Today I was given a brochure from the Opfermann company by a drug industry representative (1). According to the studies described here, evidence level 1a is supposed to have been demonstrated for DONA 200-S. There should therefore be no objections to prescribing the preparation for cases of osteoarthritis. I would like to ask for your opinion.

Dr A. KIRCHER (Specialist in internal medicine)
D-36129 Gersfeld/Rhön
Conflict of interests: none

The brochure (1) refers to a COCHRANE Review: twelve randomised controlled studies are included in this systematic review of the benefit of glucosamine (DONA 200-S) in treating osteoarthritis, which was completed in 1999 and has not been updated since (2). Of these studies, eight (3-10) have - partly severe - quality deficiencies, even according to the authors of the review. These quality deficiencies are indeed discussed in the full text, but without consequence for the meta-analysis. Moreover, the concerns are not mentioned at all in the abstract*. Only the positive result of the review is referred to there, but no information provided about the problematic underlying data.

The study with the highest quality scores is an unpublished manuscript which, at the time of inclusion in the COCHRANE Review, was said to already have been "submitted for publication". Three and a half years have passed since then without the study being published. As the largest study with 319 patients, it has a major influence on the results. Furthermore, in virtually all evaluations, it shows the greatest effect. As with the majority of the studies evaluated, the Italian glucosamine manufacturer Rotta was also involved in this one (2). We consider it unacceptable to base a systematic review on a study "submitted for publication" several years ago, but still unpublished - and therefore not assessable by anyone. By the time of going to press, we have received no response to our criticism, which we officially expressed as a comment on the Review on the COCHRANE Collaboration's home page.

Of the three published studies of acceptable quality evaluated in the Review, two compare glucosamine with placebo for osteoarthritis of the knee. One comes out with a positive result (11), while the second does not reveal any effect (12). The third study is a comparison with ibuprofen (e.g. BRUFEN) in osteoarthritis of the knee designed to show superiority. The lack of a significant difference is interpreted as equivalence (13). This conclusion is not suitable, however, as its probability of error is much greater than the generally accepted level of 5%.

Since the publication of the COCHRANE Review, two other placebo-controlled "positive" studies have appeared, which however primarily evaluated a surrogate endpoint (width of joint space) of doubtful meaningfulness (a-t 2001; 32: 17-8). Again the glucosamine manufacturer Rotta was involved in both studies (14,15). On the other hand, two more recent studies conducted independently of this manufacturer show no advantage for glucosamine over placebo (16,17).

In our view, questionable evidence is unjustifiably converted into evidence level 1a by the COCHRANE Review on osteoarthritis therapy with glucosamine (DONA 200-S). This is completely in the interests of the supplier Opfermann. The company is using the COCHRANE Review including a large-format German translation of the abstract for a promotional brochure (1).

Once again, a systematic review from the COCHRANE Collaboration, which is regarded as independent, is being misused for marketing purposes and has perhaps even been drawn up for those very purposes (cf. a-t 2002; 33: 26). A conflict of interests is not indicated in the Review. In the meantime however, within COCHRANE the suspicion of a complex conflict of interests has arisen (18).

(R = randomised study, M = meta-analysis)

1

Opfermann: promotional brochure concerning DONA 200-S, "Das Gonarthrose off-Programm", undated, article no. 8172

2

TOWHEED, T.E. et al.: Glucosamine therapy for treating osteoarthritis (COCHRANE Review). In: The COCHRANE Library, Issue 2, 2003. Oxford Update Software

3

CROLLE, G., D'ESTE, E.: Curr. Med. Res. Opin. 1980; 7: 104-9

4

D'AMBROSIO, E. et al.: Pharmatherapeutica 1981; 2: 504-8

5

DROVANTI, A. et al.: Clin. Ther. 1980; 3: 260-72

6

VAJARADUL, Y.: Clin. Therap. 1981; 3: 336-43

7

PUJALTE, J.M. et al.: Curr. Med. Res. Opin. 1980; 7: 110-4

8

LOPES VAZ, A.: Curr. Med. Res. Opin. 1982; 8: 145-9

9

REICHELT, A. et al.: Arzneim.-Forsch./Drug Res. 1994; 44: 75-80

10

QIU, G.X. et al.: Arzneim.-Forsch./Drug Res. 1998; 48: 469-74

11

NOACK, W. et al.: Osteoarthritis Cartilage 1994; 2: 51-9

12

HOUPT, J.B. et al.: J. Rheumatol. 1999; 26: 2423-30

13

MÜLLER-FASSBENDER, H. et al.: Osteoarthritis Cartilage 1994; 2: 61-9

14

REGINSTER, J.Y. et al.: Lancet 2001; 357: 251-6

15

PAVELKA, K. et al.: Arch. Intern. Med. 2002; 162: 2113-23

16

RINDONE, J.P. et al.: West. J. Med. 2000; 172: 91-4

17

HUGHES, R., CARR, A.: Rheumatology 2002; 41: 279-84

18

HERXHEIMER, A.: letter dated 26 May 2003

The abstract also refers to 16 instead of 12 studies and fails to mention that 4 out of the 5 unpublished studies have been excluded.