The Role of Hounsfield Units in Predicting Cage Subsidence After Lateral Lumbar Interbody Fusion: A Systematic Review and Meta-Analysis.

Abstract

This systematic review and meta-analysis examined cohort studies investigating the relationship between vertebral Hounsfield unit (HU) measurements on preoperative computed tomography scans and postoperative cage subsidence (CS) in patients undergoing lateral lumbar interbody fusion (LLIF). As dual-energy X-ray absorptiometry scans may inaccurately reflect local bone quality, vertebral HU of computed tomography scans have emerged as a viable alternative to predict CS. This study aimed to identify a potential HU threshold predictive of CS in LLIF for future studies to explore and verify. Four databases were searched for studies reporting CS rates and preoperative segmental HU in patients undergoing LLIF, including oblique lateral interbody fusion, extreme lateral interbody fusion, anterior-to-psoas, and transpsoas techniques for degenerative spinal conditions. Random-effects meta-analyses were conducted. Outcomes were pooled segmental HU values and the predictive performance of different HU thresholds for CS. Intrastudy risk of bias was rated using the Newcastle-Ottawa scale, and evidence certainty was assessed using Grading of Recommendations, Assessment, Development and Evaluations. Five retrospective cohort studies totaling 396 patients (271 non-CS, 125 CS) were included. The pooled segmental HU was 112 (95% confidence interval [CI]: 96-128) for patients with CS and 143 (95%CI: 128-158) for non-CS patients. The pooled mean difference was 30 (95%CI: 14-45, P < 0.01,p_egger=0.51). A threshold of ≤120 HU had the best predictive performance for CS (sensitivity: 76%, 95%CI: 32%-96%, specificity: 88%, 95%CI: 75%-95%, area under the curve: 0.903). This study affirms HU as a potential predictive tool for CS in patients undergoing LLIF for degenerative spinal diseases. Future studies should explore the cut-off of 120 HU to validate its predictive capability.