Computed tomography-derived Hounsfield unit mapping for pedicle screw planning in degenerative lumbar disease: a...

Pedicle screw loosening is a common and clinically significant complication of instrumented lumbar fusion and is reported in 15%-30% of screws in osteoporotic patients. Dual-energy X-ray absorptiometry (DEXA)-based bone mineral density (BMD) has long served as the standard preoperative assessment tool; however, its well-recognized limitations in the degenerative spine and its modest correlations with fixation strength have prompted growing interest in computed tomography (CT)-derived bone quality metrics as complementary alternatives. This study aims to synthesize the current evidence based on CT-derived Hounsfield unit (HU) measurements as preoperative predictors of pedicle screw insertional torque and postoperative screw loosening in degenerative lumbar fusion surgery, with a specific focus on trajectory-specific techniques. A narrative review of studies indexed in PubMed and MEDLINE (1990-2026) was conducted using the following search terms: 'Hounsfield units pedicle screw,' 'CT bone quality lumbar fusion,' 'screw loosening BMD,' 'trajectory HU insertional torque,' and 'vertebral bone quality score.' Studies were included if they reported quantitative bone quality metrics in adult patients undergoing lumbar instrumented fusion with outcomes of insertional torque or radiographic screw loosening. Non-human or cadaveric-only studies were excluded unless they were directly relevant to clinical threshold derivation. Multiple retrospective studies consistently demonstrated that trajectory-specific HU, measured within a cylindrical or single-slice region of interest matched to the planned screw trajectory, outperformed lumbar and femoral DEXA in correlating with insertional torque (r=0.75 vs. r=0.55-0.59) and predicting 12-month screw loosening. Three-dimensional trajectory HU and single-slice measurements both showed clinical utility, with proposed exploratory thresholds of 110-123 HU, although these values were derived from single-center, heterogeneous CT acquisition protocols and were not validated across institutions. The magnetic resonance imaging-based Vertebral Bone Quality score demonstrated comparable performance to vertebral body HU for overall loosening prediction, although trajectory-specific comparisons have not yet been published. The evidence supports an emerging framework for transitioning from systemic DEXA-based BMD to localized, trajectoryspecific CT-HU assessment as a complementary tool for preoperative pedicle screw planning. Prospective validation, standardization of measurement protocols, cross-scanner calibration, and integration into surgical navigation platforms are the critical next steps before clinical translation can be recommended.