Comparison of predictive value for cage subsidence between MRI-based endplate bone quality and vertebral bone quality scores following transforaminal lumbar interbody fusion: a retrospective propensity-matched study

Background Context Cage subsidence after lumbar fusion can lead to many adverse outcomes. Low bone mineral density (BMD) is a widely recognized risk factor for cage subsidence. Conventional methods can predict and evaluate BMD, but there are many shortcomings. Recently, MRI-based assessment of bone quality in specific parts of the vertebral body has been proposed, including scores for vertebral bone quality (VBQ) and endplate bone quality (EBQ). However, the predictive accuracy of the two scoring systems for cage subsidence after transforaminal lumbar interbody fusion (TLIF) remains unknown. Therefore, we investigated MRI-based VBQ and EBQ scores for assessing bone quality and compared their predictive value for cage subsidence after TLIF. Purpose To compare the predictive value between MRI-based VBQ and EBQ scores for cage subsidence after TLIF. Study Design/Setting A retrospective case-control study Patients Sample Patients with degenerative lumbar diseases underwent single-level TLIF at our medical center between 2014 and 2020, all of whom had preoperative MRIs available. Outcomes Measures Cage subsidence, disc height, VBQ score, EBQ score, upper and lower vertebral body bone quality (UL-VBQ) score. Methods Data were retrospectively examined for a consecutive sample of 346 patients who underwent TLIF at our medical center between 2014 and 2020. Patients who subsequently experienced cage subsidence or not were matched to each other based on propensity scoring, and the two matched groups (52 patients each) were compared using conditional logistic regression to investigate the association between the potential radiographic factors and cage subsidence. Scores for VBQ and EBQ were assessed for their ability to predict cage subsidence in the matched patients based on the area under the receiver operative characteristic curve (AUC). Results Among matched patients, those who suffered cage subsidence had significantly higher VBQ score (3.7 vs 3.1, p<0.001) and EBQ score (5.0 vs 4.3, p<0.001), and regression linked greater risk of subsidence to higher VBQ score (OR 4.557, 95% CI 1.076-19.291, p=0.039) and higher EBQ score (OR 5.396, 95% CI 1.158-25.146, p=0.032). A cut-off VBQ score of 3.4 predicted the cage subsidence among matched patients with an AUC of 0.799, sensitivity of 84.6% and specificity of 69.2%. A cut-off EBQ score of 4.7 predicted subsidence with an AUC of 0.829, sensitivity of 76.9% and specificity of 82.7%. Conclusion Higher VBQ and EBQ scores are associated with greater risk of cage subsidence following TLIF, and EBQ may perform better because of greater specificity.