Abstract
BACKGROUND CONTEXT: As lumbar degenerative diseases become more prevalent in an aging population, there is an increasing demand for surgical interventions, such as posterior lumbar interbody fusion (PLIF). However, cage subsidence (CS), observed in 23.9-54% of cases postoperatively, remains a significant complication. Several factors, including age, bone quality, and endplate damage, contribute to the risk of CS, with bone quality being among the most critical determinants. Although DEXA and QCT are widely employed to assess bone density, their routine use in preoperative evaluations is restricted by cost considerations and radiation exposure. Recent studies suggest that MRI-based vertebral body quality (VBQ) and endplate bone quality (EBQ) score offer a viable, non-invasive alternative for evaluating bone quality; however, there is limited research comparing their predictive value for CS. METHODS: In this retrospective study, 165 patients undergoing single-level PLIF surgery were included. MRI-based VBQ and EBQ score were calculated using T1-weighted images, and preoperative QCT was employed as a clinical standard. Cage subsidence was assessed based on postoperative imaging at 12-month follow-up. Statistical analyses, including t-tests, chi-square tests, and ROC curve analyses, were used to evaluate the predictive accuracy of VBQ and EBQ for CS. RESULTS: The study's findings demonstrated that both VBQ and EBQ scores were significantly correlated with QCT measurements, thereby validating their utility as indicators of bone quality. ROC analysis revealed that VBQ had superior predictive value for CS (AUC = 0.814) compared to EBQ (AUC = 0.719), with both scores demonstrating significant clinical utility in identifying patients at risk for CS. Notably, VBQ exhibited a stronger correlation with preoperative clinical outcomes compared to EBQ, underscoring its greater reliability as a predictor. CONCLUSIONS: This study highlights the effectiveness of MRI-based VBQ and EBQ score as practical, non-invasive tools for assessing bone quality preoperatively, with VBQ demonstrating superior predictive performance for CS risk. The findings underscore the potential of integrating these MRI-based assessments into routine preoperative planning to improve patient outcomes and minimize complications associated with PLIF surgery.