Magnetic resonance imaging analysis of lumbar disc changes below scoliosis fusions. A prospective study

Abstract

Study design: The authors of this prospective study examined the preoperative and 3-year postoperative magnetic resonance images of 14 patients undergoing anterior and posterior fusion and/or posterior fusion only for scoliosis. All magnetic resonance images were ready by two independent neuroradiologists, who were blinded to the purposes of the study, for the presence of disc narrowing, signal decrease on T2, or herniated nucleus pulposus before and after surgery. Particular attention was paid to the disc changes at the level directly below the end vertebral level of the fusion and two levels below the fusion in the lumbosacral spine existing before surgical intervention.

Objectives: To evaluate the potential for disc degeneration distal to long scoliosis fusions with end fusion levels in the mid to lower lumbar spine.

Summary of background data: The determination of end levels of fusion for contructs presently used to manage adult scoliotic deformity has been evaluated in terms of correction of curvature and late decompensation in coronal and sagittal plane balance after fusion. However, the natural history of the caudal, free-motion segments in terms of degeneration and/or correlation with pain has not yet been addressed.

Methods: Fourteen patients undergoing scoliosis fusion underwent magnetic resonance imaging before surgery and approximately 3 years after surgery. The scans were reviewed by two independent neuroradiologists who looked at three degenerative indices at the disc below the area of scoliosis fusion. The authors analyzed rates of change of the three degenerative indices in the pre- and postoperative magnetic resonance images and created associations between the observed changes on the magnetic resonance images and the clinical outcomes of pain, the presence or absence of solid fusion, and the need for repeat surgery.

Results: Estimates of the rates of change of the three degenerative indices one or two levels below the fusion were as follow: the chance of disc narrowing, .2-34%; the chance of a decreasing signal on T2, 5-54%, with a 23% incidence among this group; and the chance of herniated nucleus pulposus, 0-34%. There was a significant correlation between the presence of back and/or leg pain and the signal decrease one level below the fusion (P = .04).

Conclusions: If these results are corroborated in a larger sample size, surgeons who manage deformity may have to consider altering fusion levels at the time of fusion based on magnetic resonance imaging predictors. The present data may help to inform patients about the risk of developing junctional degenerative changes and potential symptoms from these changes below scoliosis fusions.