Rotation effect and anatomic landmark accuracy for midline placement of lumbar artificial disc under fluoroscopy

Purpose: Total disc arthroplasty can be a viable alternative to fusion for degenerative disc disease of the lumbar spine. The correct placement of the prosthesis within 3 mm from midline is critical for optimal function. Intra-operative radiographic error could lead to malposition of the prosthesis. The objective of this study was first to measure the effect of fluoroscopy angle on the placement of prosthesis under fluoroscopy. Secondly, determine the visual accuracy of the placement of artificial discs using different anatomical landmarks (pedicle, waist, endplate, spinous process) under fluoroscopy. Methods: Artificial discs were implanted into three cadaver specimens at L2-3, L3-4, and L4-L5. Fluoroscopic images were obtained at 0°, 2.5°, 5°, 7.5°, 10°, and 15° from the mid axis. Computerized tomography (CT) scans were obtained after the procedure. Distances were measured from each of the anatomic landmarks to the center of the implant on both fluoroscopy and CT. The difference between fluoroscopy and CT scans was compared to evaluate the position of prosthesis to each anatomic landmark at different angles. Results: The differences between the fluoroscopy to CT measurements from the implant to pedicle was 1.31 mm, p < 0.01; implant to waist was 1.72 mm, p < 0.01; implant to endplate was 1.99 mm, p < 0.01; implant to spinous process was 3.14 mm, p < 0.01. When the fluoroscopy angle was greater than 7.5°, the difference between fluoroscopy and CT measurements was greater than 3 mm for all landmarks. Conclusions: A fluoroscopy angle of 7.5° or more can lead to implant malposition greater than 3 mm. The pedicle is the most accurate of the anatomic landmarks studied for placement of total artificial discs in the lumbar spine.