Assessing the Effects of Geniculate Artery Embolization in a Nonsurgical Animal Model of Osteoarthritis.

ABSTRACT

PURPOSE: To create a nonsurgical animal model of osteoarthritis (OA) to evaluate the effects of embolotherapy during geniculate artery embolization (GAE). MATERIALS AND METHODS: Fluoroscopy-guided injections of 700 mg of sodium monoiodoacetate were performed into the left stifle in 6 rams. Kinematic data were collected before and after induction. At 10 weeks after induction, Subjects 1 and 4-6 underwent magnetic resonance (MR) imaging with dynamic contrast enhancement (DCE) and Subjects 1, 3, and 4-6 underwent angiography with angiographic scoring to identify regions with greatest disease severity for superselective embolization (75-250-µm microspheres). Target vessel size was measured. At 24 weeks after angiography, DCE-MR imaging, angiography, and euthanasia were performed, and bilateral stifles were harvested. Medial/lateral tibial and femoral condylar, patellar, and synovial samples were cut, preserved, decalcified, and scored using the Osteoarthritis Research Society International criteria. The stifle and synovium Whole-Organ Magnetic Resonance Imaging Score and Multicenter Osteoarthritis Study score were determined. The volume transfer constant (Ktrans) and extracellular volume fraction (ve) were calculated from DCE-MR imaging along the lateral synovial regions of interest. RESULTS: The mean gross and microscopic pathological scores were elevated at 38 and 61, respectively. Mean synovitis score was elevated at 9.2. Mean pre-embolization and postembolization angiographic scores were 5 and 3.8, respectively. Mean superior, transverse, and inferior geniculate artery diameters were 3.1 mm ± 1.21, 2.0 mm ± 0.50, and 1.6 mm ± 0.41 mm, respectively. Mean pre-embolization and postembolization cartilage and synovitis scores were elevated at 35.13 and 73.3 and 5.5 and 9.2, respectively. The Ktrans/ve values of Subjects 4, 5, and 6 were elevated at 0.049/0.38, 0.074/0.53, and 0.065/0.51, respectively. Altered gait of the hind limb was observed in all subjects after induction, with reduced joint mobility. No skin necrosis or osteonecrosis was observed. CONCLUSIONS: A nonsurgical ovine animal knee OA model was created, which allowed the collection of angiographic, histopathological, MR imaging, and kinematic data to study the effects of GAE.