POS1099 EFFICACY OF INTRA-ARTICULAR CORTICOSTEROID INJECTIONS IN KNEE OSTEOARTHRITIS: A SYSTEMATIC REVIEW AND META-ANALYSIS OF RANDOMIZED CONTROLLED TRIALS

Abstract

Background:Knee osteoarthritis (OA) is a frequent degenerative disease representing an important health and economic burden. Symptomatic medical treatments available include intra-articular (IA) injections of corticosteroids (GC) but their efficacy is debated. In addition, safety signals regarding cartilage damage with IA GC have been highlighted in a few studies.Objectives:To perform a meta-analysis of studies assessing IA GC efficacy and safety in knee OA.Methods:A systematic literature review and a meta-analysis of randomized controlled trials (RCTs) assessing the effect of GC IA injections versus other interventions (IA Hyaluronic Acid, IA placebo, IA NSAID, oral NSAID or physiotherapy) in knee OA was performed. The effect of the interventions on pain and function were extracted from the single studies and pooled and are presented as short term (<6weeks), medium term (6-24 weeks) and long term (>24 weeks) follow-up period. Standardized mean differences (SMD) are reported.Results:Of 520 studies screened, 23 were included in the SLR and 14 subsequently included in the MA. While IA GC showed a superior effect compared to control on both pain (SMD -0.61 (95% CI -1,25, 0.03)) and function (SMD -1.02 (95% CI -2.14, 0.10)) in short term follow-up; long term follow-up analysis favored controls (IA HA, IA NSAID, physiotherapy) for both pain (SMD 0.68 (95% CI -0.11, 1.47)) and function (SMD 0.88 (95% CI -0.36, 2.12) outcomes (Figure 1). No difference was found between interventions in the medium term. Safety data were reported in 18/23 studies (n= 1936/2314 patients); and side effects were reported as follows: arthralgia (69 IA GC patients, 146 IA HA patients, and 20 saline patients); site injection pain (7 in the IA GC group, 2 in the IA saline group, 14 in the IA HA group); 16 post injection knee swelling without signs of septic arthritis in the IA GC group and 24 in the IA HA group. In one study assessing cartilage effects of GCs, the rate of cartilage loss was greater in the GC group with a reduction of cartilage thickness at 2 year compared to placebo group. No difference was observed in the progression of cartilage denudation or bone marrow lesion. On the contrary, another study showed no effect of injections on the cartilage structure.Conclusion:We demonstrate in this work that IA GC injections reduce pain and improve function in the early phase (≤6 weeks) of treatment. In the long term (≥24 weeks), other intervention such as IA HA injections or physiotherapy seem to be more efficient, but this effect was largely driven by single studies with large effect sizes and the comparators were heterogeneous.Figure 1.Knee pain outcome at short term (≤6weeks) (A), medium term (>6 & <24 weeks) (B), and long term (≥24 weeks) (C) follow up.Disclosure of Interests:None declared.