Background/Purpose: There are currently no licensed therapies which reduce progression of knee OA. Current therapies focus on reducing pain, have limited effect, and some have significant adverse effects. Bone marrow lesions (BMLs) – ill-defined areas of high signal intensity in subchondral bone MRIs – are present in ~80% of patients with symptomatic knee OA and are associated with knee pain and structural progression. Denosumab is a potent antiresorptive monoclonal antibody targeted at RANK Ligand (RANKL). We hypothesised that in patients with symptomatic knee OA, and evidence of a BML on MRI, treatment with a subcutaneous injection of denosumab would reduce BML size, thus reducing knee symptoms.

Methods: The DISKO Trial (ISRCTN96920058) was a single centre, double-blind randomised placebo-controlled trial. Adults with symptomatic knee OA were randomly allocated 1:1 to a single injection of subcutaneous Denosumab (60mg), or matched placebo. All subjects received calcium and vitamin D supplements. Eligibility criteria: age ≥40 years; x-ray changes of tibio- or patella-femoral OA based on knee x-rays within previous 24 months; knee pain in past week ≥3 on a 0-10 numerical rating scale (NRS). Knee must have evidence of BMLs on screening knee MRI (Phillips 3T), plus Kellgren and Lawrence grade 2 or 3 disease in the painful knee and 25-OH vitamin D level > 50nmol/l. Symptoms were assessed at 3 and 6 months, repeat MRI at 6 months. Primary outcome: BML area at 6 months (mm²), from sagittal PDFS pulse sequence. Secondary outcomes: imaging parameters (BML volume), severity of knee pain (10-point NRS), other knee symptoms, and quality of life (KOOS; SF-12v2; EQ-5D-5L questionnaires).

Results: 145 men and women were assessed for eligibility. Of these, 87 were excluded, 58 taking part in the trial (30 randomised to denosumab, 28 to placebo). Mean age of randomised subjects was 64.2 years, mean BMI 29.3 Kg/m²; mean baseline pain in the past week (0-10 NRS) was 6.0 (SD 1.7); median BML area was 787.5mm² (IQR 370.0mm² to 1150.0mm²). A complete case mixed-effects model (N=46) adjusting for baseline found little difference in BML area between groups at 6 months (marginal mean difference = 52.0mm²; 95% CI -151.9 to 255.9; p = 0.62). Six months between-groups differences in BML volume (marginal mean difference = 278.0mm³; 95% CI -1877.4 to 2433.4; p = 0.80) and synovitis-effusion volume (-582.9mm³; 95% CI -3011.3 to 1845.6; p = 0.64) were also small. At 6 months, the denosumab group had slightly greater pain in the past week NRS (0-10) (marginal mean difference = -0.25 points; 95% CI -1.6 to 1.1; p = 0.75) and the KOOS pain (100-0) (marginal mean difference = 5.0 points; CI -5.4 to 15.3; p = 0.35) compared to placebo. There were no significant between-group differences in non-pain KOOS subscales or EQ-5D-5L subscales at 3 or 6 months. Sensitivity analyses using additional covariates and multiple imputation were concordant. There were 57 adverse effects (30 Denosumab; 27 placebo), 40 mild, and 1 serious adverse event, unrelated to the intervention.

Conclusion: We found no signal to suggest a beneficial effect of denosumab on symptoms or structure. These findings do not support use of subcutaneous Denosumab for symptomatic knee OA.

« Back to ACR Convergence 2023

ACR Meeting Abstracts - https://acrabstracts.org/abstract/effect-of-denosumab-on-knee-pain-and-bone-marrow-lesions-in-symptomatic-knee-osteoarthritis/