Research Fellow Rothman Orthopaedic Institute Rothman Orthopaedic Institute
Disclosure(s):
Jarod Olson, BS: No financial relationships to disclose
Introduction: Prior literature has identified Hounsfield units(HU), from computed tomography(CT) scans, as correlated with bone quality. HU categories have been proposed in relation to postoperative outcomes, including >160HU, 110-160HU, and < 110HU, with the latter being associated with increased risk of postoperative proximal junction kyphosis(PJK). This study characterizes the impact of pre-operative HU measurements on short-term surgical outcomes after revision lumbar fusion surgery.
Methods: Patients were included if they were 18 years or older and underwent a lumbar revision fusion from 2017-2022 with an accessible pre-operative CT of the lumbar spine performed within three years of the date of surgery. A patient’s HU category was measured twice, once by averaging three axial CT slices in L1, and again in the vertebrae cephalad to the upper instrumented vertebrae(UIV). These three slices were oriented in the cancellous bone just caudal to the superior endpoint, at the midpoint of the vertebral body and just cephalad to the inferior endplate. ANOVA or Kruskal-Wallis tests were used for continuous data and Chi-squared testing for categorical data comparisons with alpha set to < 0.05.
Results: This study included 77 patients with average time to fusion of 150 days. No significant differences were noted between groups for ED visits, readmission rates, and reoperation or revision rates. When grouped by L1 measurements, the groups significantly differed in screw loosening(p=0.045) with 8% of patients experiencing loosening in the >160HU group followed by 25.8% and 38.1% in 110-160HU and < 110HU, respectively. When performing sub-analysis on surgical approach for screw loosening, patients with a circumferential fusion had loosening in the < 110HU group with no patients experiencing screw loosening in the 110-160 or >160HU groups(p=0.026).
Conclusion : Decreased L1 HUs are associated with pedicle screw loosening whereas measurements of the vertebrae cephalad to the UIV are not. Using L1, patients with a HU of < 110HU and 110-160HU appear to be at substantially increased risk of screw loosening and potential hardware failure, likely secondary to low bone mineral density. Second, a circumferential approach appears to be protective against screw loosening in patients who have >110HU in the L1 vertebrae.