PIH21 EFFICACY AND SAFETY OF IMRECOXIB, CELECOXIB AND DICLOFENAC IN PATIENTS WITH OSTEOARTHRITIS: A NETWORK META-ANALYSIS

Abstract

Non-steroidal anti-inflammatory drug (NSAIDs) are widely prescribed for relief of the symptoms of osteoarthritis. The efficacy and safety of different NSAIDs are still studied, especially between the cyclooxygenase-2 inhibitor (COX-2) and non-selective NSAIDs (ns-NSAIDs). A new Cox-2 NSAID called imrecoxib was approved by China Food and Drug Administration in 2011, and the efficacy and safety compared with other NSAIDs stay unexplored. This study is aimed to access the overall efficacy and safety of imrecoxib, celecoxib (COX-2), diclofenac (ns-NSAIDs) based on the existing literature. 25 randomized controlled trials (RCTs) published from the time of database establishment to Nov 19, 2019 were systematically reviewed in 8 databases, which compared any of imrecoxib, celecoxib, diclofenac and placebo. The primary outcome was the mean change of different drugs on patient’s pain score (VAS mm); the sencondary outcomes were the rate of total adverse events and gastrointestinal (GI) events. The rate of adverse events was adjusted with standardized doses. Literature with less than 4-week treatment duration were excluded. Celecoxib (standardized mean difference [SMD] -0.52 [95%CI -0.96 to -0.09]) was significantly more effective than placebo, while imrecoxib (-0.67 [-1.97 to 0.63]) and diclofenac (-0.49 [-1.06 to 0.08]) were more effective than placebo. Celecoxib (odds ratio [OR] 1.04 [95%CI 0.84 to 1.28]) and diclofenac (1.27 [0.89 to 1.81]) were more unsafe than placebo, whereas imrecoxib was not. Similar results could be found in the analysis of the rate of GI events, celecoxib (1.05 [0.83 to 1.32]) and diclofenac (1.39 [0.95 to 2.03]) were more unsafe than placebo. It seemed that imrecoxib was associated with better efficacy and fewer adverse events. Since there were sparse RCTs have been reported for imrecoxib, the conclusion should be adopted with caution. Meanwhile, a nocebo effect was observed in the analysis of the rate of adverse events, which need further discussion.