Diffusion tensor tractography of the lumbar nerves before a direct lateral transpsoas approach to treat degenerative lumbar scoliosis

OBJECTIVE The purpose of this study was to determine the relationship between vertebral bodies, psoas major morphology, and the course of lumbar nerve tracts using diffusion tensor imaging (DTI) before lateral interbody fusion (LIF) to treat spinal deformities. METHODS DTI findings in a group of 12 patients (all women, mean age 74.3 years) with degenerative lumbar scoliosis (DLS) were compared with those obtained in a matched control group of 10 patients (all women, mean age 69.8 years) with low-back pain but without scoliosis. A T2-weighted sagittal view was fused to tractography from L3 to L5 and separated into 6 zones (zone A, zones 1-4, and zone P) comprising equal quarters of the anteroposterior diameters, and anterior and posterior to the vertebral body, to determine the distribution of nerves at various intervertebral levels (L3-4, L4-5, and L5-S1). To determine psoas morphology, the authors examined images for a rising psoas sign at the level of L4-5, and the ratio of the anteroposterior diameter (AP) to the lateral diameter (lat), or AP/lat ratio, was calculated. They assessed the relationship between apical vertebrae, psoas major morphology, and the course of nerve tracts. RESULTS Although only 30% of patients in the control group showed a rising psoas sign, it was present in 100% of those in the DLS group. The psoas major was significantly extended on the concave side (AP/lat ratio: 2.1 concave side, 1.2 convex side). In 75% of patients in the DLS group, the apex of the curve was at L2 or higher (upper apex) and the psoas major was extended on the concave side. In the remaining 25%, the apex was at L3 or lower (lower apex) and the psoas major was extended on the convex side. Significant anterior shifts of lumbar nerves compared with controls were noted at each intervertebral level in patients with DLS. Nerves on the extended side of the psoas major were significantly shifted anteriorly. Nerve pathways on the convex side of the scoliotic curve were shifted posteriorly. CONCLUSIONS A significant anterior shift of lumbar nerves was noted at all intervertebral levels in patients with DLS in comparison with findings in controls. On the convex side, the nerves showed a posterior shift. In LIF, a convex approach is relatively safer than an approach from the concave side. Lumbar nerve course tracking with DTI is useful for assessing patients with DLS before LIF.