1998 Plenary Session: Imaging Symposium

Transcription

1 1998 Plenary Session: Imaging Symposium LEARNING OBJECTIVE Understand the characteristic imaging findings of acute cerebral infarction and dural sinus thrombosis and be familiar with the current best imaging methods for diagnosis. Nontraumatic Neurologic Emergencies: Imaging Findings and Diagnostic Pitfalls 1 James M. Provenzale, MD n INTRODUCTION A number of nontraumatic neurologic conditions exist that are associated with high morbidity and mortality. These conditions often have a subacute clinical presentation, which can lead to delay in performing appropriate imaging studies. In addition, a number of these disorders may have subtle radiologic findings, which can lead to a delay in diagnosis. Recent improvements in computed tomography (CT) and magnetic resonance (MR) imaging have facilitated diagnosis of a number of these conditions. In some of these disorders (eg, dural sinus thrombosis), the radiologist may be the first to suggest the correct diagnosis because the clinical features are often ill defined and nonspecific. It is important for radiologists to be familiar with the various imaging findings associated with these neurologic emergencies to ensure early diagnosis and treatment. In this article, a representative sample of these entities are discussed and examples of pitfalls in the radiologic diagnosis are presented. The entities discussed are acute cerebral infarction, dural sinus thrombosis, and dissection of the cervicocephalic arteries. n ACUTE CEREBRAL INFARCTION Recent developments in the treatment of hyperacute stroke (cerebral infarction within 6 hours after onset) have substantially increased the importance of early diagnosis with imaging techniques. In particular, the use of recombinant tissue plasminogen activator offers the possibility of reversing ischemic deficits within the first few hours after stroke onset (1). Previously, the role of imaging was primarily to determine whether intracerebral hemorrhage was present and exclude lesions mimicking cerebral infarcts; presently, greater emphasis is placed on demonstrating the presence (and defining the limits) of the actual infarct. The CT findings of acute infarction include decreased attenuation within affected brain regions, early brain swelling, and, on occasion, increased attenuation within thrombosed arteries (eg, the so-called dense middle cerebral artery sign) (2) (Fig 1a). Early decreased attenuation Index terms: Brain, infarction, Carotid arteries, dissection, 17.74, Cerebral blood vessels, thrombosis, Computed tomography (CT), angiography, Magnetic resonance (MR), diffusion study, , Sinuses, dural, Vertebral arteries, dissection, , RadioGraphics 1999; 19: From the Department of Radiology, Duke University Medical Center, Erwin Rd, Durham, NC From the Plenary Session, Friday Imaging Symposium: Acute Radiology Where Minutes Count, at the 1998 RSNA scientific assembly. Received March 15, 1999; revision requested May 4 and received June 2; accepted June 8. Address reprint requests to the author. RSNA, 1999 September-October 1999 Provenzale n RadioGraphics n 1323

2 The major early signs of acute infarction at MR imaging include development of high signal intensity on T2-weighted images and stasis of contrast material within affected arteries on contrast material enhanced images (4). Although conventional spin-echo MR imaging can demonstrate acute infarction earlier than CT, diffusion-weighted imaging is even more sensitive for detection of stroke within the first few hours after onset. Diffusion-weighted imaging is unique in that image contrast results from varying degrees of signal loss within tissues (rather than actual signal increase, as in conventional MR imaging). After application of the gradient pulses used in diffusion-weighted imaging, normal tissue undergoes signal loss related to microscopic (Brownian) motion of water within tissue. The signal loss is proportional to a number of factors, including the rate of water diffua. b. Figure 1. Early infarct in the territory of the left middle cerebral artery in a 52-year-old man. (a) Unenhanced axial CT image shows increased attenuation of the left middle cerebral artery (arrow) due to a thrombus. (b) CT image obtained a few images cephalad to a shows subtle decreased attenuation within the left insula (arrows) relative to that in the right hemisphere (arrowhead). within an infarct can be subtle but can nonetheless be detected when a gray matter structure becomes isoattenuating relative to adjacent white matter. Examples of this phenomenon include the so-called insular ribbon sign (3) (Fig 1b) and decreased attenuation within deep gray matter structures such as the basal ganglia (Fig 2). However, in general, CT is relatively insensitive and nonspecific in diagnosis of infarction within the first few hours after onset (ie, within the therapeutic window of the major thrombolytic therapies). MR imaging in particular diffusion-weighted MR imaging has attained an important role in determination of hyperacute stroke because of the ability to demonstrate hyperacute infarction n Imaging Symposium Volume 19 Number 5

3 a. b. Figure 3. Hyperacute stroke in a 50-year-old man with bacterial endocarditis and suspected cerebral infarction. (a) Axial proton density weighted MR image shows subtle regions of high signal intensity at the gray-white matter junction (arrows). (b) Corresponding diffusion-weighted MR image shows the infarcts more clearly (arrows). The pattern of infarcts in multiple vascular distributions is consistent with embolic infarction. sion and the strength and duration of the applied gradient pulses (5). In acute infarcts, the degree of water diffusibility is decreased (ie, water diffusion is restricted), possibly due to passage of extracellular water into cells or to a decrease in extracellular water (and resultant impairment of the flow of extracellular water around cells) (5). The result is a net decrease in signal loss in acute infarcts, which thus appear hyperintense relative to normal tissue, in which more signal loss has occurred (Figs 3, 4). Figure 2. Early infarction in a 2-year-old girl with sudden onset of left hemiparesis a few days after cardiac surgery. Unenhanced axial CT image shows that the right basal ganglia (open arrows) are slightly hypoattenuating relative to the left basal ganglia (solid arrow). Note that the left basal ganglia are distinctly seen against the background of the left internal capsule (arrowhead), but the distinction between these two structures is blurred in the right hemisphere because of the lower attenuation of the right basal ganglia. September-October 1999 Provenzale n RadioGraphics n 1325

4 a. b. Figure 4. Infarcts of different ages in a 60-year-old man with new onset of right hemiparesis. (a) Axial proton density weighted MR image shows multiple regions of high signal intensity in both cerebral hemispheres (arrows). The lesions are isointense relative to one another, and their relative ages cannot be discerned. (b) Corresponding diffusion-weighted MR image shows that some of the lesions are hyperintense (open arrows), an appearance consistent with acute infarcts. In particular, one lesion that is relatively subtle on the proton density weighted image (a) is much more conspicuous (solid arrow). The remainder of the lesions seen on the proton density weighted image (a) are isointense relative to normal brain tissue on the diffusion-weighted image (b); this finding indicates that these lesions are subacute or chronic. In the evaluation of patients with suspected acute cerebral infarction, diffusion-weighted imaging offers a number of advantages over conventional spin-echo MR imaging. First, diffusion-weighted imaging is more sensitive than conventional MR imaging for detection of acute infarcts during the first few hours after stroke onset (6). Particularly in the first 12 hours after onset of neurologic symptoms, abnormalities that are inapparent (or very subtle) on T2- weighted images can be detected on diffusionweighted images (Fig 3). Second, diffusionweighted images allow more accurate discrimination of acute infarcts from older infarcts than do conventional images (on which infarcts of various ages all appear isointense) (Fig 4). Acute infarcts continue to have high signal intensity on diffusion-weighted images for approximately 7 10 days. When more than one infarct is seen on conventional MR images, comparison with diffusion-weighted images allows determination of which infarcts (if any) are acute (Fig 4). Third, diffusion-weighted imaging allows better characterization of a lesion as an acute infarct or another entity (eg, a neoplasm, which would be expected to have associated vasogenic, rather than cytotoxic, edema) n Imaging Symposium Volume 19 Number 5

5 a. b. Figure 5. Hemorrhagic venous infarction in a 68-year-old man with a 1-week history of severe headaches. (a) Unenhanced coronal T1-weighted MR image shows a hyperintense focus consistent with hemorrhage in the right temporal lobe (straight arrow). The subcortical location and the hemorrhagic nature of the lesion are typical of a venous infarct. The flow voids of the right sigmoid sinus (curved arrow) and superior sagittal sinus (arrowhead) are replaced by abnormal high signal intensity, which is consistent with thrombosis. (b) Axial T2-weighted MR image shows a hemorrhagic infarct in the left parietal lobe (straight arrows) and high signal intensity in the superior sagittal sinus (curved arrow), findings consistent with thrombosis. n DURAL SINUS THROMBOSIS Dural sinus thrombosis can occur on a spontaneous basis or be secondary to a hypercoagulable state, dehydration, infection, or dural sinus compression (7). In addition, evidence suggests that oral contraceptive use, pregnancy, and the peripartum state increase the risk of dural sinus thrombosis. The clinical manifestations are often vague and can include symptoms due to increased intracranial pressure (eg, headache, nausea, and visual blurring) (7). The thrombosis can also extend in a retrograde manner into cerebral veins and cause venous infarction (Fig 5); in this instance, focal neurologic signs can be seen. Symptoms are easily mistaken for those of tension headache, migraine headache, or other disease processes, such as pseudotumor cerebri or an intracranial mass. Furthermore, the imaging findings are easily overlooked if not specifically sought. The radiologist may be the first physician to suggest the diagnosis, which is a radiologic (and not merely a clinical) one. For these reasons, the radiologist often plays an important role in establishing the diagnosis. MR imaging and MR venography are currently the preferred techniques for making the diagnosis of dural sinus thrombosis. However, CT venography has also been shown to be useful and may eventually become the preferred method of diagnosis. Conventional CT (ie, CT performed without the helical technique) can show dural sinus thrombosis on unenhanced images as a hyperattenuating region that conforms to the expected location of a dural sinus September-October 1999 Provenzale n RadioGraphics n 1327

6 a. Figure 6. Dural sinus thrombosis in a 20-yearold woman with a 2-week history of worsening headache. (a) Unenhanced axial CT image shows increased attenuation in the superior sagittal sinus (straight arrow) and straight sinus (curved arrow), findings consistent with thrombosis. (b) Unenhanced sagittal MR image shows replacement of the flow void of the superior sagittal sinus by intermediate signal intensity (arrows), which indicates thrombosis. (c) Contrast-enhanced axial MR image shows absence of the flow void in the superior sagittal sinus (straight arrow) and straight sinus (curved arrow) with mild enhancement of the periphery of the thrombus. (Courtesy of D. James Schumacher, MD, Cape Canaveral Hospital, Melbourne, Fla.) (Fig 6); after administration of contrast material, dural sinus thrombosis is seen as a nonenhancing focus within a dural sinus (8). However, CT is generally an ineffective method of evaluating patients with suspected dural sinus thrombosis because it is insensitive and nonspecific (8). At MR imaging, the findings include replacement of the flow void of one or more dural sinuses by abnormal signal intensity (7) (Fig 6) n Imaging Symposium c. b. The signal intensity of the thrombosed dural sinus depends on the age of the thrombus and the pulse sequence chosen. Frequently, by the time imaging is performed, the blood products within the thrombus are in the stage of intracellular methemoglobin or extracellular methemoglobin. Both of these entities produce high sig- Volume 19 Number 5

7 nal intensity on T1-weighted images. Extracellular methemoglobin produces high signal intensity on T2-weighted images (Fig 5); thus, the abnormal signal intensity within the thrombosed dural sinus is evident with all pulse sequences. However, intracellular methemoglobin typically produces low signal intensity on T2-weighted images and can thus mimic the flow void of a patent dural sinus. This finding is a potential pitfall in diagnosis of dural sinus thrombosis but can be avoided by carefully evaluating the dural sinuses with all pulse sequences (9). Venous infarcts secondary to retrograde extension of thrombus into cortical veins are frequently subcortical and hemorrhagic; the thrombus itself has the signal intensities outlined earlier in this paragraph. After administration of contrast material, the walls of the affected dural sinus often enhance (Fig 6); the thrombus is then typically seen as an unenhancing region with rim enhancement (9). On rare occasions, the thrombus itself also enhances. At MR venography, dural sinus thrombosis is characterized by absence of signal within one or more dural sinuses. Either time-of-flight or phase-contrast MR venography can be used to make the diagnosis. One potential pitfall of use of time-of-flight MR venography is that thrombus that is hyperintense on T1-weighted images can simulate flow-related enhancement on MR venograms. This pitfall can be avoided by careful evaluation of routine spin-echo images or by acquisition of phase-contrast images, on which the thrombus will not appear hyperintense (9). CT venography is an excellent alternative to MR imaging for diagnosis of dural sinus thrombosis. CT venography is similar to CT angiography in terms of the helical data acquisition but differs by using a second delay after contrast material administration (10), thus allowing optimal visualization of the venous system. Bone detail is subtracted from the resultant images with a thresholding technique; thereafter, data can be analyzed from source images or reconstructed with a variety of methods, such as maximum intensity projection. The appearance of dural sinus thrombosis at CT venography is similar to that at MR venography (ie, a filling defect within the thrombosed dural sinus, alone or in association with a venous infarct) (10). Advantages of CT venography over MR imaging include a shorter examination time and the ability to study patients in whom MR imaging is contraindicated or difficult to perform. In addition, CT venography greatly increases the sensitivity of CT for the diagnosis of dural sinus thrombosis because direct visualization of the dural sinuses is possible. n DISSECTION OF THE CERVICO- CEPHALIC ARTERIES Dissection of the carotid and vertebral arteries accounts for 5% 20% of strokes in young and middle-aged adults (11,12). A number of misconceptions regarding this entity exist, including the beliefs that dissection of these arteries occurs primarily in patients with underlying vasculopathies (eg, Marfan syndrome, fibromuscular dysplasia, and Ehlers-Danlos syndrome) and is usually posttraumatic. In fact, an underlying vasculopathy is found in only a small percentage of cases. Furthermore, the majority of carotid and vertebral artery dissections in the general population are unrelated to overt trauma. In many instances, patients have experienced minor trauma near the time of onset of symptoms, although the trauma is usually so insignificant as to be unnoticed at the time of occurrence. The most common symptom of dissection of the cervicocephalic arteries is headache or neckache, which is often worse on the side of the affected artery. Oculosympathetic paresis (Horner syndrome) is seen in a minority of patients with carotid artery dissection; when accompanied by headache, oculosympathetic paresis is suggestive of the diagnosis (11). Transient ischemic attacks or stroke may be the initial manifestation and appears to be more common with intracranial dissections than with dissections of the extracranial arterial segments. The diagnosis of dissection of the cervicocephalic arteries should be considered in any young or middle-aged patient with transient ischemic attacks or complete stroke. Carotid artery dissection is approximately twice as common as vertebral artery dissection. The most common site of carotid artery dissection is within the cervical portion of the internal carotid artery, within a few centimeters of September-October 1999 Provenzale n RadioGraphics n 1329

8 a. Figure 7. Carotid artery dissection in a 44-year-old man with onset of neck pain while skiing. (a) Unenhanced axial T1-weighted MR image shows a rim of high signal intensity around the flow void of the right internal carotid artery (arrowhead). (b) Lateral angiogram of the right internal carotid artery shows a long segment of narrowing (arrows) beginning a few centimeters beyond the carotid bifurcation, a typical site for dissection of the extracranial segment of this artery. b. the carotid bifurcation (Figs 7, 8). The supraclinoid segment of the internal carotid artery is the most common intracranial site. Approximately 65% of vertebral artery dissections occur at the level of the C1-C2 vertebral complex, possibly due to stretching of the artery over the lateral mass of the C2 vertebral body during lateral rotation of the head or unusual head motions. The diagnosis of arterial dissection can be made by a number of means, including conventional angiography, MR imaging, CT angiography, and duplex ultrasonography (US). At conventional angiography, the variety of possible findings include arterial narrowing (Fig 7), occlusion, an intimal flap, and a pseudoaneurysm (Fig 8). MR imaging and MR angiography have a high sensitivity and specificity for detection of arterial dissection and are presently the preferred noninvasive means of making the diagnosis (13,14). However, preliminary studies of CT angiography have also shown great promise in this regard (15,16). Because both MR imaging and CT angiography are noninvasive, they are optimal for follow-up imaging of patients in whom the diagnosis of dissection has already been made (16). At MR imaging, the most common finding of arterial dissection is a rim of high signal intensity (which represents intramural hematoma) Figure 8. Pseudoaneurysm due to carotid artery dissection in a 55-year-old woman with right-sided pulsatile tinnitus. Lateral angiogram of the right internal carotid artery shows a pseudoaneurysm (arrow) due to arterial dissection at the junction of the cervical and petrous segments of the artery. around the flow void of the affected artery (Fig 7). The arterial lumen may be normal in caliber, stenosed, or occluded. The length of the arterial segment with the rim of abnormal signal intensity varies from patient to patient but is typically on the order of a few centimeters (Fig 7). Other findings can include a pseudoaneurysm (Fig 8) or, in unusual circumstances, a double 1330 n Imaging Symposium Volume 19 Number 5

9 lumen similar to that seen in aortic dissection (13). Essentially the same appearances can be seen at CT angiography and conventional angiography. At MR angiography, the intramural hematoma typically has signal intensity intermediate between the high signal intensity of flowing blood and the signal intensity of background soft tissue. The primary findings of arterial dissection at duplex US are an echogenic intimal flap or echogenic thrombus (17). At Doppler US, the primary finding is an abnormal Doppler signal with an accompanying gray-scale image that does not show a large focus of atheromatous plaque as a cause of the abnormal Doppler signal (17). These US findings are considered suggestive of the diagnosis; at the vast majority of institutions, further imaging with conventional angiography, MR imaging, or CT angiography is performed on a standard basis to confirm the diagnosis. The primary treatment of arterial dissection is anticoagulation, which is intended to prevent cerebral infarction from emboli arising from platelet-fibrin aggregates along the exposed arterial wall. After anticoagulant therapy, imaging findings usually normalize after a few months. However, worsening clinical findings or imaging features occur in a small minority of cases (approximately 10%) and are usually considered an indication for surgical treatment or endovascular stent placement. n CONCLUSIONS This article presents the imaging findings of a number of neurologic conditions that may be encountered by the radiologist in the emergency department setting. Because the clinical features of these conditions are often nonspecific, in many cases, the radiologist may be the first person to propose the correct diagnosis on the basis of the imaging findings. Therefore, it is important that radiologists working in the emergency department setting be familiar with the entities described herein to minimize the time to diagnosis. n REFERENCES 1. National Institute of Neurological Disorders and Stroke rt-pa Stroke Study Group. Tissue plasminogen activator for acute ischemic stroke. N Engl J Med 1995; 33: Tomsick TA, Brott TG, Chambers AA, et al. Hyperdense middle cerebral artery sign on CT: efficacy in detecting middle cerebral artery thrombosis. AJNR 1990; 11: Truwit CL, Barkovich AJ, Gean-Marton A, Hibri N, Norman D. Loss of the insular ribbon: another early CT sign of acute middle cerebral artery infarction. Radiology 1990; 176: Yuh WTC, Loes DJ, Greene GM, et al. MR imaging of cerebral ischemia: findings in the first 24 hours. AJNR 1991; 12: Moseley ME, Kucharczyk J, Mintorovich J, et al. Diffusion-weighted MR imaging of acute stroke: correlation with T2-weighted and magnetic susceptibility enhanced MR imaging in cats. AJNR 1990; 11: Sorensen AG, Buonanno FS, Gonzalez RG, et al. Hyperacute stroke: evaluation with combined multisection diffusion-weighted and hemodynamically weighted echo-planar MR imaging. Radiology 1996; 199: Zimmerman RD, Ernst RJ. Neuroimaging of cerebral venous thrombosis. Neuroimaging Clin North Am 1992; 2: Virapongse C, Cazenave C, Quisling R, Sarwar M, Hunter S. The empty delta sign: frequency and significance in 76 cases of dural sinus thrombosis. Radiology 1987; 162: Provenzale JM, Joseph GJ, Barboriak DP. Dural sinus thrombosis: findings on CT and MR imaging and diagnostic pitfalls. AJR 1998; 170: Casey SO, Alberico RA, Patel M, et al. Cerebral CT venography. Radiology 1996; 198: Bogousslavsky J, Regli F. Ischemic strokes in adults younger than 30 years of age: cause and prognosis. Arch Neurol 1987; 44: Lisovoski F, Rousseaux P. Cerebral infarction in young people: a study of 148 patients with early cerebral angiography. J Neurol Neurosurg Psychiatry 1991; 54: Provenzale JM. Dissection of the internal carotid and vertebral arteries: imaging findings. AJR 1995; 165: Levy C, Laissy JP, Raveau V, et al. Carotid and vertebral artery dissections: three-dimensional time-of-flight MR angiography and MR imaging versus conventional angiography. Radiology 1994; 190: Lelerc X, Godefroy O, Salhi A, Lucas C, Leys D, Pruvo JP. Helical CT for the diagnosis of extracranial carotid artery dissection. Stroke 1996; 27: Lelerc X, Lucas C, Godefroy O, et al. Helical CT for the followup of cervical internal carotid artery dissections. AJNR 1998; 19: Gardner DJ, Gosink BB, Kallman CE. Internal carotid artery dissections. J Ultrasound Med 1991; 10: September-October 1999 Provenzale n RadioGraphics n 1331