MRI-based endplate bone quality score independently predicts cage subsidence after anterior cervical corpectomy fusion.

Abstract

To reduce the amount of radiation that patients receive during surgery, surgeons can evaluate the quality of the bone prior to surgery using computed tomography (CT) or dual-energy X-ray absorptiometry. Recently, lumbar spine vertebral bone quality has been evaluated using an MRI-based scoring system. However, few studies have investigated the connection between cage subsidence in patients following cervical interbody fusion and site-specific MRI bone evaluation. It is unknown how cage subsidence following anterior cervical corpectomy decompression and fusion is related to MRI-based endplate bone quality assessment. To create a similar MRI-based cervical spine scoring system (C-EBQ) and to investigate the predictive value of the MRI cervical endplate bone quality (C-EBQ) score for cage subsidence after ACCF. The patients' demographic, surgical, and radiological data were collected. Cage subsidence was defined as fusion segment height loss ≥ 3 mm. Multivariate logistic regression models were developed to determine correlations between potential risk factors and subsidence, and simple linear regression analyses of statistically significant indicators were performed. Among the patients who underwent single-level ACCF, 72 met the requirements for inclusion. The C-VBQ scores also improved from 2.28 ± 0.12, indicating no subsidence, to 3.27 ± 0.35, which indicated subsidence, and the C-EBQ scores improved in both the nonsubsidence group (1.95 ± 0.80) and the subsidence group (2.38 ± 0.54). There was a statistically significant difference (p < 0.05) among the groups. Higher C-EBQ scores were strongly correlated with subsidence in the multivariate analysis (odds ratio [OR] = 17.249, 95% CI = 2.269 to 7.537, P < 0.001), and the C-VBQ score was the major independent predictor of subsidence following ACCF ([OR] = 4.752, 95% CI = 3.824 to 8.781, P < 0.05). The C-EBQ score outperformed the C-VBQ score (75.6%) in terms of predictive accuracy, with a ROC curve indicating an 89.4% score. After ACCF, cage subsidence was strongly correlated with higher C-EBQ scores on preoperative MRI. Assessing C-EBQ before ACCF may be a useful way to estimate the likelihood of postoperative subsidence.