骨性影像学参数对胸腰椎骨折PLC损伤诊断效能的研究

Correlation Between Posterior Ligamentous Complex Injury in Thoracolumbar Fractures and Osteal Imaging Parameters

目的评估CT检查中骨性影像学参数(椎体楔变角、局部后凸角、椎体压缩率、棘突间距增加值和椎弓根间距增加值)对后方韧带复合体(PLC)损伤的诊断效能。方法回顾性分析2013年1月至2016年3月间外伤导致胸腰段(T_(11)~L_2)骨折的患者52例,所有患者进行CT平扫及多平面重组,测量椎体楔变角、局部后凸角、椎体压缩率、棘突间距增加值及椎弓根间距增加值等骨性参数,根据MRI结果,将PLC状态分为完整和损伤。先将骨性参数通过t检验进行筛选,将有统计学意义的骨性参数纳入二元Logistic回归模型分析单个参数与PLC损伤的相关性,使用受试者工作特性曲线(ROC)分析骨性影像学参数对PLC损伤的诊断效能。结果52例患者中,MRI显示23例出现PLC损伤。椎体楔变角(P=0.710)与局部后凸角(P=0.120)在无PLC损伤与PLC损伤组间无明显统计学差异。Logistic回归分析表明椎体压缩率与PLC损伤无明显相关(P=0.053)。棘突间距增加值(P=0.001)及椎弓根间距增加值(P=0.046)与PLC损伤存在相关。ROC分析显示:棘突间距增加值曲线下面积最大(AUC=0.807),其次是椎弓根间距增加值(AUC=0.666)。结论在各项骨性参数中,棘突间距增加值和椎弓根间距增加值是最有效的预测参数。当棘突间距增加值>2.2 mm,椎弓根间距增加值>1.0 mm时,两者联合诊断PLC损伤的敏感度为70%,特异度为93%。

Objective To evaluate the reliability of a group of parameters of CT in the diagnosis of PLC disruption. Methods Assessment of 52 cases of PLC disruption in thoraeolumbar fracture. Patients were separated into two groups by MRI findings of their PLC integrity. CT was then conducted to collect data of osteal parameters including SIEA, LK, LOVBH, ISD, and IPD. Significant variables, at P 〈 0.05, were chosen for further analysis. Parameters were then analyzed using Logistic Regression Model to evaluate their correlation with PLC disruption. An ROC curve was then utilized to calculate the approximate probability of PLC damage at different ranges of measurement values for each ofthe parameter studied. Results SIEA （P=0.71） and LK （P=O. 12）were found to have negative correlation with PLC integrity in ttest. LOVBH were also found to have a negative correlation in logistic regression （ P = 0. 053）. ISD and IPD were found to have a positive correlation in identification of PLC injury. In ROC curve analysis, the AUC of ISD was 0.807, AUC of ISD was 0. 666. Conclusion The ISD and IPD were considered to be significant predictors of PLC injury. ISD difference greater than 2.2 mm and IPD difference greater than 1.0 mm, or ISD difference greater than 1.2 mm and IPD difference greater than 2.7 mm are effective predictors of PLC injury; sensitivity was 70% and specificity was 93%.