Computed tomography Hounsfield unit–based prediction of pedicle screw loosening after surgery for degenerative lumbar spine disease

OBJECTIVE The authors investigated the relation between Hounsfield unit (HU) values measured on CT and the risk of pedicle screw loosening in patients who underwent lumbar pedicle screw fixation for degenerative lumbar spine disease. METHODS Patients who were treated with lumbar pedicle screw fixation between July 2011 and December 2015 at the authors’ department were reviewed. Age, sex, BMI, smoking and diabetes histories, range of fixation, and fusion method were recorded as the basic patient information. The HU values for lumbar bone mineral density (BMD) for the L1, L2, L3, and L4 vertebra were measured on CT scans. Logistic regression analysis was used to identify the independent influencing factors of pedicle screw loosening. RESULTS A total of 503 patients were included in the final analysis. The pedicle screw loosening rate at the 12-month follow-up was 30.0% (151 of 503 patients). There were no significant differences in sex, BMI, or histories of smoking and diabetes between the patients with (loosening group) and those without (nonloosening group) screw loosening (p > 0.05). The mean HU value of L1–4 was lower in the loosening group than the nonloosening group (106.3 ± 33.9 vs 132.6 ± 42.9, p < 0.001). In logistic regression analysis, being male (OR 2.065; 95% CI 1.242–3.433), HU value (OR 0.977; 95% CI 0.970–0.985), length of fixation (OR 3.616; 95% CI 2.617–4.996), and fixation to S1 (OR 1.699; 95% CI 1.039–2.777) were the independent influencing factors for screw loosening. CONCLUSIONS HU value measured on CT was an independent predictor for pedicle screw loosening, and lower HU value was significantly correlated with higher risk of screw loosening.