J Neurosurg Spine. 2023 Oct 20:1-8. doi: 10.3171/2023.8.SPINE23450. Online ahead of print.
ABSTRACT
OBJECTIVE: The purpose of this study was to compare and contrast lumbar bone quality and osteoporosis/osteopenia screening results via dual-energy x-ray absorptiometry (DEXA), CT, and MRI.
METHODS: A consecutive series of 426 candidates screened for lumbar disc replacement over a 5-year period beginning in 2018 was reviewed. Patients with a preoperative lumbar spine DEXA scan and a CT and/or MRI scan were included. The primary outcome measures included the bone mineral density (BMD) and osteoporosis or osteopenia classification from DEXA scans, Hounsfield units (HUs) for CT, and vertebral bone quality (VBQ) assessment for MRI. Patients were included if they had a DEXA scan within 1 year of an MRI or CT scan. DEXA BMD scores from composite or level-by-level reports were recorded. Asynchronous MRI and CT measurements were conducted using PACS. Interrater and intrarater reliability scores were generated for both CT and MRI measurements and ranged from 1.000 for MRI L1-4 scans to 0.683 for MRI VBQ.
RESULTS: All 3 types of scans were statistically significantly correlated with one another; however, CT was more strongly correlated with the lumbar DEXA value (r = 0.439, p < 0.001). The correlation between MRI VBQ and DEXA was -0.103, (p < 0.045). The CT level-by-level measurements correlate with the corresponding level-by-level DEXA BMD values (correlation ranging from 0.531 to 0.289, p < 0.001 to p = 0.007). CT HU values were more strongly related to osteoporosis/osteopenia classification based on DEXA T-scores than were MRI VBQ values. Receiver operating characteristic analyses found that the area under the curve was 0.817 for CT and 0.539 for MRI.
CONCLUSIONS: These results demonstrate that CT HUs more closely correlate to DEXA scores than MRI VBQ in this population of patients undergoing surgery for symptomatic disc degeneration. Thus, CT may be an alternative to DEXA for assessing VBQ in this population.
PMID:37862719 | DOI:10.3171/2023.8.SPINE23450
