Abstract
BACKGROUND: The objectives are to compare the efficacy of intra-articular hyaluronic acid (IA-HA) alone and in combination with anti-inflammatory drugs (IA-HA + AI), corticosteroids (CS) or non-steroidal anti-inflammatory drugs (NSAIDs) in clinical trials and in vivo and in vitro studies of osteoarthritis (OA). METHODS: Data in the BIOSIS, CINAHL, Cochrane Library, EMBASE and Medline databases were collected and analyzed. Random effects models were used to compute the effect size (ES) of the mean difference in pain reduction scores from baseline and the relative risk (RR) of adverse events. The ES of histological scores in vivo and cartilage metabolism in vitro were also calculated. We conducted sensitivity analysis of blinding and intention-to-treat (ITT), compared IA-HA combined with CS vs. IA-HA alone in trials, and compared the effects of HA + AI vs. AI alone in vitro, including anabolic and catabolic gene expression. RESULTS: Thirteen out of 382 papers were included for data analysis. In clinical trials, the ES of pain reduction scores within the 1st month was -4.24 (-6.19, -2.29); 2nd-12th month, -1.39 (-1.95, -0.82); and within one year, -1.63 (-2.19, -1.08), favoring IA-HA + AI (P < 0.001). The ES of RR was 1.08 (0.59, 1.98), and histological scores was 1.38 (-0.55, 3.31). The ES of anabolic gene expression was 1.22 (0.18, 2.25), favoring HA alone (P < 0.05); catabolic gene expression was 0.74 (-0.44, 1.53), favoring HA alone; and glycosaminoglycans remaining was -2.45 (-5.94, 1.03). CONCLUSIONS: IA-HA + AI had greater efficacy for pain relief than IA-HA alone within a one-year period. However, HA + AI down-regulated the ACAN gene when compared with HA alone in vitro.