How Lumbar Disc Geometrics Affect the Risk for Rod Fracture in Patients Undergoing Adult Spinal Deformity Surgery: An Initial Report

Abstract

Introduction Rod fracture (RF) may have significant consequences for patients, including pain, loss of deformity correction, and the need for revision surgery. Risk factors associated with RF includes age, previous spine surgery, insufficient sagittal plane correction, rod material and bending, and pedicle subtraction osteotomy (PSO). Lumbar disc geometrics, regardless of PSO, may be a risk factor affecting RF rates. Out Objective is to evaluate disc geometrics as a risk factor for RF in patients undergoing surgical correction for adult spinal deformity (ASD). Materials/Methods: A retrospective review of a single surgeon's practice from 2010 to 2014 was performed. All patients who underwent long fusion constructs for adult deformity with a minimum of 2-year follow-up with radiographs were included. Patients were divided into 2 Groups: Group 1 (RF) - patients who sustained a rod fracture, Group 2 - patients without RF. Results Thirty-seven of 41 patients met inclusion criteria. Group 1 - included 11 patients (30%) with RF, all requiring revision surgery. Group 2 - included 26 patients (70%) who did not have evidence of RF at 2-year follow-up. There was no significant difference between the 2 Groups regarding age, weight, construct length (12 versus 11 levels), apical lordosis level (L3), pre or post-operative lumbar lordosis, or mean number of non-fused lumbar discs (3). Twenty-one of the 37 patients (54%) had a PSO performed. There was no difference in the prevalence of PSO between the 2 Groups: 5/11 patients (45%) in Group 1 (RF) versus 17/26 patients (65%) in Group 2. Nineteen of the 37 patients (51%) had previous lumbar spinal surgery. Measurements comparing intervertebral disc geometry 2 levels immediately cranial and caudal to the PSO or the apical lumbar vertebra were noted to have a significant difference in the mean cranial disc height (9.5mm versus 6.5mm, p < 0.01), mean cranial diameter (40mm versus 34mm, p < 0.01), mean cranial volume (13284mm3 versus 8145mm3, p < 0.01), and mean caudal volume (10635mm3 versus 7771mm3, p = 0.02). Caudal disc heights and diameters were larger in the RF group, but did not reach statistical significance. Conclusions The overall prevalence of RF was 27% with no difference in PSO rate between the 2 Groups. Patients with RF had significantly larger non-fused disc heights, diameters and volumes. The increased disc geometrics may allow for increased micro-motion in the anterior column resulting in increased failure rates. Lumbar interbody support in those discs at risk may improve stability and decrease RF rates. Further data analysis and clinical correlation is required.