INTRODUCTION: Diagnostic imaging following cervical fusion often involves magnetic resonance imaging (MRI) to evaluate neural structures and computed tomography (CT) and/or dynamic radiographs to evaluate instrumentation and arthrodesis. CT and radiographs use ionizing radiation that increases risk for malignancy. Using MRI alone would protect patients and decrease healthcare utilization. METHODS: Single institution retrospective study evaluating patients who received postoperative CT and MRI following ventral cervical fusion procedures. Radiographic standards were used to evaluate for arthrodesis or pseudo-arthrosis. Blinded observers evaluated signal intensity using a region of interest (ROI) over the instrumented disc space, homogenous cerebral spinal fluid (CSF), and distal uninstrumented vertebral body (VB). A quotient (ROIinterspace/ROICSF and ROIinterspace/ROIVB) was used to create T1- and T2-interspace interbody scores (IIS). Additionally, T1 hypointense lines along endplates and facet joints were evaluated for degree of discontinuity. RESULTS: Study included 60 patients (50% female) with 110 instrumented levels and mean age 51.7 years-old. For T1-weighted MRI, median T1-IISCSF for arthrodesis was 176.2 compared to 130.9 for pseudo-arthrosis (p < 0.0001) while T1-IISVB for arthrodesis was 68.5 and pseudo-arthrosis was 52.7 (p < 0.0001). For T2-weighted MRI, median T2-IISCSF for arthrodesis was 27.7 and 14.2 for pseudo-arthrosis (p < 0.0001) while T2-IISVB for arthrodesis was 67.9 and 41.0 for pseudo-arthrosis (p < 0.0001). Median T1 hypointense discontinuity was graded at the endplates, left facet, and right facet as 1.5, 0.25, and 0.5 compared to pseudo-arthrosis as 1.0, 0.0, and 0.0, respectively (p < 0.0001, p = 0.002, and p = 0.001). The greatest area under the curve (AUC) after receiver operator characteristic (ROC) analysis was T1-IISVB (0.7743). CONCLUSIONS: We describe a novel scoring system that may help determine arthrodesis versus pseudo-arthrosis on MRI following ventral cervical fusion. Postoperative symptomatic patients may only require MRI for workup, which would protect patients from harm while decreasing healthcare utilization.
Read full abstract