Diagnosis of blunt traumatic aortic injury 2007: Still a nemesis

Abstract

In recent years, the use of multidetector computed tomography (MDCT) for the diagnosis of acute thoracic injury in blunt trauma has expanded. MDCT has shown high accuracy for the diagnosis or exclusion of injury to the aorta and its primary branches, decreasing the need for thoracic angiography and allowing earlier treatment of this often rapidly fatal lesion. With increasing use of MDCT, more subtle injuries and variants of vascular anatomy are being recognized that create pitfalls in the diagnosis. Of perhaps more concern is the recognition that aortic injury can occur with little or no associated mediastinal hematoma, the principle chest radiographic finding indicating a need for further imaging. The importance of recognizing unusual sites of aortic injury, congenital variants of mediastinal anatomy, the precise extent of injury, and the anatomic pathology present as key factors in deciding among treatment options is emphasized.

Introduction

The use of computed tomography (CT) for the diagnosis of traumatic aortic and major arterial branch vessel injury has evolved significantly over the last three decades. In the early 1980s, reports began to appear describing this CT application and some early series indicated that the method was highly accurate to diagnose or exclude this potentially fatal injury [1], [2], [3]. However, general acceptance of contrast-enhanced CT for this application was slow. Not all studies indicated a high level of reliability for CT [4], [5]. Many thoracic surgeons, familiar with thoracic angiographic, were generally more comfortable with this modality, which remained the “reference standard” in many centers well into the 1990s

Prior to widespread availability of CT scanning in emergency and trauma centers, the admission chest radiograph combined with angiography served as the diagnostic tools to assess the aorta and its major branches in high-speed decelerating trauma [6], [7]. The approach depended on the premise that essentially all patients who harbored these injuries would manifest an abnormal mediastinal contour by chest radiography secondary to mediastinal hemorrhage. The abnormal mediastinum would be the trigger to perform thoracic angiography as the definitive study. Unfortunately, chest radiography proved a relatively poor screening test for aortic injury [6], [7], [8]. A pristine, sharply demarcated mediastinal contour was required to exclude mediastinal hemorrhage and thus, by inference, major arterial mediastinal injury. Often however, chest radiographs demonstrated mediastinal abnormalities due to the well-known limitations of bedside anteroposterior supine chest radiography in the setting of trauma and non-vascular chest abnormalities. Even the presence of hemomediastinum was far more often a false positive sign as an indicator of aortic injury [9].

As CT technology advanced, first with the introduction of helical CT, and later multi-row detector CT (MDCT), the quality of images obtained in the axial plane, multiplanar reformations, surface contour, and volume rendering techniques improved in step. Over the past decade, these advances have made CT the usual definitive screening test for major thoracic vascular injury, without the requirement of arteriography or transesophageal echocardiography to exclude or confirm the diagnosis, with the rare exception of an equivocal CT result or a technically inadequate study [9], [10], [11].

While the improved spatial resolution, better overall image quality, and supplemental post-processing techniques have been a great boon to the success of CT for the diagnosis of aortic-great vessel injury, these advances have also led to recognition of greater numbers of normal variants of vascular anatomy and subtler vascular injuries that were less likely to be appreciated using earlier conventional CT scans obtained with wider slice thickness, greater motion artifacts, poorer quality multiplanar reformations, and less consistent peak aortic contrast opacification.

In some ways this technological leap has been the equivalent of turning to a new, more powerful, microscope lens, suddenly introducing a new world of anatomic findings, not apparent through the weaker objective. This article reviews the major MDCT imaging findings of typical aortic and great vessel injuries, demonstrates a variety of anatomic variants that can simulate injury, and shows the wide spectrum of injuries that occur, from very minimal to gross lesions. The development and increasing use of non-surgical treatment options as blood pressure regulation and arterial stent-grafts for temporary or definitive management of these injuries has made precise localization and characterization of injuries more important than ever for optimal therapy. Using our higher level of CT imaging quality it will become clear that some “established concepts” regarding the diagnosis of aortic trauma are on somewhat shaky ground.

In this article, the term aortic injury will be taken to include injury to the intrathoracic aortic branches as well.