Noninvasive evaluation of suspected thoracic aortic disease by contrast-enhanced computed tomography.

Traditionally, suspected thoracic aortic disease has been evaluated by aortography, which has associated risks because it is invasive. With the introduction of computed tomography (CT), a noninvasive alternative has become available. In the present retrospective study, the potential clinical value of CT in providing correct diagnoses and pertinent information required for current therapies is evaluated. For this purpose, results of CT in 200 patients examined for suspected thoracic aortic disease were compared with "hard" (surgical and autopsy findings) or "soft" (follow-up clinical information) evidence of the true diagnoses. Aortographic results, available in 51 patients (26%), were also compared with available clinical evidence. Excluding inadequate examinations, the diagnostic accuracy of the independently and blindly interpreted results of CT and aortography were similar (86% and 87%, respectively) in patients with true diagnoses confirmed by hard evidence. When patients evaluated for aortic damage from acute blunt chest trauma were also excluded, the accuracy was 90% for CT and 86% for aortography. Specifically, CT was 83% and 67% accurate in proved (i.e., confirmed by hard evidence) type A and type B dissections, respectively (75% for both type A and B by aortography). In the 183 patients suspected of having thoracic aortic disease not attributable to acute blunt chest trauma and with follow-up information, 91% would have been managed appropriately based on their CT evaluation alone. In the subset of patients who underwent aortographic evaluation as well, 91% would have been managed appropriately based on aortography alone. The accuracy for combined CT and aortography was 94%.(ABSTRACT TRUNCATED AT 250 WORDS)