Introduction

Many drugs have been proposed for treating osteoarthritis (OA), but chondro-protective agents deserve some attention. As the name indicates, these agents are intended to protect the cartilage during the many years of disease evolution. They are usually derived from constituents of the cartilage tissue itself, and exhibit many different properties in lab experiments that are interpreted as being potentially beneficial in the progressive cartilage remodeling and destruction that takes place in OA.

A group of investigators in Belgium have recently reported their findings with glucosamine sulfate (GS) treatment in a 3-year clinical study conducted in patients suffering from OA.

Methods

This was a prospective, randomized, placebo-controlled clinical trial in which patients were randomly allocated to receive either glucosamine sulfate (1500 mg equivalent as powder) or placebo, once daily for 3 years. Outpatients consulting the Bone and Cartilage Metabolism Research Unit of the University Hospital Centre in Liege, Belgium were included in the study if they were older than 50 years and had a clinical and radiological diagnosis of OA of the femoro-tibial compartment according to the American College of Rheumatology criteria. Patients with evidence of inflammatory or post-traumatic disease as well as those with abnormal liver or renal function were not allowed in the study. Concomitant use of analgesic paracetamol or non-steroidal anti-inflammatory drugs was allowed during the whole study.

The primary radiological outcomes variables were changes in joint-space narrowing. The osteoarthritic symptoms were assessed by the Western Ontario and McMaster Universities (WOMAC) index for pain, physical function and stiffness at baseline and after 3 years.

Of the 106 patients included in each group, 71 who received placebo and 68 who received GS had a 3-year assessment. The intent-to-treat analysis reported was somewhat unusual, as it did not use the last available assessment carried forward, but rather assigned the placebo response of completers (i.e. those likely to fare the best) to both placebo and GS missing data; this introduced a systematic bias by artificially reducing variance.

Results

The mean joint space narrowing was smaller in the GS group. The estimated difference from placebo amounted to 0.24 mm for mean joint space narrowing and 0.33 mm for maximum joint space narrowing.

The total WOMAC score in the placebo group was slightly worse (10%) after 3 years than at baseline, but this worsening was not statistically significant. By contrast, the total WOMAC score improved by 12% in the GS group, which was statistically significantly different not only from baseline but also from the placebo group. Additional intake of 'rescue' analgesic or NSAID was similar in both groups.
There were no between-group differences in safety or adverse events reports. Approximately 20% of patients in each group withdrew from the study because of adverse events, but no details are reported as to which type of adverse event led to discontinuation or whether they were study-drug related. It is not known if other commonly used treatments for OA (e.g. physical therapy, exercise, weight maintenance programs, which are also known to influence WOMAC scores beneficially) were also allowed or recorded.

Conclusions and recommendations

Although claiming to have assessed joint structure changes during this study, the authors also recognize the limitations of the trial results with respect to radiological outcomes and the still relatively short observation period for a disease like OA. Indeed radiological findings of small joint space narrowing in the knee are likely indicators of meniscal extrusion or may result from positioning differences because of pain when the knee needs to be fully extended. In addition, OA is now recognized as being a disease of the entire joint, with profound modifications occurring also in adjacent tissue structures, including bone. Therefore the focus on cartilage and its presumed X-ray associated changes should be considered as somewhat unreliable and may be poor estimates of disease progression.

However, the unexpected results relate to reported clinical changes. Assuming validity of the intent- to-treat results in the statistical sense, the magnitude of clinical effect of GS treatment would be assessed as small to moderate, based on the well-validated WOMAC instrument. Nevertheless, a moderate improvement is clinically relevant and at least equal to that seen with other treatments of OA.

Other clinical investigations have reported a similar symptomatic effect with much shorter administration of glucosamine-type substances, but this long-term investigation suggests a sustained clinical usefulness of these products in OA. However, more long-term placebo- controlled studies are required to demonstrate that these drugs are indeed slowing progression of the disease. It is not known whether these beneficial clinical observations can be extrapolated to other pharmaceutical preparations containing glucosamine sulfate.