Predicting osteoporosis-related complications in lumbar spine surgery using Hounsfield unit and vertebral bone quality scores: A 5-Year follow-up study with principal component analysis insights

Authors:
Publisher website
Google Scholar
Abstract
Objective Osteoporosis is a well-known risk factor for incident vertebral fractures (VFs) and postoperative mechanical complications. However, the specific Hounsfield unit (HU) and vertebral bone quality (VBQ) cutoff values for predicting these complications, as well as the diagnostic and clinical significance of the differences between these two measurements, remain unclear. This study sought to evaluate the predictive value of HU and VBQ scores and to establish specific cutoff points for osteoporosis-related complications in lumbar spine surgery. Additionally, we analyzed whether combining HU and VBQ scores enhanced their predictive usefulness and whether the differences between them were clinically significant using principal component analysis (PCA). Methods We examined 189 patients with lumbar spinal stenosis who underwent microscopic bilateral decompression using a unilateral approach or short segmental fixation between 2010 and 2016. Patients who underwent prior lumbar surgery were excluded, and a minimum 5-year follow-up was conducted. Surgical indications were based on preoperative evaluation of segmental instability. HU values and VBQ scores were measured using computed tomography and magnetic resonance imaging, respectively. Osteoporosis-related complications, including incident VFs, and mechanical complications, such as proximal junctional fractures, cage subsidence, and screw loosening, were assessed. Results For predicting osteoporosis-related complications, cutoff points of HU < 95.5 and VBQ > 3.49 were identified, both demonstrating good predictive accuracy without a significant difference in the area under the curve. Multivariate logistic regression confirmed low HU and high VBQ as independent risk factors (adjusted odds ratio = 5.57 and 3.42, respectively; p < 0.001). Additionally, PCA revealed that the concordance of these scores significantly distinguished patients with complications from those without complications (odds ratio, 3.77; p < 0.001), while the differences between these scores lacked clinical significance. Conclusion This study demonstrated that both HU and VBQ scores effectively predicted postoperative osteoporosis-related complications, with no significant difference in predictive ability. Optimal cutoff values were identified at approximately ≤ 100 for HU and ≥ 3.5 for VBQ, and the difference between these two scores lacked clinical significance. These cut-offs for HU and VBQ can serve as practical thresholds for risk assessment, enabling spinal surgeons to use these scores as opportunistic tools for identifying patients at an elevated risk of osteoporosis-related complications after lumbar spine surgery.

This publication has 25 references indexed in Scilit: