Glucosamine sulphate and osteoarthritis

Sir—Jutta Halbekath and colleagues are biased by gross misunderstandings and misinterpretations. The poor correlation between symptom and structure outcomes in osteoarthritis is well accepted. Such poor correlation may apply also to the effects of diseasemodifying drugs.

However, our secondary analysis, which we briefly present in the paper and as an abstract elsewhere,

shows that glucosamine sulphate provided symptom relief independently of structure outcomes. Thus, Halbekath and colleagues’ assumption is wrong.

The final difference in minimum joint-space narrowing was not lower than, but identical to that expected

in our per-protocol sample-size calculation. The significant symptom improvement with glucosamine sulphate, the reduced proportion of patients with severe structure deterioration, and the antagonism of the natural joint-space narrowing make our findings clinically relevant. The European Drug Agency states that given the favourable modification of the natural history of osteoarthritis by structure changes, a large proportion of patients will have long-term clinical benefit.

The symptom-modifying effects of glucosamine sulphate, even for shorter treatment periods, is well described in the reports we referenced. Nevertheless, we can confirm that, although this effect was beyond the goals of our study, the symptoms that were affected developed early. The effect was significant compared with placebo throughout treatment. Furthermore, contrary to the groundless statement by Halbekath and co-workers, there were no baseline imbalances between groups, since the small apparent numerical differences in the WOMAC index were far from significant, whereas the changes at the end of treatment were significant. In addition, there are no discrepancies between table 3 and figure 2: the table shows the primary endpoint represented by the mean percentage variation. This representation makes clinical interpretation easier than the mean change in absolute index scores reported for the subscales in the figure. Elementary arithmetic teaches that these two ways of expressing the same results do not give superimposable numbers.

The study calendar dates are from May, 1994 (first patient in), to December, 1998 (last patient out).

We agree with R B Payne on the importance of the integrated approach to the medical management of osteoarthritis, including pharmacological and non-pharmacological measures. Unfortunately, we could not assess the possible additive effects of glucosamine sulphate and exercise in our trial.

We thank Layinka Swinburne for pointing out an effect that was not noted in our study, nor ever in the clinical development of glucosamine sulphate to our knowledge.

Mark Goldstein’s hypothesis is based on speculation and not experimental or clinical evidence. Glucosamine sulphate was well tolerated in our trial and the attribution of the lack of an adverse event (namely episodes of decreased blood pressure) to a possible problem, is misleading. However, four patients receiving placebo and two taking glucosamine sulphate had cardiovascular events (myocardial infarction, ischaemia, or angina). Our efficacy and safety results are applicable only to crystalline glucosamine sulphate, which is a prescription drug in more than 40 countries. the periodic safety update reports for 1995–2000 estimate a total of 17 042 492 patient-months of treatment, with no reports of atherosclerosis or cardiovascular events.