Stein Jill, Laor Tal, Carr Preston, Zbojniewicz Andrew, Cornwall Roger
Department of Radiology, Cincinnati Children's Hospital Medical Center, Cincinnati, OH.
Division of Orthopaedic Surgery, Cincinnati Children's Hospital Medical Center, Cincinnati, OH.
J Hand Surg Am. 2017 Dec;42(12):1030.e1-1030.e11. doi: 10.1016/j.jhsa.2017.07.001. Epub 2017 Aug 16.
Neonatal brachial plexus palsy (NBPP) frequently causes glenohumeral dysplasia. Quantification of this dysplasia on magnetic resonance imaging can determine the need for and the success of nonsurgical or surgical intervention. However, we hypothesize that the variable position of the scapula on the thorax between affected and unaffected shoulders affects dysplasia measurements.
Magnetic resonance imaging studies were analyzed from 19 NBPP patients (ages 0.8-18 years; median, 2.4 years) without prior shoulder surgery. Three reviewers measured the glenoid version angle (GVA) and percentage of humeral head anterior to the midscapular line (PHHA) on standard axial images ("thoracic axial") and on reformatted axial images aligned perpendicular to the scapular plane ("scapular axial"), which corrects for scapulothoracic position. Scapular tilt and protraction were measured to assess their impact on the difference between thoracic and scapular GVA and PHHA measurements. Intra- and interrater reliability were calculated for GVA and PHHA on both views.
The GVA of the affected shoulder was significantly greater on thoracic than on scapular images, by an average of 5° and as much as 34°. The PHHA was significantly less in the affected shoulders on thoracic than on scapular images, by an average of 5% and as much as 33% of humeral head width. The difference in GVA, but not PHHA, between thoracic and scapular axial images in the affected shoulder correlated with scapular tilt. Unaffected shoulders showed no significant difference in GVA or PHHA between thoracic and scapular axial images. Interrater reliability ranged from fair to substantial and did not differ between thoracic and scapular images.
Thoracic axial images overestimate the severity of glenohumeral dysplasia in NBPP, owing at least in part to the variable position of the scapula on the thorax. This confounding effect must be considered in interpretation of axial quantitative measures of glenohumeral dysplasia in NBPP.
TYPE OF STUDY/LEVEL OF EVIDENCE: Diagnostic III.
新生儿臂丛神经麻痹（NBPP）常导致盂肱关节发育不良。通过磁共振成像对这种发育不良进行量化可确定非手术或手术干预的必要性及成功率。然而，我们推测患侧与未患侧肩部之间肩胛骨在胸廓上的位置变化会影响发育不良的测量结果。
对19例未接受过肩部手术的NBPP患者（年龄0.8 - 18岁；中位数2.4岁）的磁共振成像研究进行分析。三位评估者在标准轴位图像（“胸廓轴位”）以及垂直于肩胛骨平面重新格式化的轴位图像（“肩胛骨轴位”）上测量盂肱关节面倾斜角（GVA）和肱骨头位于肩胛中线前方的百分比（PHHA），后者可校正肩胛胸廓位置。测量肩胛骨倾斜度和前伸度，以评估它们对胸廓和肩胛骨GVA及PHHA测量差异的影响。计算两种视图下GVA和PHHA的评估者内及评估者间可靠性。
患侧肩部在胸廓图像上的GVA显著大于在肩胛骨图像上的GVA，平均大5°，最大相差34°。患侧肩部在胸廓图像上的PHHA显著小于在肩胛骨图像上的PHHA，平均少5%，最多相差肱骨头宽度的33%。患侧肩部胸廓轴位图像与肩胛骨轴位图像之间GVA的差异（而非PHHA的差异）与肩胛骨倾斜度相关。未患侧肩部胸廓轴位图像与肩胛骨轴位图像之间的GVA或PHHA无显著差异。评估者间可靠性从中等到高度，胸廓图像和肩胛骨图像之间无差异。
胸廓轴位图像高估了NBPP中盂肱关节发育不良的严重程度，至少部分原因是肩胛骨在胸廓上的位置变化。在解释NBPP中盂肱关节发育不良的轴位定量测量结果时，必须考虑这种混杂效应。
研究类型/证据水平：诊断性研究III级
