Effects of glucosamine sulfate on the use of rescue non-steroidal anti-inflammatory drugs in knee osteoarthritis: Results from the Pharmaco-Epidemiology of GonArthroSis (PEGASus) study

Abstract Background and objective The use of Symptomatic Slow-Acting Drugs in Osteoarthritis (SYSADOAs) may be expected to decrease the use of concomitant medications for rescue analgesia, including non-steroidal anti-inflammatory drugs (NSAIDs). The Pharmaco-Epidemiology of GonArthroSis (PEGASus) study was designed to assess this possibility. Methods PEGASus was a cohort study of continuous recruitment of patients with "dynamic" exposure to the investigated SYSADOA (crystalline glucosamine sulfate, glucosamine hydrochloride, chondroitin sulfate, diacerein, and avocado–soybean unsaponifiables, all at approved dosages). Investigators were rheumatologists or general practitioners randomly selected from French telephone lists. Patients diagnosed with knee osteoarthritis (OA) were recruited when consulting an investigator for a symptom flare and were prescribed, or not, one of the SYSADOAs as per clinical judgment. Follow-up visits were as per routine medical practice in the 12 months following enrollment, with telephone interviews after 1 month and at 4-month intervals thereafter up to 24 months. Use of NSAIDs was recorded, as well as the dynamism of treatment exposure consisting of continuing the prescribed SYSADOA, switching, discontinuation or initiation of a SYSADOA. Patient exposure was expressed in 2-month time units, with any NSAID use as Yes/No binary outcome during each unit. Odds ratios [OR and 95% confidence interval (CI)] of NSAID use were calculated for periods of exposure to each SYSADOA, by multivariate logistic regression for an 80% power and 95% confidence to see a decrease of at least 15%. Results This report consists of the full data pertaining to crystalline glucosamine sulfate, while results of other SYSADOAs were summarized as available from the French Health Authority (HAS) website (www.has-sante.fr). Of 6451 patients in the PEGASus cohort, 315 patients received crystalline glucosamine sulfate, they were exposed for 481 2-month time units and had an incident use of NSAIDs of 18.7%. In the control cohort (9237 time units) NSAID incident use was 23.8%. Crystalline glucosamine sulfate significantly decreased the risk of NSAID consumption by up to 36% (OR = 0.64; 95% CI: 0.45–0.92) in the primary analysis foreseen by the protocol; OR was 0.74 (95% CI: 0.54–1.01), i.e. at the very limit of significance, in a sensitivity analysis accounting for an extension of the study and of the control cohort. None of the other SYSADOAs showed any hint of a decrease in the use of NSAIDs. Conclusion Crystalline glucosamine sulfate was the only SYSADOA that decreased the use of NSAIDs in this pharmaco-epidemiology study in patients with knee OA.