摘要
目的探讨基于MRI和CT评估患者椎体骨密度进而预测成人退变性脊柱侧凸术后近端交界性后凸的诊断效能。方法回顾性收集2010年1月至2020年12月于四川大学华西医院骨科行长节段固定融合手术且具有完整术前全脊柱X线片、CT、MRI等影像学资料的成人退变性脊柱侧凸患者68例,男16例、女52例,年龄(66.87±6.65)岁(范围54~80岁)。采用术前腰椎MRI T1WI进行骨质量评分(vertebral bone quality score,VBQ)、采用CT测量L_(1)椎体亨氏单位(Hounsfield unit,HU),并根据术后影像学评估是否发生近端交界性后凸(proximal junctional kyphosis,PJK)分为PJK组和非PJK组。比较两组年龄、性别、体质指数、合并症、腰椎VBQ值、L_(1)HU值,对比两组手术前后的影像学参数(包括骨盆入射角、腰椎前凸角、矢状面轴向距离、骨盆倾斜角、骶骨倾斜角、骨盆入射角-腰椎前凸角、T1骨盆角、上端固定椎螺钉朝向),分别绘制受试者工作特征曲线(receiver operating characteristic curve,ROC)以分析腰椎VBQ值、L_(1)HU值及二者联合指标的诊断效能、灵敏度和特异度。结果PJK组17例、非PJK组51例。术前PJK组的VBQ值、L_(1)HU值分别为3.10±0.43、99.76±16.34,非PJK组分别为2.62±0.37、115.27±13.46,差异均有统计学意义(t=3.896,P<0.001;t=4.482,P<0.001)。VBQ值和L_(1)HU值的ROC曲线下面积(area under curve,AUC)分别为0.773[95%CI(0.633,0.914)]和0.814[95%CI(0.680,0.949)]、灵敏度分别为82.4%和76.5%、特异度分别为70.6%和76.5%、最大约登指数和临界值分别为0.530、2.895和0.530,111.0,二者联合指标AUC为0.895[95%CI(0.795,0.995)],灵敏度、特异度分别为94.1%、82.4%,最大约登指数和临界值为0.765,0.734。结论VBQ值和L_(1)HU值均可有效预测退变性脊柱侧凸术后近端交界性后凸,L_(1)HU值准确率略高于VBQ值,二者联合诊断准确率最高。
Objective To explore the measurement method and its diagnostic performance based on MRI and CT measurement of vertebral bone density in patients to predict proximal junctional kyphosis after degenerative scoliosis surgery.Methods Retrospectively included patients who underwent long-segment fixation and fusion surgery at the Department of Orthopedics,West China Hospital of Sichuan University from January 2010 to December 2020 and had complete preoperative whole-spine X-rays,CT,MRI and other imaging examination results,and were followed up on schedule.68 cases of adult degenerative scoliosis,16 male,52 women,aged 66.87±6.65 years(range,54-80 years).The patients were measured based on preoperative lumbar spine MRI T1WI bone quality score(vertebral bone quality score,VBQ)and CT-based Hounsfield(HU)value,and the patients were divided into postoperative proximal junction kyphosis group or non-proximal junction kyphosis group based on the results of postoperative imaging examinations.The age,gender,BMI,comorbidities,lumbar spine VBQ score,L_(1) CT HU value and various imaging parameters before and after surgery were compared between the two groups of patients,including pelvic incident,lumbar lordosis,sagittal vertical axis,pelvic tilt,sacral slope,pelvic incidence-lumbar lordosis,T1 pelvic angle,Upper instrumented vertebrae screw angle,etc.The receiver operating characteristic curve(ROC)was drawn to analyze the diagnostic efficacy,sensitivity and specificity of VBQ score,HU value and their combined indicators.Results Seventeen cases were included in the PJK group and 51 cases in the non-PJK group.The preoperative VBQ and HU values of the PJK group were 3.10±0.43 and 99.76±16.34 respectively,while those of the non-PJK group were 2.62±0.37 and 115.27±13.46 respectively.The differences were statistically significant(t=3.896,P<0.001;t=4.482,P<0.001).The area under curve(AUC)of VBQ was 0.773[95%CI(0.633,0.914)],the sensitivity and specificity are 82.4%and 70.6%respectively,the AUC of HU value was 0.814[95%CI(0.680,0.949)],the sensitivity and specificity are 76.5%and 76.5%respectively.The AUC of the two combined indicators was 0.895[95%CI(0.795,0.995)],and the sensitivity and specificity were 94.1%and 82.4%respectively.The maximum Youden index and the critical value were respectively,VBQ value 0.530,2.895,HU value 0.530,110.0,the combined index 0.765,0.734.Conclusion Both VBQ and L_(1) HU value can accurately predict proximal junctional kyphosis after degenerative scoliosis surgery.The accuracy of HU value was slightly higher than that of VBQ value.The diagnosis accuracy of the combined index was the highest.
作者
王哲
陈骞
黄勇
吴锐邦
赵一泽
刘立岷
宋跃明
丰干钧
Wang Zhe;Chen Qian;Huang Yong;Wu Ruibang;Zhao Yize;Liu Limin;Song Yueming;Feng Ganjun(Department of Orthopedic Surgery and Orthopedic Research Institute,West China Hospital,Sichuan University,Chengdu 610041,China)
出处
《中华骨科杂志》
CAS
CSCD
北大核心
2024年第11期748-755,共8页
Chinese Journal of Orthopaedics
关键词
脊柱侧凸
骨密度
磁共振成像
体层摄影术
近端交界性后凸
Scoliosis
Bone density
Magnetic resonance imaging
Tomography
Proximal junctional kyphosis