123. A novel comprehensive model to describe vertebral body motion following interbody fusion

Abstract

<h3>BACKGROUND CONTEXT</h3> Subsidence of interbody fusion spinal implants occurs in up to 30% of interbody fusions. Previous methods to quantify subsidence of interbody implants have focused on quantifying changes in disc height or magnitude of erosion of the implant into the vertebral body. These studies focus on the description of subsidence in a cranial-caudal direction. A comprehensive description of subsidence should report cranial-caudal shift, anterior-posterior shift and angular shift of one vertebra with respect to its adjacent vertebra. <h3>PURPOSE</h3> The purpose of the study is to develop and validate a novel method that describes the motion of one vertebral body with respect to its adjacent vertebral body in terms of translation and rotation. We hypothesize this method would have satisfactory interobserver reliability. <h3>STUDY DESIGN/SETTING</h3> Retrospective review of lumbar spine radiographs at a tertiary academic center. <h3>PATIENT SAMPLE</h3> Patients undergoing anterior lumbar interbody fusion at L4-5 or L5-S1 with at least one year radiographic follow-up. <h3>OUTCOME MEASURES</h3> The amount of cranial-caudal shift, anterior-posterior shift and angular shift on preoperative, postoperative, and 1-year postoperative radiographs was recorded according to the novel measurement technique. <h3>Methods</h3> A retrospective review of lateral lumbar spine radiographs was performed by seven physicians (3 residents, 2 fellows, and 2 attendings). The review included radiographic analysis of 10 patients who underwent L4-5 or L5-S1 anterior interbody fusion. Preoperative, postoperative (day 1), and 12-month postoperative radiographs were analyzed. Each vertebra was defined in two-dimensional space by 4 peripheral points at the corners of the vertebral body and a centroid. The upper vertebra was designated as a reference plane by which the motion of the lower vertebra was measured. We measured the movement of the lower vertebra in terms of anterior-posterior translation, cranial caudal translation, and angular change immediately following lumbar interbody fusion and 12 months postoperatively to detect translational and rotational subsidence. Inter-observer reliability was measured using SPSS software. <h3>Results</h3> A model was successfully developed to describe the angular and translational change of one vertebral body with respect to its adjacent segment. The intraclass correlation coefficients of anterior-posterior, cranial-caudal, and angular change were calculated to be 0.75, 0.90, and 0.86, respectively. The average standard deviations of translational and rotational change were 3.3% and 2°, respectively. <h3>Conclusions</h3> A model to comprehensively describe vertebral body motion as a method of characterizing subsidence was successfully developed. The model had good to excellent reliability. Further refinement of the protocol may yield improved intraclass correlation coefficients. This method of measuring vertebral body motion is applicable to measuring modes of failure beyond subsidence such as fracture or delayed fusion. <h3>FDA DEVICE/DRUG STATUS</h3> This abstract does not discuss or include any applicable devices or drugs.