Different cervical vertebral bone quality scores for bone mineral density assessment for the patients with cervical degenerative disease undergoing ACCF/ACDF: computed tomography and magnetic resonance imaging-based study

Abstract

BackgroundBone mineral density (BMD) is important for the outcome of cervical spine surgery. As the gold standard of assessing BMD, dual-energy X-ray absorptiometry scans are often not ordered or go unreviewed in patients’ charts. As the supplement, MRI-based vertebral bone quality (VBQ) was found to accurately predict osteopenia/osteoporosis and postoperative complications in lumbar spine. However, discussion of the efficiency of VBQ in cervical spine is lacking. And measurement methods of VBQ in cervical spine are diverse and not universally acknowledged like lumbar spine. We aimed to compare the predictive performance of three kinds of different Cervical-VBQ (C-VBQ) scores for bone mineral density assessment in patients undergoing cervical spine surgery. HU value of cervical spine was set as a reference.MethodsAdult patients receiving cervical spine surgery for degenerative diseases were retrospectively included between Jan 2015 and Dec 2022 in our hospital. The VBQ scores and HU value were measured from preoperative MRI and CT. The correlation between HU value/C-VBQs (named C-VBQ1/2/3 according to different calculating methods) and DEXA T-score was analyzed using univariate linear correlation and Pearson’s correlation. We evaluated the predictive performance of those two parameters and achieved the most appropriate cutoff value by comparing the receiver operating characteristic (ROC) curves.Results106 patients (34 patients with T ≥ − 1.0 vs 72 patients with T < − 1.0) were included (mean age: 51.95 ± 10.94, 48 men). According to Pearson correlation analysis, C-VBQ1/2/3 and HU value were all significantly correlated to DEXA T-score (Correlation Coefficient (r): C-VBQ1: − 0.393, C-VBQ2: − 0.368, C-VBQ3: − 0.395, HU value: 0.417, p < 0.001). The area under the ROC curve (AUC) was calculated (C-VBQ1: 0.717, C-VBQ2: 0.717, C-VBQ3: 0.727, HU value: 0.746). The AUC of the combination of C-VBQ3 and HU value was 0.786. At last, the most appropriate cutoff value was determined (C-VBQ1: 3.175, C-VBQ2: 3.005, C-VBQ3: 2.99, HU value: 299.85 HU).ConclusionsDifferent MRI-based C-VBQ scores could all be potential and alternative tools for opportunistically screening patients with osteopenia and osteoporosis before cervical spine surgery. Among them, C-VBQ calculated in ASIC2–C7/SIT1-CSF performed better. We advised patients with C-VBQ higher than cutoff value to accept further BMD examination.