Multislice CT has changed the initial management of trauma patients with multiple injuries and now plays a central role in their primary diagnoses. Many radiology departments have implemented whole-body CT scanning for polytrauma patients.
Multislice CT has changed the initial management of trauma patients with multiple injuries and now plays a central role in their primary diagnoses. Many radiology departments have implemented whole-body CT scanning for polytrauma patients.1
The ability to image large areas with previously unsurpassed spatial and temporal resolution has also improved the quality of CT angiography. When assessing injuries with CT, it takes little additional effort to visualize associated vessels as well. MSCT-based angiography is consequently showing promise as an accurate screening technique for vessel damage, including blunt carotid artery injury.
The observed incidence of blunt carotid artery injury is rising. This is due to the increasing number of high-velocity accidents involving high-powered modern cars and improved assessment of trauma patients, often with state-of-the-art imaging equipment. The reported incidence of blunt carotid artery injury ranges from 0.03% to 3%.2-4 It is associated with a significant overall mortality of up to 33% and a neurological morbidity of up to 38%.4,5
For patients with blunt carotid artery injury, survival rates, therapeutic success, and permanent sequelae depend greatly on initial management. While 96% of patients develop symptoms, late onset of characteristic symptoms can be expected in approximately 40% of patients. This may delay clinical diagnosis for up to five months after the initial trauma.4 These injuries are also challenging to diagnose because their overall incidence is low, and they are often masked by concomitant head trauma or coma in polytrauma patients.
No study has proven that screening for blunt carotid artery injury prevents morbidity or mortality. It has been shown, however, that patients who are diagnosed and treated while asymptomatic have a lower incidence of cerebrovascular sequelae, with the rate of stroke decreasing from 54% in untreated patients to 14% in treated patients.2 Cerebral ischemia is usually present by the time patients become symptomatic, and it may worsen the prognosis considerably. These current data strongly suggest that screening is justified, and the idea has become accepted more widely.
The optimal imaging-based screening method for blunt carotid artery injury remains a subject of much debate. Digital subtraction angiography is the accepted standard of reference for evaluating carotid arteries. It can detect and localize a dissection, pseudoaneurysm, occlusion, or transaction accurately. Compared with CTA, however, DSA is invasive, expensive, resource-intensive, and time-consuming, and it involves a 1% risk of stroke.6 DSA is consequently reserved for cases in which blunt carotid artery injury is already suspected.
Ultrasound is widely available, and it can diagnose blunt injuries of the common carotid artery and the proximal internal carotid artery with high sensitivity.7 But its assessment of the most frequently affected vessel segments of the internal carotid artery, close to or in the petrous bone, is insufficient.
Promising results have been reported for MR angiography.8 Its widespread use in trauma patients is limited by equipment availability, difficulties monitoring patients during scanning, and, most important, the overall length of the imaging procedure.
MSCT-based angiography is now an established tool for the evaluation of many vascular diseases.9 Most experience of carotid artery CTA concerns the classification of atherosclerotic disease, with many studies demonstrating a good or excellent correlation between DSA and CTA.10,11 This suggests that MSCT will characterize and localize traumatic carotid artery lesions correctly as well. Possible lesions include intimal irregularities, dissections, formation of a flap or thrombus (Figures 1 and 2), pseudoaneurysms (Figures 3 and 4), vessel occlusions, and transections.
Earlier studies using single-slice CT scanners raised controversy over the use of CTA to detect traumatic carotid artery lesions.12,13 Two prospective studies using single-slice technology found CTA to be inadequate for screening of blunt carotid artery injury, owing to a sensitivity of less than 70%.3,8
Prospective randomized studies comparing multislice CTA and DSA are still lacking; there are no published trials comparing the two techniques because it is simply not possible to screen every polytrauma patient using DSA. However, new data indicate the promise of multislice CTA for screening. Biffl et al demonstrated that multislice CTA can detect all clinically significant injuries.14 Bub et al reported a specificity of 88% to 98% and a sensitivity of 83% to 92% when comparing multislice CTA and DSA retrospectively.15 In another retrospective study, Elijovich et al found the accuracies of multislice CTA and MRA to be comparable.16
Limited information is available about the image quality ultimately attainable with multislice CTA and the spectrum of pathologies that may be confidently delineated with this technology. The quality of CTA images is critically dependent on contrast density and artifacts from various sources.
Two factors play a key part in vessel contrast density. High luminal attenuation, exceeding 300 HU, facilitates vessel identification. Vessels will show up well against soft tissues with lower densities of around 90 HU. Contrast-related artifacts-halo and edge blur artifacts, for example-also have a direct impact on the accuracy of vessel imaging. A phantom vessel study demonstrated no difference between various contrast densities with respect to halo artifacts.17 Investigators demonstrated improved edge definition with higher contrast density, of approximately 300 to 350 HU, but they also found decreasing accuracy of stenosis measurements with increasing vessel contrast density. They attributed this to the presence of blooming effects. Such artificial effects are believed to be caused by expanding vessel edges and more complete edge pixel saturation with higher contrast densities.
Halo and blooming artifacts are likely to obscure small vessel wall hematomas related to blunt carotid artery injury. Claves et al demonstrated an optimal contrast density of 150 to 250 HU for the overall evaluation of vessels using nonionic contrast media.17
Artifacts unrelated to contrast media are an important limitation of multislice CTA. The rapid scan times mean that artifacts from vessel pulsation are not a problem. The topographical proximity of vessels to osseous structures in the skull base and the shoulder bones produce typical artifacts on multislice CTA. These artifacts can be caused by beam-hardening effects of bone, where low-energy photons are absorbed more strongly than high-energy photons, resulting in distorted projections. They can also occur when x-ray scattering generates a bandlike increase of attenuation values. Using correction algorithms and reducing slice thickness can lessen these effects.18
Vessel differentiation in the petrous portion of bone can be further complicated by the high attenuation of both contrast-filled vessels and bone. The matched-masked bone elimination technique can be used here to avoid bone artifacts. The technique subtracts an enhanced from an unenhanced CT scan to remove bone pixels from the 3D CTA data set.19 This method is promising, though it does not remove beam-hardening artifacts completely. The 15-minute postprocessing time required also limits the use of this technique in a trauma setting.
The majority of artifacts on multislice carotid CTA are due to dental hardware. These artifacts can obscure the target vessel, making diagnostic evaluation impossible. The artifacts crucially hamper imaging of the cervical part of the internal carotid artery, the location most commonly affected by traumatic dissections. This is a serious limitation.
The major sources of metal artifacts are, as with bone artifacts, beam-hardening and photon scattering. With metal, however, these artifacts can be more pronounced. The severity of artifacts can depend on the size, shape, density, atomic number, and position of the metal object and the patient's size and shape.20
Beam-hardening correction functions used for reconstruction of CT data are generally derived and calibrated for body tissues. They are consequently not accurate for metal objects. Reconstructed CT images may show broad dark and bright streaks emanating from metal objects, with typical attenuation values above 2000 HU. Although specific metal artifact correction functions have been developed in the past, they are incompatible with existing reconstruction methods.21
Multislice CTA compares well against other noninvasive imaging methods as a screening test for blunt carotid artery injury. It is true that MRA is not hampered by x-ray artifacts, and its images are not degraded by dental hardware.22 MRA quality is influenced by different artifacts, however, notably those related to patient movement and blood flow.
MRA provides detailed images of the vessel wall and lumen, which is a significant advantage. Difficulties distinguishing acute-stage vessel wall hematomas have been reported, though.23 A vessel wall hematoma may be the only sign of blunt carotid artery injury in some patients. Many vessel lesions detected by ultrasound are unspecific. A vessel wall hematoma may be overlooked or misdiagnosed on ultrasound examination.24 Even DSA, the standard of reference, offers limited imaging of the outer vessel wall and surrounding soft-tissue changes.25
Multislice CTA holds great promise as an effective and reliable noninvasive screening test for blunt carotid artery injury in trauma patients. It can be integrated into the whole-body trauma CT scans that are already performed routinely at many institutions. This makes an early and rapid diagnosis of a potentially life-threatening dissection of the carotid artery both possible and practicable to make.
DR. BORISCH is a staff radiologist and PROF. DR. FEUERBACH is director in the department of diagnostic radiology at the University Hospital Regensburg in Germany.