Single-slice computed tomography-derived pedicle screw trajectory Hounsfield unit values predict screw loosening after lumbar fusion: a retrospective study in Japan.

Yamamoto, Shuji; Komatsubara, Satoshi; Fujiki, Takaaki; Mitamura, Katsuya; Ishikawa, Masakazu · Asian Spine J · 2025

retrospective_cohort · Level III

Where this comes from

Abstract

Retrospective observational study. This study aimed to determine the predictive utility of Hounsfield unit (HU) values, derived from single-slice computed tomography (CT) images, for identifying postoperative pedicle screw (PS) loosening in patients undergoing lumbar interbody fusion. The HU values were specifically measured along the planned PS trajectory. PS loosening is a frequent complication following spinal fusion surgery, particularly in individuals with compromised bone quality. Although prior research has explored the correlation between screw loosening and general vertebral HU values or bone mineral density, the specific predictive capability of a simple, single-slice measurement along the screw path remains a gap in the literature. We retrospectively reviewed data from 47 patients (212 screws) who underwent lumbar interbody fusion between November 2020 and September 2022. Preoperative CT images were used to measure HU values in both the vertebral body and along the PS trajectories. Patients were then stratified into two groups based on the presence (L group, n=9) or absence (no-loosening [NL] group, n=38) of screw loosening identified on 1-year follow-up CT. Comparative analyses were conducted, followed by a receiver operating characteristic curve analysis to establish the optimal predictive HU cutoff. The incidence of postoperative screw loosening was 11.8% (25/212). While no significant inter-group differences were observed regarding age, bone mineral density, or vertebral body HU values, the PS trajectory HU was significantly lower in the L group (111±40) compared to the NL group (157±74) (p <0.001). Receiver operating characteristic curve analysis identified an optimal HU cutoff of 123. The area under the curve was 0.7106, with an odds ratio of 3.18, sensitivity of 0.31, and specificity of 0.41. Lower HU values measured along the PS trajectory using single-slice CT images were significantly associated with a higher risk of screw loosening. This straightforward method may serve as a valuable tool for identifying high-risk patients during the preoperative planning phase of spinal fusion surgery.

Anatomy