis of scientific evidence from the hype of unregulated non-medical fantastic promotion. An editorial based on sound and unbiased chemical, pharmacokinetic, and clinical knowledge would have surely contributed to dispelling much of the confusion about glucosamine. Unfortunately this was not the case.

Authors' reply

EDITOR—We are pleased to respond to two of the 16 rapid responses posted on bmj.com in response to our editorial on glucosamine as their authors make some valid points.1

Sonnino compares the Cochrane reviews on how harmful non-steroidal anti-inflammatory drugs can be and how effective glucosamine is. This may be true, but it does not relate to the main thrust of our editorial. We have never denied that the published evidence suggests that glucosamine works, although we (like the Cochrane reviewers) did point out the need for larger studies. Our main point is that commercial biases and publication biases, for example, might affect the validity of these data. These are not issues covered by Cochrane reviews, which only report on the quality of published evidence. Finally, we, like Sonnino, reported that a recently published trial indicated that glucosamine might have an effect in modifying the structure, although we urge caution in the interpretation of these data as there are unresolved methodological issues surrounding the use of radiographic measures of joint space narrowing.2

Setnikar raises five issues.

Firstly, glucosamine is not sulphated. This is correct, and we apologise for adding to the confusion about this.

Secondly, oral doses do reach the chondrocytes—again we accept that there is some evidence to support this view.

Thirdly, he takes issue with our suggestion that cartilage might play little part in pain generation in osteoarthritis. But we maintain that, although this might theoretically be possible, there is no direct evidence for cartilage involvement; what little evidence there is on the source of pain in osteoarthritis would favour bone.3

Fourthly, Setnikar maintains that 1500 mg daily is the established optimum dose. We agree that this is the standard dose mentioned in the literature, but we were unable to identify any dose-response studies published, in English, on which this figure was based, and trials by Hughes et al and Houpt et al show that glucosamine is not effective at 1500 mg daily. 4 5

Finally, he raises an important point about avoiding media hype in medicine and basing recommendations on valid scientific evidence. We agree—in our editorial we included only trials and reviews, ignoring the multitude of comments that promote the use of glucosamine. We also highlighted the fact that the number of comments, reviews, letters, and editorials will soon overtake the amount of primary evidence.