Repurposing an anti-epilepsy glutamate receptor antagonist for the prevention of post-traumatic osteoarthritis

Purpose: Osteoarthritis (OA) affects 27 million people in the US and has no disease modifying treatments. Since joint injuries lead to OA, intervention at the time of injury could prevent injury-induced OA. Synovial fluid (SF) glutamate concentrations increase in arthritis. We previously found that activation of kainate (KA) and AMPA glutamate receptors (GluRs) increases interleukin-6 release and causes pain, that AMPA and KA GluRs localise to osteoarthritic bone, cartilage and synovium and that NBQX (AMPA/KA GluR antagonist) reduces knee swelling, gait abnormalities and joint destruction in a rat inflammatory arthritis model, and two models of post-traumatic OA (PTOA): rat medial meniscal transection (MNX), mouse anterior cruciate ligament (ACL) rupture. Here we tested a GluR antagonist (Drug A) which was originally developed for epilepsy treatment and has passed Phase 1 clinical trials, and investigated the window of opportunity for treatment based on SF glutamate concentrations at various times following human ACL rupture. We hypothesised that drug A will reduce pain, inflammation and degeneration in the mouse ACL rupture PTOA model and tested safety and efficacy of the maximum tolerated dose (MTD) and the minimum effective dose (MED) of this drug. Methods: SF glutamate levels (ELISA, Novus) were measured at different times post-ACL injury in 24 patients (18 men, 6 women). For mouse ACL rupture, the right stifle is held in flexion and a single 12N load (ElectroForce® 3200, BOSE) applied to rupture ACL in anaesthetised 12-week-old C57Bl6J mice (Animals (Scientific Procedures) Act 1986. Project License: P287E87DF). A single 10μl intra-articular (i.a.) injection of the drug A MTD (dissolved in 100mM NaOH, n=4) was administered to mice immediately following ACL rupture, whilst additional mice received an i.a. injection of vehicle (100mM NaOH, n=4) immediately following ACL rupture. A second experiment examined the effects of two i.a. injections (one immediately following ACL rupture (day 0) and one 24 hours later (day 1)) of the MTD, 1:5 and 1:10 dilution of MTD to establish the MED (n=5, with 2x i.a. of 100mM NaOH vehicle). Contralateral knees (intact ACL) were used as controls. All mice received analgesia (Temgesic) prior to ACL rupture. Over 21 days, lameness was scored and stifle swelling measured (callipers, mean of 3 measurements, days 0, 1, 2, 3, 7, 14, 21). On day 21, all animals were culled and knees scored for degradation (OARSI score, mean of 2 observers blinded to treatment) and inflammation. All interventions, assessments and measurements were performed blinded to treatment and powered to detect differences between ACL rupture and healthy knees (n=5). Results: Glutamate concentrations varied significantly with time post ACL injury (ANOVA p=0.03), being greater at 0-20 (58.3μM±5.84, p<0.05, n=12) and 21-100 (70.64μM±10.32, p<0.01, n=8) weeks post-injury, compared with 100-500 (34.9μM±7.79, n=7) weeks post-injury (Fig. 1a). In vivo, by day 2, drug A (single dose MTD) had reduced mean knee swelling by ∼50% and did not differ significantly from day 0 pre-ACL rupture measurements, whereas vehicle control knee swelling was significantly increased until day 7 (p<0.01, GLM). Mean lameness scores were reduced by drug A over time (GLM, Day p<0.001, Treatment p<0.001) although pairwise comparisons on any given day were not significantly different. Drug A MTD also reduced histological inflammation (∼60%, p<0.001, ANOVA, Fig. 1b) and degradation (∼40%, Fig. 1c) scores at day 21. Two i.a. injections of all 3 drug A doses significantly reduced knee swelling compared to vehicle on day 1, with the two lower doses maintaining this reduction on days 2 and 21 (p<0.05, GLM). All doses reduced mean lameness scores, with the lowest dose significantly reducing lameness score compared to vehicle on days 3 (p<0.05) and 7 (p<0.01, GLM). Day 21 histological degradation scores were also reduced (∼25%) the most by the lowest double dose of drug A. Conclusions: This study provides evidence that an AMPA/KA GluR antagonist, already approved for use in humans, reduces inflammation, lameness and joint damage in PTOA. We also show, for the first time, a window of opportunity to administer these drugs in humans following ACL injury, when glutamate levels are still high (up to 100 weeks post-injury). Intriguingly, the lowest dose in this study appeared to be the most effective for all readouts. Limitations of the study are potential detrimental effects of the 100mM NAOH vehicle (although this does not increase degradation or inflammation score over ACL rupture alone, Gilbert et al. JOR) and the statistical power being limited to detect differences between ACL rupture and healthy, as opposed to ACL rupture with drug or vehicle.