Chest radiography for the diagnosis of acute aortic syndrome

Abstract Purpose We sought to assess the diagnostic accuracy of routine chest radiography for the acute aortic syndrome (dissection, intramural hematoma, penetrating ulcer, or nondissecting aneurysm). Methods During a 6-year period, 216 patients (143 men, 73 women; mean [± SD] age, 58 ± 17 years) underwent chest radiography for suspected acute aortic syndrome. Chest films were re-evaluated blindly for aortic disease, based on an overall impression using standard criteria such as widening of the aortic contour and mediastinal shadow. Findings were matched to tomographic images, anatomical inspection, or both, as the criterion standard; aortic disease was confirmed in 109 (50%) of the patients. Results Chest radiography had a sensitivity of 64% (70/109) and a specificity of 86% (92/107) for aortic disease. Sensitivity was 67% (38/57) for overt aortic dissection, 61% (22/36) for nondissecting aneurysm, and 63% (10/16) for intramural hemorrhage or penetrating ulcer. However, sensitivity was lower for pathology confined to the proximal aorta (47% [21/45]) than for disease involving distal aortic segments (77% [49/64]). A receiver operating characteristic curve analysis of aortic diameters failed to identify a threshold for the diagnosis of aortic disease. Conclusion Chest radiography is of limited value for diagnosing the acute aortic syndrome, particularly for conditions confined to the ascending aorta. Since a definitive diagnosis is required in any patient with clinically suspected acute aortic syndrome, routine chest radiography should be replaced by tomographic aortic imaging.