Crus Atrophy: Accuracy of Computed Tomography in Diagnosis of Diaphragmatic Paralysis.

Abstract

The aim of this study was to measure the association between crus atrophy as depicted by computed tomography (CT) and fluoroscopic diagnosis of hemidiaphragmatic paralysis in patients with suspected diaphragmatic dysfunction. A retrospective review of patient data was approved by our institutional review board and was HIPPA-compliant. We reviewed 90 patients who had undergone diaphragmatic fluoroscopy; 72 had CT scans available for measurement of crus thickness at the levels of the celiac and superior mesenteric arteries and the L1 vertebral body. Receiver operating characteristic analysis was used to determine the threshold of crus thinning that best distinguished a paralyzed hemidiaphragm from a nonparalyzed one. Hemidiaphragmatic paralysis caused significant crus thinning at the celiac artery level (mean±SD, 1.7±0.6 vs. 3.6±1.3 mm, P=0.017, on the right; 1.1±0.4 vs. 3.0±1.4 mm, P=0.001, on the left) and the L1 vertebral level (mean±SD, 1.5±0.7 vs. 4.4±1.6 mm, P=0.018, on the right; 1.5±0.6 vs. 3.6+1.7 mm, P=0.017, on the left). On axial CT, thinning to ≤2.5 mm at the celiac artery level identified paralysis of the hemidiaphragm with a sensitivity of 100% and a specificity of 86% on the right and with a sensitivity of 100% and a specificity of 64% on the left. On coronal CT, thinning to ≤2.5 mm at the L1 vertebral level identified paralysis of the hemidiaphragm with a sensitivity of 100% and a specificity of 88% on the right and with a sensitivity of 100% and a specificity of 77% on the left. Atrophy of the crus assessed by CT is a good discriminator of paralyzed versus nonparalyzed hemidiaphragm in patients with suspected diaphragmatic dysfunction.