Glucosamine-Chondroitin Shown Not Inferior to Celecoxib for Knee OA Pain

April 28, 2014

(OARSI 2014) A 600-patient randomized trial has found a glucosamine-chondroitin formulation non-inferior to celecoxib for the pain of knee osteoarthritis. Also, unlike placebo, it shows noteworthy effects on pain regions in the brain.

Two studies reported at the Osteoarthritis Research Society International Meeting last week in Paris offered new evidence in favor of the effects of glucosamine and chondrotin sulfate for knee osteoarthritis (OA). Glucosamine hydrochloride and chondroitin sulfate (GHCl+CS) relieved pain of knee OA, and was non-inferior to celecoxib. The supplement also reduced the activity of pain centers in the brain, as measured by functional MRI (fMRI).

In the randomized trial of glucosamine-chondroitin and celecoxib, 606 patients who met American College of Rheumatology criteria for knee OA, had Kellgren-Lawrence grade 2 or 3 radiologic severity, and WOMAC pain of >301 (on a 0-500 scale) were randomized in double-blind, double-dummy fashion to either GHCl 250 mg + CS 200 mg three times a day or celecoxib 200 mg per day. (This was the dose of GHCl+CS used in the NIH-funded Glucosamine Arthritis Intervention Trial published in the New England Journal of Medicine in 2006.) The primary outcome was mean decrease in WOMAC pain at six months; the non-inferiority margin was 40 (on a 0-500 scale).

Five hundred and twenty-two patients (86.1%) completed the trial. At randomization, the mean WOMAC pain was about 370 in each group; at six months it was about 185 in each group, which met the non-inferiority margin.

The multicenter MOVES trial is sponsored by the pharmaceutical firm Bioiberica.

The randomized double-blind trial that  used fMRI to compare chondroitin sulfate (CS) versus placebo measured the brain response of 24 patients who were using the supplement and 27 patients in the placebo group. CS significantly reduced fMRI activity in key regions of the pain-processing network.

An algometer was used to apply painful pressure on the patella surface and on the knee medial interline during the acquisition of two fMRI sequences, before and after CS or placebo. Subjects also rated the subjective pain. The main outcome was attenuation of the response to the painful stimulus in the pain-processing system of the brain.

The difference in patient-reported subjective pain was not statistically significant, but the fMRI images did show a significant effect for CS. The fMRI of patella pain showed a greater post-treatment reduction in signal in the CS group in a posterior mesencephalon region including the periaqueductal gray (PAG). The CS group also showed more reduction in the primary somatosensory cortex, including the cortical representation of the leg.