Magnetic resonance imaging artifact following anterior cervical discectomy and fusion with a trabecular metal cage

Abstract

OBJECT This study was undertaken to evaluate the impact of postoperative MRI artifact on the assessment of ongoing spinal cord or nerve root compression after anterior cervical discectomy and fusion (ACDF) using a trabecular tantalum cage or bone autograft or allograft. METHODS The authors conducted a retrospective review of postoperative MRI studies of patients treated surgically for cervical disc degenerative disease or cervical instability secondary to trauma. Standard ACDF with either a trabecular tantalum cage or interbody bone graft had been performed. Postoperative MR images were shown twice in random order to each of 3 assessors (2 spine surgeons, 1 neuroradiologist) to determine whether the presence of a tantalum interbody cage and/or anterior cervical fixation plate or screws imparted MRI artifact significant enough to prevent reliable postoperative assessment of ongoing spinal cord or nerve root compression. RESULTS A total of 63 patients were identified. One group of 29 patients received a tantalum interbody cage, with 13 patients (45%) undergoing anterior plate fixation. A second group of 34 patients received bone auto- or allograft, with 23 (68%) undergoing anterior plate fixation. The paramagnetic implant construct artifact had minimal impact on visualization of postoperative surgical level spinal cord compression. In the cage group, 98% (171/174) of the cases were rated as assessable versus 99% in the bone graft group (201/204), with high intraobserver reliability. In contrast, for the assessment of ongoing surgical level nerve root compression, the presence of a tantalum cage significantly decreased visualization of nerve roots to 70% (121/174) in comparison with 85% (173/204) in the bone graft group (p < 0.001). When sequences using turbo spin echo (TSE), a T2-weighted axial sequence, were acquired, nerve roots were rated as assessable in 88% (69/78) of cases; when only axial T2-weighted sequences were available, the nerve roots were rated as assessable in 54% (52/96) of cases (p < 0.01). The presence of anterior plate fixation had minimal impact on visualization of the spinal cord (99% [213/216] for plated cases vs 98% [159/162] for nonplated cases; p = 1.0) or nerve roots (79% [170/216] for plated cases vs 77% [124/162] for nonplated cases; p = 0.62). CONCLUSIONS Interbody fusion with tantalum cage following anterior cervical discectomy imparts significant paramagnetic artifact, which significantly decreases visualization and assessment of ongoing surgical level nerve root, but not spinal cord, compression. Anterior plate constructs do not affect visualization of these structures. TSE T2-weighted sequences significantly improve nerve root visualization and should be performed as part of a standard postoperative protocol when imaging the cervical spine following interbody implantation of materials with potential for paramagnetic artifact.