Up to 40% of Chronic Low Back Pain cases involve facet joints ¹ which may be denervated by subcutaneous lumbar medial branch nerve (MBN) radiofrequency ablation. In a recent clinical study of MRI-guided Focused Ultrasound (MRgFUS) ablation of the facet joint ², no improvement in pain scores was reported by 40% of the patients, potentially due to incomplete destruction of all small nerves innervating the joint. Technical feasibility for using MRgFUS to target the MBN before it branches into small nerves of the facet joint was recently demonstrated in an acute swine model ³. While feasible, currently, direct FUS ablation of spine is not clinically practiced due to potential risk of damaging spinal cord and lumbar nerve roots. The main goals of this project were to determine whether direct MRgFUS ablation of the lumbar MB nerve leads to functional changes and to study the extent of the thermal damage to the targeted and adjacent tissues, including neurological structures. Three Yorkshire swine underwent unilateral ablation of the MB nerves using ExAblate 2000 FUS system and a GE 3 T MRI scanner after institutional animal care and use committee protocol approval. The intubated animals were positioned into a tilted-supine position to provide access to the space between the transverse (TP) and articular processes (AP)(Fig 1a). Two to four nerves were ablated in each animal using 3 - 9 sonications per nerve. Ultrasound frequency was 1.35 MHz and acoustic energy was 500 - 710 J. Thermal dose was calculated by the treatment planning software. Contrast-enhanced (CE) imaging was performed immediately after the treatment using MRI and 48-hours post-treatment - using computerized tomography (CT). The animals were maintained on preventative pain medications for 48 hours following the treatment. Assessment of pain, behavior, ambulation and gait was performed twice a day. After euthanasia, lumbar spine was fixed in formalin and decalcified. Decalcified sample was sectioned transversely into 5 mm thick slices, analyzed for gross pathology and processed routinely for histology, using hematoxylin and eosin (H&E) staining. All targeted nerves received lethal thermal dose (Fig 1b) in a location corresponding to non-perfused region visible between the transverse and articular processes on CE-MRI (Fig 1c,d) and CE-CT images (Fig 2), with tissue near the nerve root, superior to the ablation, appearing normal (Fig 2). No changes in bone contrast uptake were visible on CT. On CE-MRI bone surface near ablation appeared slightly more hypo-intense than bone of the untreated side (Fig 1c, arrowhead). No changes in animal behavior and functions were observed and no signs of superficial or deep pain were detected. There was no difference in animals’ ambulation, gait or pain between the treated and untreated sides. On gross pathology (Fig 2), the targeted region exhibited changes consistent with thermal damage and hemorrhage was observed as brown rim surrounding the ablation region in muscle and penetrating 2-5 mm into the bone. The gross appearance of the nerve root slices was normal. Histological analysis confirmed that changes in bone tissue did not extend deep towards the spinal canal leaving 5 to 7 mm of intact bone between the ablation and the spinal canal (Fig 3). Compared to a normal nerve (Fig 4a), treated nerve exhibited stromal hyalinization and nuclear pyknosis (Fig 4b) consistent with coagulative necrosis. The vessels near the ablated nerve appeared necrotic with signs of thrombosis. Muscle adjacent to the nerve was necrotic, with macrophages infiltrating the periphery of the ablation zone (inflammatory reaction). Compared to normal bone tissue (Fig 4c), the bone adjacent to the MBN exhibited coagulative necrosis: shrinkage, cytoplasmic hypereosinophilia and nuclear pyknosis of osteoblasts and hematopoietic cells, necrotic vessels, and hemorrhage. Spinal cord showed no histological changes. The nerve root sampled immediately above the ablation site (Fig 5) appeared normal.

Direct FUS targeting of the lumbar MBN achieves thermal necrosis of the nerve with minimal necrosis of the adjacent bone and muscle tissue. Visualization of spine bone necrosis is challenging on MRI due to short relaxation time of the cortical bone. The extent of bone necrosis is limited to the superficial layer of the bone, with no changes in the spinal cord, confirming the protective effect of spine bone acoustic attenuation ⁴. No changes in the nerve roots demonstrate that small size of the individual FUS lesions makes FUS ideally-suited for neurological applications requiring minimal ablation “foot-print” to spare adjacent critical structures.

The functional and histological outcomes of this study provide preliminary evidence of the safety of direct MRgFUS ablation of the MBN, potentially broadening the bone MRgFUS applications to spine malignancies.