Abstract
BACKGROUND: Numerous complications of lumbar fusion surgery have been reported, with adjacent segment disease (ASD) being one of the most important. Few studies describe the effect of sagittal, horizontal screw angles and distance of pedicle screw apex to superior endplate on the incidence of ASD in lumbar spine. The purpose of this retrospective study is to evaluate the hypothesis that unsatisfactory pedicle screw insertion positions would increase the likelihood of ASD. METHODS: Outpatients with lumbar spinal stenosis underwent posterolateral lumbar fusion at L4-S1 with a least 2-year follow-up were studied. ASD at L3-L4 was defined as a condition in which intervertebral disk narrowing, posterior vertebral opening, and vertebral slippage progress at the last follow-up in comparison with the postoperative. Independent t test was performed to compare data between two groups; Spearman analysis was performed to analyze the relationship between two continuous variables. Multivariate binary logistic models were performed to identify the independent risk factors of ASD. The receiver operating characteristic (ROC) curve was performed to measure model discrimination and Hosmer-Lemeshow (H-L) test was used to measure calibration. ROC curve evaluated the discrimination ability of sagittal screw angle and distance in predicting incidence of ASD. RESULTS: Patients in ASD group exhibit higher incidence of osteoporosis, higher Visual analogue scale (VAS), Oswestry disability index (ODI), bigger sagittal screw angle, shorter distance of pedicle screw apex to superior endplate than those in non-ASD group (p < 0.05). VAS, ODI at the last follow-up were positively correlated with Pfirrmann grade of L3-4 disk and sagittal screw angle, while negatively correlated with distance of screw apex to superior endplate (p < 0.05). Multivariate binary logistic model indicated that follow-up time (odds ratio [OR] 1.637, 95% confidence interval [CI] 1.186-2.260), distance of screw apex to superior endplate (OR 0.150, 95% CI 0.067-0.336), sagittal screw angle (OR 2.404, 95% CI 1.608-3.594) were statistically significant. The models showed great discrimination and calibration. The area under the curve of ASD identified by sagittal angle and distance was 0.895 and the cut-off values were 5.500° and 6.250 mm, respectively. CONCLUSION: Sagittal screw angle and distance of screw apex to superior endplate were significantly associated with the risk of ASD.