Shen Canghai, Xu Haiwei, Xu Baoshan, Zhang Xiaoguang, Li Xiaofeng, Yang Qiang, Ma Xinlong
Tianjin Medical University, Tianjin, China.
Department of Spine Surgery, Tianjin Hospital, Tianjin, China.
J Back Musculoskelet Rehabil. 2018;31(3):525-532. doi: 10.3233/BMR-170972.
In addition to conventional magnetic resonance imaging (cMRI), diffusion tensor imaging (DTI) has been investigated as a potential diagnostic and prognostic tool for patients with degenerative cervical myelopathy (DCM).
To assess the efficacy of cMRI and DTI parameters in prediction of surgical outcome in DCM patients.
One hundred and forty-two patients with DCM who underwent presurgical cMRI and DTI of the cervical spine were included. Quantitative parameters obtained by cMRI included compression ratio (CR), transverse area (TA), and signal intensity ratio (SIR). DTI was evaluated for apparent diffusion coefficient (ADC) and fractional anisotropy (FA). The Japanese Orthopaedic Association (JOA) score and recovery rate were used to evaluate clinical outcomes. A JOA recovery rate < 50% was defined as a poor surgical outcome. The relationship of surgical outcome with various imaging parameters was examined. Receiver operating characteristic (ROC) curves were used to measure the predictive ability and determine the best cut-off values of the quantitative parameters.
By ROC curve analyses of imaging parameters, the largest area under the ROC curve (AUC) was for FA (0.750), followed by ADC (0.719), TA (0.716), SIR (0.673), and CR (0.591). The cut-off values with the best compromise between sensitivity and specificity were set at 0.390 for FA, 1.344 × 10-3 mm2/s for ADC, 46.02 mm2 for TA, 1.556 for SIR, and 26.56% for CR. Multivariate logistic regression model revealed that JOA score ⩽ 8 points, TA ⩽ 46.02 mm2, and FA ⩽ 0.390 were independently associated with poor surgical outcome. The AUC value for the three-predictor model was 0.871, indicating strong predictive discrimination, and was significantly higher than the AUC value for the model containing only the JOA score (0.763; P= 0.003).
JOA score is a reasonable predictor of surgical outcome in DCM. However, a model inclusive of TA and FA provides superior predictive ability. Thus, quantitative analysis of cMRI and DTI is useful for predicting surgical outcome in DCM.
除传统磁共振成像（cMRI）外，弥散张量成像（DTI）已被研究作为退行性颈椎病（DCM）患者潜在的诊断和预后工具。
评估cMRI和DTI参数对DCM患者手术结果的预测效能。
纳入142例接受颈椎术前cMRI和DTI检查的DCM患者。cMRI获得的定量参数包括压迫率（CR）、横截面积（TA）和信号强度比（SIR）。评估DTI的表观扩散系数（ADC）和分数各向异性（FA）。采用日本骨科协会（JOA）评分和恢复率评估临床结果。JOA恢复率<50%被定义为手术结果不佳。研究手术结果与各种影像参数的关系。采用受试者工作特征（ROC）曲线测量预测能力并确定定量参数的最佳截断值。
通过对影像参数的ROC曲线分析，ROC曲线下最大面积（AUC）为FA（0.750），其次是ADC（0.719）、TA（0.716）、SIR（0.673）和CR（0.591）。在敏感性和特异性之间取得最佳折衷的截断值设定为：FA为0.390，ADC为1.344×10⁻³mm²/s，TA为46.02mm²，SIR为1.556，CR为26.56%。多因素逻辑回归模型显示，JOA评分≤8分、TA≤46.02mm²和FA≤0.390与手术结果不佳独立相关。三预测因子模型的AUC值为0.871，表明具有较强的预测判别能力，且显著高于仅包含JOA评分的模型的AUC值（0.763；P=0.003）。
JOA评分是DCM患者手术结果的合理预测指标。然而，包含TA和FA的模型具有更好的预测能力。因此，cMRI和DTI的定量分析有助于预测DCM患者的手术结果。
