CT-based surrogate parameters for MRI-based disc height and endplate degeneration in the lumbar spine.

Abstract

This study investigated potential use of computed tomography (CT)-based parameters in the lumbar spine as a surrogate for magnetic resonance imaging (MRI)-based findings. In this retrospective study, all individuals, who had a lumbar spine CT scan and MRI between 2006 and 2012 were reviewed (n = 198). Disc height (DH) and endplate degeneration (ED) were evaluated between Th12/L1-L5/S1. Statistics consisted of Spearman correlation and univariate/multivariable regression (adjusting for age and gender). The mean CT-DH increased kranio-caudally (8.04 millimeters (mm) at T12/L1, 9.17mm at L1/2, 10.59mm at L2/3, 11.34mm at L3/4, 11.42mm at L4/5 and 10.47mm at L5/S1). MRI-ED was observed in 58 (29%) individuals. CT-DH and MRI-DH had strong to very strong correlations (rho 0.781-0.904, p < .001). MRI-DH showed higher absolute values than CT-DH (mean of 1.76mm). There was a significant association between CT-DH and MRI-ED at L2/3 (p = .006), L3/4 (p = .002), L4/5 (p < .001) and L5/S1 (p < .001). A calculated cut-off point was set at 11mm. In the lumbar spine, there is a correlation between disc height on CT and MRI. This can be useful in trauma and emergency cases, where CT is readily available in the lack of an MRI. In addition, in the middle and lower part of the lumbar spine, loss of disc height on CT scans is associated with more pronounced endplate degeneration on MRIs. If the disc height on CT scans is lower than 11mm, endplate degeneration on MRIs is likely more pronounced. Level III, a retrospective study.