Statistics from Altmetric.com
We would like to thank Zeng et al1 for their comments on the Multicentre Osteoarthritis interVEntion trial with SYSADOA (MOVES)2 and the Glucosamine/chondroitin Arthritis Intervention Trial (GAIT)3 and its long-term follow-up.4
In MOVES,2 glucosamine plus chondroitin sulfate was non-inferior to celecoxib (primary outcome: mean decrease in Western Ontario and McMaster (WOMAC) pain score) in 606 patients with knee osteoarthritis with moderate–severe pain. A placebo arm was not included because celecoxib, the reference treatment, had proven efficacy versus placebo in knee osteoarthritis, for example5 ,6; so, this was felt to be unethical and unnecessary. While Zeng et al1 agreed with this decision, they brought up some interesting points.
The original 24-week randomised, controlled GAIT trial3 compared the effects of glucosamine and/or chondroitin sulfate, celecoxib and placebo on knee osteoarthritis pain. There was a significant benefit of glucosamine plus chondroitin sulfate over placebo (primary outcome: 20% decrease in WOMAC pain score at 24 weeks) among a prespecified subgroup of patients with moderate–severe pain. The 24-month GAIT follow-up,4 as Zeng et al1 pointed out, failed to demonstrate significant differences between treatment arms. However, the 24-week GAIT trial3 included 1583 patients (1258 (79%) completed the study), while the 24-month follow-up4 only included 662 patients (349 (only 53%) completed). Furthermore, GAIT was designed and powered (sample size required: 1270) to examine efficacy over 24 weeks, not 24 months. The long-term extension study was found to be underpowered, with an unusually high dropout rate and very low pain levels at baseline; therefore, the results of GAIT3 were likely to be more reliable than those of the follow-up study.4
Zeng et al1 also mentioned the Long-term Evaluation of Glucosamine Sulfate (LEGS) study,7 which randomised 605 patients with knee osteoarthritis to glucosamine and/or chondroitin sulfate or placebo. LEGS8 found that glucosamine plus chondroitin sulfate significantly reduced joint space narrowing, but not pain, versus placebo (there were two primary outcomes). However, direct comparison of the results of MOVES2 and LEGS7 was not possible as there were important differences in study design because the LEGS study7 examined joint space narrowing while MOVES2 studied pain. Patients in LEGS7 had less severe pain than those in MOVES2 (baseline mean standardised WOMAC pain 33.5 vs 74.2). In addition, patients in LEGS could use non-steroidal anti-inflammatory drugs or opioid analgesics throughout the study, which could have masked any pain differences between groups, while in MOVES, patients could only take acetaminophen as rescue, but not in the 24 h before clinical evaluation. Lastly, MOVES used pharmaceutical-quality chondroitin sulfate and glucosamine (Bioiberica SA, Barcelona, Spain), whereas the products used in LEGS were dietary supplements. As discussed in a recent publication,8 the source and purity of these extracted natural products are important, and not all products have the same efficacy.
Zeng et al1 also pointed out that some studies of glucosamine and chondroitin sulfate included manganese ascorbate, which could have improved efficacy. It is always difficult to compare between different studies, which is why the results of GAIT,3 which directly compare glucosamine plus chondroitin sulfate with placebo and show a significant effect among those with moderate–severe pain at 24 weeks, are important. MOVES2 supported this result by showing a pain response to glucosamine and chondroitin sulfate that was non-inferior to celecoxib.
Competing interests MCH is a consultant to Bioiberica SA, Bristol Myers Squibb, Eli Lilly, EMD Serono SA, Iroko Pharmaceuticals, Novartis Pharma AG, Pfizer, Samumed and Theralogix, and owns stock in Theralogix.
Provenance and peer review Commissioned; internally peer reviewed.