The Association Between Hounsfield Units and Mechanical Failure in ASD Patients

• Published in: Global spine journal 2024-10, p.21925682241291519
• Main Authors: Heegaard, Martin, Ingadottir, Kristin, Ragborg, Lærke, Dahl, Benny, Hansen, Lars Valentin, Ohrt-Nissen, Søren, Gehrchen, Martin
• Format: Article
• Language: English
• Subjects: Original Research
• ISSN: 2192-5682; 2192-5690
• DOI: 10.1177/21925682241291519

Retrospective Cohort Study. Low bone mineral density (BMD) is a known risk factor for revision surgery in patients with adult spinal deformity (ASD). Hounsfield units (HUs) on CT scans have been suggested as a proxy for assessing BMD. This study aimed to determine HUs in the lumbar region and their association with mechanical failure in patients undergoing ASD surgery. We included ASD patients undergoing surgery from 2010-2020 with minimum 2-year follow-up. We excluded patients without preoperative CT scans, or a CT scan more than 1 year before surgery. Mechanical failure was defined as proximal junctional failure, pseudarthrosis, or implant failure requiring revision surgery. On preoperative CT scans, HUs were measured on 3 axial slices on each vertebra from L1-L5 and, if available, at UIV and UIV + 1. We included 170 patients, mean age 63 (±12) years, with 108 (64%) females, and 13 [IQR 10-16] instrumented levels. Mechanical failure occurred in 27% (n = 46) of patients at 2-year follow-up. Mean lumbar HUs were 146 (±51) in the mechanical failure group and 135 (±52) in those without revision ( = .232). Area under the curve was 0.58 (95% CI: 0.48-0.68), corresponding to no to low discriminatory power in predicting mechanical failure using lumbar HUs. Univariate logistic regression revealed no significant difference between mechanical failure and lumbar HUs (OR = 1.00, 95% CI: 1.00-1.01, = .239). We found no association between mechanical failure and HUs on preoperative CT scans in ASD patients. Thus, we cannot recommend using HUs to predict mechanical failure in these patients.