BackgroundCytokines such as tumor necrosis factor (TNFα) have been shown to elicit inflammatory and catabolic events in the joints of patients with osteoarthritis. Recent RCTs demonstrated that TNFα inhibition has no effect on pain and MRI-detected synovitis or bone marrow lesions in patients with erosive hand OA1-3. However, the progression of bone erosions was reduced in a subgroup of patients with more clinically swollen distal interphalangeal joints in one RCT1. Consequently, it remains possible that TNFα inhibition may have beneficial effects in specific subgroups of patients with a high inflammatory component.ObjectivesWe aimed to evaluate the efficacy and safety of a TNFα inhibitor, adalimumab (ADA), in a proof-of-concept study in patients with inflammatory OA of the knee.MethodsOKINADA was a 52-week, randomized, double-blind, placebo-controlled, parallel-group study done at 11 sites in Canada (NCT02471118). Eligible participants were adults (aged ≥18 years) with a diagnosis of OA of the index knee and classified according to American College of Rheumatology criteria, including radiological evidence of OA (Kellgren-Lawrence grades 2 or 3) with clinical signs of knee effusion. Subjects had persistent knee pain of ≥ one month duration with a pain score of ≥ 4 (0-10 NRS) in the index knee at screening and baseline despite conventional treatment with maximum tolerated acetaminophen and/or non-steroidal anti-inflammatory drug. Patients were randomly assigned (1:1) to receive subcutaneous 40 mg ADA every 2 weeks or placebo (PBO). Primary endpoint was the Outcome Measures in Rheumatology and Osteoarthritis Research Society International set of responder criteria (OMERACT-OARSI) at week 16 defined as: (1) improvement in pain or function ≥50% and an absolute change ≥20 mm; or (2) improvement of ≥20% with an absolute change ≥10 mm in at least two of the following three categories: pain, function, and patient’s global assessment. Secondary endpoints included: the Knee Injury and Osteoarthritis Outcome Score (KOOS) for the domains of pain, activities of daily living (ADL), OA symptoms, sport and recreation function (SRF), and knee-related quality of life (QoL), patient’s global assessment of disease status (PGAD), investigator global assessment of disease status (IGAD), and expanded Target Joint Assessment (TJA) score.ResultsA total of 59 patients were randomized (29 to PBO, 30 to ADA). The primary endpoint was not met: OMERACT-OARSI combined (ADA: 9 [30.0%] vs PBO: 7 [24.1%], p=0.62). For KOOS pain, ≥20% improvement was noted in 11 (36.7%) ADA vs 7 (24.1%) PBO patients (p=0.30), and ≥50% improvement in 5 (16.7%) ADA vs 6 (20.7%) PBO patients (p=0.69). There were no significant treatment-group differences in baseline to 16-week change in continuous secondary endpoints (ADA vs PBO: KOOS ADL 6.5 vs 8.4 (p=0.71), KOOS QoL 10.1 vs 7.4 (p=0.66), KOOS symptoms 7.8 vs 11.5 (p=0.42), KOOS SRF 5.8 vs 7.7 (p=0.76), PGAD -1.0 vs 0.1 (p=0.10), IGAD -1.5 vs -2.1 (p=0.30), TJA -2.4 vs -2.2 (p=0.87) or in lab markers (ESR, CRP). There were 11 withdrawals (4 ADA, 7 PBO) of which 2 were for adverse events (1 ADA, 1 PBO) and 2 for increasing knee pain (1 ADA, 1 PBO). No new safety signals were identified and there were no serious adverse events.ConclusionAlthough the treatment was safe, short-term treatment with anti-TNFα therapy does not appear to provide clinically meaningful improvements in OA symptoms in patients with established radiographic knee OA. Analyses of structural endpoints will be reported when results are available.