The 2005 Annual Scientific Meeting of the Society of Interventional Radiology was held March 31 through April 5, 2005, in New Orleans. At the meeting, Etiology and Clinical Manifestation of Stroke was presented by Rafael H. Llinas, MD, assistant professor of neurology, Johns Hopkins Medicine, Baltimore. Llinas notes that while stroke is the third most common cause of death among US residents 40 or more years old, it is the leading cause of morbidity requiring long-term care (outranking even Alzheimer disease). The United States has 4 to 5 million residents affected by stroke, which has a US incidence rate of 600,000 to 700,000 per year and contributes to 150,000 US deaths annually.

All of the clinical manifestations of stroke are of sudden onset: headache, unilateral or bilateral loss of vision, abnormal speech, vertigo, abnormal gait, and unilateral weakness. Roughly 15% of strokes are due to primary hemorrhages, but the remaining 85% are ischemic. About 20% of these are attributed to atherosclerotic cerebrovascular disease (hypoperfusion or arteriogenic emboli); another 25%, to penetrating artery disease; 20%, to cardiogenic embolism of many types, including valve disease, ventricular thrombi, and atrial fibrillation; 30%, to no known cause; and 5% to several less common causes, including drug abuse, prothrombic states, dissections, arteritis, and migraine/vasospasm. The primary risk factors for stroke are smoking (the most prominent risk), arterial hypertension, diabetes, hypercholesterolemia, elevated plasma homocysteine, and three factors often found together: angina, myocardial infarction, and limb claudication.

The Interventional Management of Stroke Study

Ajay K. Wakhloo, MD, PhD, is chief of neuroendovascular surgery and interventional neuroradiology, departments of radiology, neurological surgery, and biomedical engineering, at the University of Miami. At the 2005 Annual Scientific Meeting of the Society of Interventional Radiology, held March 31 through April 5, 2005, in New Orleans, he gave a presentation entitled IMS Study and Intracranial Devices. He attributes 20% to 30% of strokes to carotid artery disease and notes that the annual cost of care for stroke patients, combined with productivity lost due to stroke, is roughly $43 billion (109) in the United States. Nonetheless, 80% of all strokes in the United States should be preventable by 2010 if risk reduction and the development of new treatments continue. At present, recanalization and neuroprotection determine outcomes for acute stroke patients. Unfortunately, fewer than one fifth of stroke patients arrive for care within 2 hours of the event, greatly hampering the usefulness of intravenous recombinant tissue plasminogen activator (IV rtPA) therapy. Wakhloo estimates that complete recanalization occurs in fewer than 10% of internal carotid artery (ICA) occlusions and 25% of middle cerebral artery occlusions.

The Interventional Management of Stroke (IMS) Study1 investigated the treatment of acute ischemic stroke using combined intravenous and intra-arterial recanalization. The study screened 1,477 subjects and enrolled 80 who scored 10 or more on the National Institutes of Health Stroke Scale and met modified National Institute of Neurological Disorders and Stroke criteria. They were given IV rtPA (0.6 mg per kg, with a maximum dose of 60 mg), with 15% as the initial bolus and the remainder administered over the following 30 minutes. Of the 80 subjects, 77 underwent immediate angiography. If a clot was detected, intra-arterial (IA) rtPA was administered as a 2-mg bolus within and past the clot, for a total dose of 22 mg given over a 2-hour period or until Thrombolysis in Acute Myocardial Infarction (TIMI) grade 3 recanalization occurred. This combined IV/IA therapy was given to 62 patients. The mean time that elapsed between the onset of symptoms and the IV treatment was 2 hours 16 minutes. Angiography took place a mean of 47 minutes later, and IA treatment was given a mean of 34 minutes after that. ICA occlusion or stenosis was noted in 28 subjects, who were given IA rtPA as a 2-hour infusion.

Of the 62 patients who received combined therapy, 11% had TIMI 3 complete recanalization. Of the 56% of subjects who had TIMI 2 or 3 partial or complete recanalization, 34% had favorable outcomes, defined as a modified Rankin Scale score of 0 to 1 after 3 months. Of subjects at TIMI 0 or 1, only 12% had the same favorable outcomes. The 16 subjects who received IA rtPA within 3 hours of symptom onset had favorable outcomes in seven cases (43%). The 24 subjects who were given the same treatment within 3 to 4 hours of onset had favorable outcomes in only three cases (13%). Overall, there was 16% mortality within 3 months, symptomatic intracranial hemorrhage (ICH) within 36 hours in 6.3% of cases, parenchymatous hemorrhage ICH (dense hematoma) in 7.5%, asymptomatic ICH within 36 hours in 43%, and serious systemic bleeding events within 36 hours in 3%.

These results, Wakhloo says, indicate that the next step to be developed in stroke intervention is immediate recanalization that does not use fibrinolytic agents (or uses them only minimally). Promising technologies include some of the more than 65 intracranial devices now available. These can be categorized as low-intensity ultrasound used within the clot itself to promote thrombolysis; neuroprotection (cooling); mechanical clot disruption using photoacoustic devices, wires, grippers, snares, and balloons; and clot suction. Use of these devices can also be combined with thrombolysis or fibrinolysis.


  1. IMS investigators. Combined intravenous and intra-arterial recanalization for acute ischemic stroke: The Interventional Management of Stroke Study. Stroke. 2004;35:904.
K. Kyes

Periprocedural Stroke Evaluation

Interventional radiology procedures can result in strokes, even when distal protection devices are used. It is uncommon for a major stroke to be the result of an interventional procedure, but stroke can be severe when it does occur. Llinas recommends documentation of a brief evaluation for stroke both before and after interventional procedures.

There are two rapid evaluations that may be useful for this purpose. The Cincinnati Stroke Scale is often used by emergency medical technicians in the field, by triage nurses, and by emergency physicians. Although it requires only 30 seconds to perform and is based on only three physical findings, it identifies acute stroke consistently. Patients who have one of the three findings as a new symptom have a 72% probability of acute stroke, Llinas says, and this probability increases to 85% if all three findings are present. Arm drift, the first finding, is assessed by asking the patient to hold out both arms while the eyes are closed. A positive finding consists of asymmetrical arm motion (if both arms drift equally or fail to move at all, this test is negative). The second finding is speech abnormality; the patient is asked to repeat a phrase. If the patient cannot speak, cannot repeat the test phrase, uses inappropriate words, or has slurred speech, this test is considered positive. The third finding is facial droop, assessed by observing the face for asymmetry.

At Johns Hopkins Hospital and Johns Hopkins Bayview Medical Center, both in Baltimore, neuroradiology fellows employ a 60-second stroke evaluation of the frontal, left temporal, parietal, and occipital lobes of the brain. This rapid, focused approach is used (and documented) before, during, and after neuroradiology procedures. It is particularly important to document any deficits found before a procedure, Llinas says, so that they will not be mistakenly categorized as adverse effects of the procedure itself. The 60-second screening has five steps. First, the patient is asked to follow the evaluator’s finger visually; this assesses the frontal-lobe control of eye movements. Second, the patient is asked to say, “It’s a sunny day in Baltimore.” This assesses speech centers in the frontal, temporal, and parietal lobes. Third, the patient is asked to smile, lift the arms for 10 seconds, and lift the legs (if practical) for 5 seconds to assess the frontal motor area. Fourth, the patient attempts to touch the same area with the left hand, right hand, and both hands; this assesses the parietal lobe. Fifth, the patient is asked which of the evaluator’s hands is moving to assess the vision area of the occipital lobe. Llinas notes that anesthesia itself may cause positive findings in some cases, so experience is needed to perform the evaluation properly.

If the 60-second screening performed during the procedure indicates any change from the preprocedure screening results, the sheath used for the procedure should not be removed and the neurology department should be called. Before sheath removal, reshooting the side that was manipulated should be considered. Vascular occlusion should be ruled out, and an acute head CT study should be obtained before any thrombolytic agents are used. Antihypertension therapy should also be considered.

Last-Chance Lysis

Later in the SIR stroke-management session, Llinas presented Outcomes of Stroke Lysis: The “Hail Mary” Approach. The name of this approach is taken from the term for a last-second shot at the end of a basketball game, attempted from wherever the player happens to be (in other words, the situation may not be ideal, but some action must be taken before time runs out for the game or, in this case, for the patient). The approach was developed by Johns Hopkins Medicine because the benefit of intravenous recombinant tissue plasminogen activator (IV rtPA) decreases after 3 hours have passed since the onset of stroke symptoms. Because too much time has elapsed, only 11% of Johns Hopkins Medical’s stroke patients are given IV rtPA. Intra-arterial (IA) thrombolysis, likewise, is used to treat only a minority of patients and is not performed at many facilities. In order to provide it consistently to patients at both Johns Hopkins campuses who cannot be given IV rtPA, it was necessary to institute a new treatment paradigm, Llinas reports.

Before this change, IA thrombolysis had been performed only by interventional neuroradiologists. Llinas describes the available staff as excellent but over-extended. For this reason, Johns Hopkins investigators sought to determine whether IA rtPA could be given safely and effectively without biplane imaging by staff members who were not interventional neuroradiologists and, in some cases, not interventional radiologists.

Figure 1. Patient with basilar artery occlusion was comatose by the time angiography was performed. Prognosis without intervention was poor, with expected mortality of 75%-80%. (Click the image for a larger version.)

Of course, there were several potential problems to be addressed. Interventional neuroradiologists were clearly the best choice for the administration of IA thrombolysis because their skill in navigating intracranial vessels is highly developed. Even those with great expertise may find this difficult without a biplane machine. In less experienced hands, navigation of the intracranial circulation could waste considerable time; in stroke treatment, speed is crucial. Given these limitations, the investigators chose not to pursue the direct infiltration of rtPA into clots when no interventional neuroradiologist was available.

Instead, neuroradiologists and body interventional radiologists performed thrombolysis. Both were skilled at entering the extracranial circulation. The chosen technique was cannulation of the arterial system leading to the artery containing the clot. A multidisciplinary approach was developed, involving vascular neurologists, neurological critical care teams, and radiologists.

The new protocol has been employed for 10 patients since 2002, being reserved for those who were considered otherwise certain to die or become locked in (unable to move anything but their eyes, but aware, which Llinas describes as worse than death). The protocol consists of placement of the catheter into the most distal artery with which the radiologist feels comfortable (and that can be reached quickly). For middle cerebral artery (MCA) or anterior cerebral artery clots, this is usually the proximal internal carotid artery; for basilar clots, it is the vertebral origin of the artery. Eptifibatide is infused at 70g per kg, followed by rinsing; then the dose and rinsing are repeated. Prefrozen 5-mg boluses of rtPA are next infused, followed again by rinsing. This step is repeated three times if the clot is in the anterior circulation and four to five times if it is in the posterior circulation. Llinas stresses that the regular IA rtPA protocol is used instead if an interventional neuroradiologist is available.

Figure 2. Hail Mary protocol consists of placement of catheter into most distal artery with which radiologist is comfortable: in the case of a basilar artery occlusion (a), the vertebral origin of the clot. Patient above was successfully recanalized (b). Images courtesy of Rafael H. Llinas, MD, Johns Hopkins Medicine, Baltimore.

Of the 10 patients treated using the Hail Mary protocol, four had undergone large MCA strokes and six had basilar artery occlusions. Angiography showed 100% occlusion in all 10 cases, and all had large strokes evolving. Their prognoses without intervention were all poor, with expected mortality of 80% or more for the four MCA stem occlusions and 75% to 80% or more for the six basilar artery thromboses. These patients were all comatose by the time angiography was performed.

Seven of the 10 patients had successful recanalization, and five of these were living independently 3 months after the event. Two of the seven experienced reocclusion within 24 hours, despite anticoagulant therapy, and subsequently died. The three remaining patients did not experience successful recanalization and died within a week of the event (one with symptomatic hemorrhages).

Although the group treated was small, it showed a recanalization rate of 70% and a 3-month independence rate of 50%. The Prolyse in Acute Cerebral Thromboembolism, or PROACT II, study1 of IA recombinant pro-urokinase delivered via catheter placed against the clot (without clot manipulation) showed a recanalization rate of 66% and a 3-month independence rate of 40%. Because most of the world’s hospitals lack biplane imaging and interventional neuroradiologists, the multidisciplinary Hail Mary approach merits further investigation, Llinas reports.

Imaging and Stroke

Endovascular Treatment of Acute Ischemic Stroke: Need for Imaging Triage was presented by Colin P. Derdeyn, MD, associate professor, Mallinckrodt Institute of Radiology, and the departments of neurology and neurological surgery, Washington University School of Medicine, St Louis, at the 2005 Annual Scientific Meeting of the Society of Interventional Radiology, held March 31 through April 5, 2005, in New Orleans. Derdeyn states that the need for imaging-based triage in stroke patients is based on the fact that intravenous recombinant tissue plasminogen activator (IV rtPA) is effective if given within 3 hours of stroke onset in patients who show no CT evidence of brain hemorrhage. Because there is no laboratory test available to detect hemorrhage, imaging is needed to exclude hemorrhage before IV rtPA can be given to patients with nonconvulsive neurological deficits of sudden onset. CT is readily available for this purpose at most medical centers, although MRI is also adequate in detecting hemorrhage.

Figure 1. The PET study shows reduced cerebral blood flow (CBF) and normal oxygen metabolism (cerebral metabolic rate for oxygen (CMRO2). This is a patient with a complete occlusion of the carotid artery and reduced flow owing to inadequate collateral supply. The oxygen extraction fraction (OEF) is increased. This is a compensatory mechanism to preserve normal oxygen metabolism when the delivery has fallen (CBF). The presence of increased OEF is a major risk factor for future stroke in patients with occlusive cerebrovascular disease and has also been used to identify ischemic but viable tissue in acute stroke patients. (Click the image for a larger version.)

The availability of reliable, rapid screening, Derdeyn says, would help clinicians understand which patients could benefit most from therapy and, at the same time, could lengthen the window during which therapy is helpful. Using anatomic screening, the size and location of an occlusive thrombus can be identified, while physiological screening indicates the viability of tissue and helps to quantify the risk of hemorrhage. Possible screening methods are CT angiography (CTA)/CT perfusion and MRI/MR angiography (MRA). Positron-emission tomography (PET) can also be used to identify surviving tissue (using the cerebral metabolic rate for oxygen) and tissue at risk (using the oxygen-extraction fraction). When MRI is used, diffusion can indicate dead tissue; ischemic (but viable) tissue can be identified using a perfusion indicator. CT imaging uses a reduced cerebral blood volume (CBV) or a very high mean transit time (MTT) to show dead tissue. It uses increased CBV and MTT with reduced cerebral blood flow (CBV divided by MTT) to show the penumbra surrounding the core area of damage.

Figure 2. The MR study shows a small DWI lesion with much larger CBF and MTT lesions in a patient with an acute stroke. The MRA shows no flow in the middle cerebral artery. A study done 3 days later (bottom row) shows recanalization of the vessel with some persistent permanent area of injury but improvement in MTT and CBF lesion volumes. DWI = diffusion-weighted imaging. Images courtesy of Colin P. Derdeyn, MD, Mallinckrodt Institute of Radiology, St Louis. (Click the image for a larger version.)

Complex changes take place both after infarction and upon reperfusion. There is a relationship between hemorrhage risk and the amount of ischemic tissue. The properties of the ideal imaging test for stroke are hemorrhage exclusion, speed, and the ability to provide both anatomical and physiological information. Clinical trials now under way are investigating the usefulness of CT, MRI, and, to a lesser extent, transcranial Doppler, PET, and single-positron emission computed tomography. Diffusion-perfusion mismatching, as detected using MRA, indicates viable tissue, but there can be perfusion processing problems, and access to MRA scanning can be difficult to obtain. CTA/CT perfusion, which correlates well with MRI diffusion-perfusion, is accurate for the diagnosis of major arterial occlusion and can show areas of irreversible tissue infarction and penumbral tissue.

K. Kyes

Kris Kyes is technical editor of Decisions in Axis Imaging News.


  1. del Zoppo GJ, Higashida RT, Furlan AJ, et al. PROACT: a phase II randomized trial of recombinant pro-urokinase by direct arterial delivery in acute middle cerebral artery stroke. Stroke. 1998;29:4-11.