|
Neurology India, Vol. 50, (Suppl. 1), Dec, 2002, pp. S30-S36 Imaging in Acute Ischemic Stroke : Relevance to Management M. Moonis Department of
Neurology,
University of Massachusetts Medical School,
Worcester MA 01655, USA. Code Number: ni02160 Summary With the advent of thrombolytic therapy in the treatment of acute ischemic stroke, it has become increasingly important to identify the suitable patients for whom such therapy may be useful. The success of reperfusion therapy depends on salvaging ischemic tissue at risk (penumbra). Imaging techniques continue to evolve. MRI with diffusion weighted and perfusion imaging can identify the penumbra (diffusion-perfusion mismatch). MR Angiography provides additional information about large and medium size vessel occlusion. However, MRI is limited by its lesser availability and slower acquisition times. Ultrafast perfusion CT scans are more widely available and seem capable of identifying the ischemic tissue at risk. Newer techniques of perfusion CT and triphasic perfusion CT are becoming more refined and along with CT Angiography provide information not only of the penumbra but also of large vessel occlusion. Patients with large vessel occlusion of the internal carotid and middle cerebral artery are best treated by intra-arterial thrombolytic therapy whereas branch occlusions are suitable for intravenous thrombolysis. Patient selection, based on the present and evolving MRI and CT techniques would provide a more rational application of treatment (greater chances of reperfusion and minimizing the possibility of symptomatic intracerebral hemorrhage). Key words : Thrombolytic therapy, MR angiography, CT angiography. Introduction Increasing refinements in stroke management with acute thrombolytic therapy and trials of neuroprotective therapy have been paralleled by advances in neuroimaging techniques. Rapid improvement in the scanning times and more sensitive methods of defining acute ischemia continue to evolve in computerized tomography (CT) and magnetic resonance imaging (MRI) scans. The most significant advances have been in the field of newer MRI techniques; diffusion weighted imaging (DWI) with the ability to define ischemic events very early and perfusion imaging (PI) which helps to define the volume of brain tissue at risk.1-6With the evolution of ultrafast CT scanners, perfusion CT and CT angiography are becoming important tools in acute ischemic stroke management. Current techniques seem capable of defining the ischemic penumbra as well as the site of large vessel arterial occlusion.7 Ischemic tissue accumulates intracellular water and this forms the basis of echoplanar DWI. The result of accumulation of intracellular water is reduction of the apparent diffusion co-efficient (ADC) with a hypointense signal on the ADC maps and a hyperintense signal on the DWI, reflective of cytotoxic edema. The combination is specific to ischemic tissue and is useful in separating other conditions that appear hyperintense on DWI. These include seizure foci, transient global ischemia, migraine, multiple sclerosis and certain tumors.3,8,9 Unlike ischemic infarcts, ADC values are not affected in these other conditions. Although most DWI changes reflect irreversible ischemia, several animal and clinical studies suggest that these deficits at least early on in the ischemic process may be partially reversible. Kidwell et al found that of the 9 patients with transient ischemic attacks and early DWI hyperintensity, almost half had resolution of their symptoms on follow-up studies indicating that the initial DWI lesions may not always reflect irreversible infarction.10 Current techniques of PI involve either administration of an exogenous bolus of a contrast agent that shortens T2 (BT-PI) or arterial spin labeling techniques (ASLPI). BT-PI utilizes serial multislice T2WI to monitor the signal reduction associated with the passage of the bolus. This is used to generate time curves of signal intensity change that can then be used to estimate the time to peak (TTP), relative mean transit time (MTT), relative cerebral blood flow (CBF) and cerebral blood volume (CBV). Of these, TTP and MTT are utilized to estimate the ischemic area at risk.11-14 PI abnormalities are a composite of the area of infarction and the area at risk of infarction (penumbra). When used with DWI, PI deficit outside the area of DWI hyperintensity approximates the ischemic penumbra. This is potentially salvageable tissue and typically in untreated patients the area of infarction (DWI) will evolve into the region of the PI deficit.15,16 Evolution of thrombolytic therapy has revolutionized stroke management. Patients with acute ischemic stroke treated with thrombolytics intravenously within 3 hours and intraarterially within 6 hours have a 30% better relative outcome as compared to patients who did not receive this treatment.17-20 Patients treated within the first ninety minutes show a better recovery than those treated in the 90-180 minute period.21 However, the number of patients treated thus far is disappointingly low (10-12%). While part of the problem is in delay in arriving at the hospital, we are probably not treating all patients who may be appropriate for treatment.22 Selection of appropriate patients based on clinical examination and a CT scan can be misleading.23 For rational thrombolytic therapy, imaging tools need to be sensitive (identify ischemic infarcts) and specific (exclude other causes of acute neurological deficits and hemorrhage). Furthermore, it should differentiate large vessel and cardioembolic small vessel strokes from lacunar infarcts and other non-embolic strokes. DWI-PI in identification of lesions suitable for thrombolysis Detection of acute ischemic infarcts: DWI is a very sensitive technique in identifying acute ischemic infarcts. DWI changes in ischemic tissue may be evident within minutes both in experimental and clinical studies. The changes may persist up to a week or even longer in some studies.24-26 In a study of subacute ischemic strokes, persistent abnormalities on DWI were seen in more patients than on conventional MRI after a period of 6-10 days. Furthermore DWI provided information not evident on conventional MRI in 30% of the patients.27 In a large prospective study DWI increased the sensitivity of the diagnosis from 48 to 94%. Similar results have been reported by several other studies comparing DWI with CT or conventional T2 weighted MRI imaging. Since DWI usually represents irreversible damage (infarction), it gives valuable information on the size of the stroke at a time when thrombolysis is being considered. Currently, DWI hyperintensity greater than one-third of the arterial territory would be a contraindication to thrombolytic therapy.28-30 In patients with previous strokes DWI is advantageous in differentiating acute from chronic lesions.31-33 Fitzek reported a sensitivity and specificity approaching 100% in patients with at least one previous imaging abnormality.33 In our experience, 9 patients imaged within the first 48 hours had a negative CT scan while all patients demonstrated a DWI abnormality. In 45% of cases DWI changed the original diagnosis to ischemic small vessel stroke, which influenced subsequent management. Differentiation of stroke sub-types : Lacunar vs Embolic : Even though the National Institute of Neurological disorders and Stroke (NINDS) trial had demonstrated an equal benefit of thrombolytic therapy for small vessel strokes, lacunar strokes remain a relative contraindication to thrombolytic therapy because of their spontaneous potential of complete recovery and the risk of intracranial hemorrhage with rt-PA use. However, clinical differentiation of large vessel from small vessel strokes may be inaccurate in 20-30% of cases. The sensitivity of the diagnosis is increased with DWI compared to conventional CT and T2 weighted MRI techniques. This is highlighted in several studies that demonstrate that up to 50% of clinically diagnosed small vessel strokes evolved into large cortical strokes.8,9,23,34-36 Furthermore, location of the DWI hyperintensity is helpful in delineating the extent of the irreversible ischemic lesion and by their location differentiate small vessel from large vessel stroke.27 In a prospective study of stroke patients presenting with a clinically well-defined lacunar syndrome, the authors found multiple, simultaneous DWI abnormalities in nearly 16% of cases. One out of 6 cases that appear to be lacunar stroke on clinical evaluation may actually be embolic and possible candidate for thrombolytic therapy.37 While perfusion CT scans and CT Angiograms (CTA) continue to evolve with faster scan times and more accurate delineation of the area of potential infarction, these techniques have not been widely tested against DWIPI.38,39 Identification of watershed infarcts : DWI has been helpful in defining border zone infarcts that usually result from hypotensive crisis and not thromboembolism. Clinical and early CT diagnosis of watershed infarcts is notoriously inaccurate. DWI on the other hand identifies these lesions and combined with PI helps to differentiate these from similar lesions resulting from thromboembolism or critical carotid artery disease. Matching multiple DWI-PI abnormalities correlated with embolic strokes while abnormal perfusion with normal DWI correlated with transient hypoperfusion. Extensive PI deficits in one or more arterial territories suggested superimposed severe carotid artery disease.11 Combined DWI-PI imaging correlated well with angiographic lesions during intra-arterial therapy.36 PI imaging in itself detects early, potentially reversible ischemic changes in stroke and the extent of these deficits suggests whether the main stem or a branch occlusion has occurred. Branch occlusions correlate with a smaller PI deficit. The estimated lesion volumes correlate well with the scores on National Institute of Health Stroke Scale (NIHSS) and the Rankin Scale (RS).40,41 Detection of Intracerebral hemorrhage : CT scan has been believed to be a more sensitive tool for detection of intracerebral hemorrhage, which is an absolute contraindication to thrombolytic therapy. However, recent techniques (echoplanar gradient recall echo sequences) have been shown to detect hemorrhage with comparable accuracy.14,42-44 In a prospective study of patients with primary intracerebral hemorrhage, Roob et al found MRI evidence of additional micro bleeds in 54% of all cases. In the majority of patients micro bleeds were located simultaneously in multiple brain regions.42 Utility of a combined DWI-PI in acute thrombolytic therapy Although not universally accepted because of time limitation and a reported study indicating reversibility of DWI lesions within 24 hours in some cases, DWIPW imaging offers a powerful combination for rational thrombolytic therapy in acute ischemic stroke. Furthermore it is increasingly becoming possible to image patients within 3-6 hours of onset of their symptoms.45 Arterial occlusion results in 3 zones in the involved territory. The central core of irreversibly infarcted tissue, the surrounding critically perfused potentially viable and outer zone of normally perfused tissue. This middle zone of critically perfused tissue; the ischemic penumbra is potentially salvageable tissue if reperfusion occurs in time.3,46,47 There may be a differential susceptibly of sub-cortical and posterior circulation compared to the large anterior circulation occlusions. Ideally, before attempted re-perfusion, it is important to identify the penumbra i.e. that an area of salvageable tissue is present since there is a substantial iatrogenic risk of causing intracerebral hemorrhage. This may be achieved with DWI-PI imaging. Perfusion deficits define the potential area of infarction while to a large extent DWI identifies the area of irreversible infarction. Over time as evident from several large clinical series the DWI deficit expands to match the perfusion deficit.45 Treatment with thrombolytic therapy should be instituted when there is a PI>DWI volume or an isolated perfusion deficit (DWI-PI mismatch). On the other hand, a DWI-PI match would imply a completed stroke and little role for thrombolysis. Even though studies of thrombolytic intervention indicate a definite time window for intervention; up to 3 hours for intravenous and up to 6 hours for intra-arterial thrombolysis there are subgroups that may fall outside this range. While the second European study ECASS-II revealed no overall benefit of rt-PA intravenously at 6 hours, it also demonstrated no significant increase in the risk of hemorrhage as compared to 3 hours. Given this information, patients who fall outside the time frames specified, but still demonstrate a DWI-PI mismatch should be candidates for thrombolysis.43,45,48 Role of MRI in assessment of reperfusion following thrombolysis Thrombolytic therapy can only be useful if there is effective re-perfusion established. Transcranial Doppler (TCD) has been utilized as a non-invasive tool to determine reperfusion flow. Results of several studies reveal a better outcome in patients with reestablished flow and correlates well with the NIHSS as well as functional scales.49,50 More recently, DWIPI has been shown to be a sensitive technique in determining reperfusion. After successful reperfusion, the final PI deficits are less than or equal to the initial DWI hyperintensity and correlate with functional outcome.40 DWI: Predicting hemorrhagic transformation Hemorrhagic transformation of an embolic stroke is a common phenomenon seen to some degree in 50-60% of all embolic strokes. However, clinically significant hemorrhagic transformation is associated with a worse outcome. While studies of contrast extravasation on CT and perfusion deficits on HMPOA-SPECT provide some measure of the possible risk, DWI is beginning to have greater utility in this area.51-53 In a prospective study, Tong et al assessed the utility of DWI in predicting hemorrhagic conversion. The authors analyzed 17 patients with DWI within 8 hours and at 1 week after the onset of an acute ischemic stroke. ADC for each pixel of the whole ischemic area was calculated. ADC values below 550 x 10-6 correlated well with subsequent hemorrhagic transformation. The authors did not describe the clinical outcome in these cases. Nevertheless, if these results are consistently replicated in other studies, they may provide yet another tool to assess the safety and tailor thrombolytic therapy.54 Other studies have shown DWI volumes to be predictive of the long term functional outcome and correlate with NIH stroke scale and Rankin Scale.27,46-47,55 Can MRI and MRA replace preoperative conventional angiogram? Conventional angiograms in the best of centers report a 1% morbidity and mortality. However, most vascular surgeons would like to obtain information about the extent of hemodynamic compromise and the cross collateral flow to the compromised side before planning a carotid endarterectomy(CEA). BT-PI can be utilized to obtain the same information noninvasively. The MTT, TTP, CBF and CBV are compared from side to side. On the side of the critical stenosis, MTT and TTP are increased and CBF and CBV reduced. With perfusion MRA, this allows determination of the extent of cross collateral flow. This technique if validated by the larger experience would allow for non-invasive pre-operative assessment.56-58 TCD monitoring during CEA reveals multiple echogenic signals indicative of multiple upstream emboli during the procedure. DWI has been utilized to assess the extent of cerebral microembolism following a CEA. In 53 consecutive patients with CEA who had no apparent postoperative neurological deficits intraoperative TCD detected silent cerebral emboli in 17%; however, subsequent infarcts as evidenced by DWI hyperintensity were found in only 1%, suggesting that most microembolic signals detected on TCD are not associated with cerebral infarction.59 Role of perfusion CT and CT angiography in acute stroke management CT scan with its simplicity, availability and low cost is always an attractive alternative to MRI scan. Unenhanced CT scans are relatively insensitive to identifying infarcted tissue in the early period and cannot identify the tissue at risk (penumbra). More recently, perfusion (contrast enhanced CT) with rapid image acquisition made possible with recent scanners, especially helical CT are actually able to trace the contrast entry into the acquired brain slices and the mean transit time (MTT) can be calculated. MTT is equal to cerebral blood flow (CBV) / cerebral blood volume (CBF). The linear relationship between dye concentration and signal intensity is an advantage over MR perfusion where this relationship is nonlinear and leads to some overestimation of CBF.60 In a recent study by Max Wiltermark and colleagues, CBV and CBF were measured on perfusion CT with rapid slice acquisition on 4 contiguous 10 mm slices. Results were compared to delayed MRI. CBV of < 2.5ml/100g was associated with irreversible infarction. Tissue with CBF < 34% of the contralateral side was considered as tissue at risk of infarction (equivalent of the penumbra identified on DWI-PI mismatch). Reperfusion was associated with near resolution of this impaired CBF and significant improvement in the NIHSS. The limitations of this technique are assessment based on selected slices and radiation exposure and an indirect comparison with DWI-PI MRI imaging. However, radiation exposure was not greater than 2 non-contrast CT scans and probably not significant in the overall scheme.7 Another technique, triphasic CT (TPCT) where after injection of a contrast bolus, helical ultrafast CT scanning is done to delineate the early, middle and late phases of contrast enhancement. The procedure requires approximately 5 minutes and a special software program. Perfusion deficits are determined in the 3 phases of TPCT and graded as severe or moderate. Severe perfusion deficits refer to decreased attenuation compared to the contralateral side with no or few leptomeningeal collaterals visible on the any of the 3 phases of imaging. Moderate perfusion deficits referred to either a) decreased attenuation on the early phase and normal attenuation on the middle and late phases and slow collateral filling on the middle and late phases or b) decreased attenuation on the early and middle phases and normal attenuation on the late phase with slow collateral filling. The extent of moderate and severe perfusion deficits were quantified to a) less than 1/3 of the presumed middle cerebral artery territory, b) less than 50% of the MCA territory and c) greater than 50% of the MCA territory. The authors posited that the zone of severe perfusion deficit was equal to the ischemic core and the zone of moderate deficit equal to the penumbra. This then represented the tissue at risk and a target for thrombolytic therapy. If this is confirmed, it may provide an alternate rapid method of determining the penumbra and more rational application of thrombolytic therapy.61 Future directions Magnetic resonance spectroscopy (MRS) is another promising technique in the evaluation of acute stroke and assessment of patients for thrombolytic therapy. MRS is a technique of measuring metabolite peaks in applied pixels on the MRI. The common peaks are lactate, N-Acelylaspartate (NAA), Creatinine and Choline. The ratio of the NAA to lactate peaks is important in assessment of acute stroke. NAA peak is reduced with irreversible neuronal damage while lactate increases after ischemic injury and may return to normal with early re-perfusion. Animal studies indicate that a mismatch between NAA and lactate (reduced or normal NAA and elevated lactate) would suggest an incomplete lesion (ischemic penumbra) and may be another tool to assess patients for reperfusion therapy.30 In a prospective study Wild et al performed MRS in addition to T2W MRI in 11 patients imaged 24 to 72 hours later. In the center of the T2 lesions, there was a sharp reduction in the NAA peak, the intermediate zone; presumably the penumbra revealed a lesser reduction and the NAA peaks were normal beyond the margins of the lesion. Unfortunately, NAA could not be correlated to DWI in this series.62 Other MRI techniques are finding application in the evaluation of acute stroke. Kamran et al have described a hyperintense vessel sign (HSV) on fluid attenuated inversion recovery (FLAIR). The authors correlated these findings with cerebral angiography in selected cases. HSV was an early MRI sign in 10% and indicated the need for reperfusion therapy. There was a correlation with the extent of HSV sign, NIHSS and functional outcome.63 Diffusion tensor imaging (DTI) is a MRI technique that allows visualization of the DWI hyperintensity as well as its directionality. DTI has been applied to the visualization of cerebral diaschisis as well as activation of alternative regions after an acute stroke. This has the potential to provide prognostic information and along with functional MRI, promises to be a tool in assessment of long-term recovery after a stroke.64 References
Copyright 2002 - Neurology India. Also available online at http://www.neurologyindia.com |
|