点击显示 收起
the Department of Clinical Neurosciences, Foothills Medical Centre, Calgary, Alberta, Canada (A.M.D.)
the University of Calgary, Calgary, Alberta, Canada (M.D.H., P.A.B.)
Henry Ford Hospital, Detroit, Mich (B.S., S.P.)
Mount Sinai School of Medicine, New York, NY (S.R.L.).
Abstract
Background and Purpose— The importance of early ischemic change (EIC) on baseline computed tomography (CT) in the decision to thrombolyze the patient with acute ischemic stroke has been controversial. ASPECTS is a semiquantitative scale that scores the extent of EIC within the middle cerebral artery territory. We examined whether ASPECTS could be a treatment modifier by systematically reviewing the CT scans in the NINDS rtPA Stroke Study.
Methods— Six hundred eight of the 624 CT scans were available and of sufficient quality. One of 2 teams (n=3 each) of expert ASPECTS readers evaluated each scan for an ASPECTS value using a consensus score approach. Each team was blind to all clinical information except symptom side and blind to follow-up imaging and outcome information. ASPECTS values were stratified before analysis. Multivariable logistic regression was used to determine if an ASPECTS by treatment interaction existed on treatment response, outcome, and intracerebral hemorrhage risk.
Results— A total of 57.2% (348 of 608) of scans showed EIC with an ASPECTS <10. ASPECTS dichotomized into 8 to 10 and <8 did not have a treatment-modifying effect on good outcome but showed a trend to lower mortality at 90 days with tPA (relative risk 0.67, 95% confidence interval 0.41 to 1.06, P=0.10). ASPECTS 8 to 10 were associated with a trend to larger benefit of tPA with a number needed to treat (NNT) of 5 versus ASPECTS 3 to 7 with a NNT of 8.
Conclusion— There was no evidence of treatment effect modification by the baseline ASPECTS value in the NINDS rtPA Stroke Study. Therefore, exclusion of patients for thrombolysis within 3 hours of symptom onset based on EIC is not supported by our data. There is a trend to reduced mortality and increased benefit to rtPA if the baseline CT scan is favorable (ASPECTS >7).
Key Words: computed tomography diagnostic imaging ischemia thrombolytic therapy stroke, acute
Introduction
The efficacy of tissue plasminogen activator (rtPA) for acute ischemic stroke within 3 hours of symptom onset was demonstrated in the NINDS rtPA Stroke Study.1 Among multiple baseline clinical factors, only early time to treatment predicted a better outcome with treatment.2,3 Computed tomography (CT) was used to exclude intracranial hemorrhage (ICH) before the allocated treatment was administered. The extent of early ischemic change (EIC) on the baseline CT did not influence patient eligibility in the trial. Patients benefited independent of stroke severity. However, a small proportion of patients developed symptomatic intracerebral hemorrhage (ICH).
There is accumulating evidence suggesting that EIC on CT before the administration of acute stroke therapies can predict both functional outcome and the risk of ICH.4–6 However, in patients solely treated within 3 hours of symptom onset, the NINDS investigators showed a 31% EIC prevalence and EIC was not a modifier of the treatment effect. The detection of EIC was low when compared with other studies, although this may be the result of earlier CT scanning in the 0- to 3-hour window trial compared with the other 0- to 6-hour trials.7,8 Furthermore, the concept that the risk of ICH after thrombolysis is related to the degree of EIC has been recently challenged.8
We hypothesized that: (1) the use of Alberta Stroke Programme Early CT Scale (ASPECTS)5 on the baseline CT scan from The NINDS rtPA Stroke Study would increase the detection rate of EIC; and (2) a differential treatment effect with rtPA could be detected dependent on the extent of EIC in the middle cerebral artery (MCA) territory as quantified by ASPECTS.
Methods
The NINDS rtPA Stroke Study was a multicenter, prospective, double-blind, placebo-controlled, randomized trial of intravenous rtPA for acute ischemic stroke performed from January 1991 through October 1994. Methodology of the trial has been previously published.1
A noncontrast CT scan of the brain was mandatory before enrollment in the study to rule out ICH. All CT scans were performed on third- or fourth-generation CT scanners. All the baseline CT scans were obtained with 10-mm slice thickness and consisted of the following technical factors: 120kV, 170mA, matrix size 512x512.3 second scanning time for posterior fossa, and 2-second scanning time for supratentorial compartment. Window levels and window width were optimized for gray/white matter distinction. (S. Patel, personal communication, 2003)
For this study, all baseline CT scans were reevaluated retrospectively by one of 2 groups of 3 expert CT readers using ASPECTS. Follow-up CT scans were not reevaluated. Hemorrhage detection9 and infarct volume measurement (methodology described in Appendix 2 of reference 10)10 were already available from previous review of follow-up CT scans. This studies’ readers were kept blind from this and all other follow-up CT information, however. ASPECTS was assessed by systematically scoring each of 10 regions on the CT scan and assigning a score of one for a normal and 0 for a region showing signs of ischemia. Signs of ischemia are defined as x-ray hypoattenuation, loss of the gray–white boundary (which is the result of x-ray hypoattenuation of the gray matter), and/or effacement of cortical sulci. ASPECTS is only scored for acute changes. The regions are subdivided into 2 levels, which involve the basal ganglionic axial cuts and supraganglionic cuts. All axial cuts are used to examine individual regions. The individual regions include subcortical structures (lentiform nucleus, caudate nucleus, and posterior limb of internal capsule) and cortical structures (insula, M1 through M6; see Figure 1 in reference 5).5 The only clinical information given to reviewers of the scans was symptom-side. Symptom-side information was derived from the baseline NIHSS score based on side of predominant motor weakness, aphasia, or hemispatial neglect. Readers remained blinded to treatment assignment.
Review occurred over a 2-day period with all scans randomly assigned to each group of 3 readers. One member from each group switched teams the second day to ensure uniform rating between groups. Three raters agreed on a consensus ASPECTS value after discussing each scan. All members of ASPECTS review had experience using the ASPECTS methodology with a detailed tutorial performed before initiating reading.
Favorable outcome was defined as a modified Rankin scale of 0 to one at 3 months, Barthel index 95, and NIHSS 1. Symptomatic ICH was defined as neurologic deterioration within 24 hours of treatment associated with ICH.
Statistical Analysis
The primary outcomes were excellent functional outcome defined as a 90-day modified Rankin scale score less than 2, the occurrence of symptomatic ICH, and death. Secondary outcomes were excellent neurologic outcome defined as a 90-day NIHSS score less than 2 and good functional activity recovery defined as a 90-day Barthel Index score of 95 or greater. ASPECTS was dichotomized at less than or equal to 7 or greater than 7, and trichotomized into greater than 7, 3 to 7, and less than 3. The primary hypothesis was that a differential treatment effect on the primary outcome existed according to whether the baseline ASPECTS was greater than, less than, or equal to 7. The power to detect an interaction, estimated as described by Hsieh,11 with an odds ratio for the interaction term greater than or equal to 1.25 was low at 21%. Secondary analyses considered alternate outcomes using the ASPECTS dichotomy (>7/7) as well as the ASPECTS trichotomy (>7/3 to 7/<3). Trichotomizing the ASPECTS allowed for more detailed understanding of the ASPECTS 7 group. Trends in outcomes were assessed using Cuzick nonparametric test for trend across ordered groups. We used derived additive interaction terms as described by Rothman12 to help explain the relationship between ASPECTS and treatment as predictors of outcome and mortality. The use of such terms arguably makes the analysis more directly clinically relevant.
The data are described with standard descriptive statistics and stratified analysis. Logistic regression was used to assess whether an ASPECTS by treatment interaction was present for the primary outcomes using a likelihood ratio test. One-way analysis of variance was used to assess the relationship among lesion volume by ASPECTS categories. The number-needed-to-treat and the number-needed-to-harm were calculated as the inverse of the absolute risk difference accordingly, and 95% confidence intervals for each are reported.13
Results
Of the 624 patients enrolled in the trial, 620 CT scans were available for interpretation and 608 CT scans were deemed of sufficient quality to allow for reasonable interpretation of EIC. There were 308 patients treated with placebo and 300 patients treated with rtPA. No differences in baseline characteristics or outcome were detected between 16 patients excluded and 608 included in this study.
The distribution of ASPECTS values were skewed. In 43.9% (n=267) of baseline CT scans, ASPECTS values were 10 (no evidence of EIC in MCA territory). Seven of these patients had an ASPECTS value of 10 and either evidence of a new posterior cerebral artery territory (n=5) or a new brain stem stroke (n=2). Therefore, 57.2% of patients (n=348) showed evidence of EIC. The median ASPECTS value was 9 (interquartile range 7 to 10) and was the same value in both the 0 to 90 and 91 to 180 cohorts (P=0.892). The median ASPECTS value in the patients showing ischemia in greater than one third of the MCA territory was 5 (interquartile range 3 to 8) compared with 9 (interquartile range 7 to 10) in patients with ischemia in less than one third of the MCA territory (P<0.001). A weak negative correlation (P=–0.31) between ASPECTS and stroke severity (baseline NIHSS) was observed. Age and baseline serum glucose were not correlated with ASPECTS.
In the primary analysis, there was no evidence of a multiplicative interaction between the ASPECTS value and tPA treatment in a simple logistic regression model when ASPECTS was dichotomized at 7 versus >7. There was no evidence for masking confounding by the baseline NIHSS score, age, baseline glucose, or stroke onset to treatment time. This lack of multiplicative interaction was also true for additional outcomes of NIHSS 0 to one, Barthel Index >95, and occurrence of symptomatic ICH. However, the risk of death with rtPA treatment compared with placebo tended to be lower in the ASPECTS 8 to 10 group (12% vs 18%; relative risk [RR]=0.67, 95% confidence interval 0.41 to 1.06, P=0.10). No mortality benefit was observed in the rtPA arm with ASPECTS 0 to 7 group (RR=1).
To better understand this trend to lower mortality, we trichotomized the ASPECTS score into 3 groups based on extent of early change (ASPECTS >7, 3 to 7, <3). Prognostically, across the 3 groups, a lower ASPECTS category predicted poorer outcome (mRS 2 to 6) (P=0.015), increased mortality (P=0.002), but not increased symptomatic ICH (P=0.15). When death or symptomatic ICH were considered as a composite outcome, poorer outcome with lower ASPECTS category remained significant (P=0.003). In every case, rtPA treatment improved the proportion of patients achieving an excellent functional outcome. In the ASPECTS >7 group, a mortality benefit is seen with rtPA treatment and in the ASPECTS <3 group, mortality was increased with rtPA compared with placebo (Figure 1).
Examination of the absolute risk differences (Figure 1) among patients with baseline ASPECTS >7 showed improved outcomes (48% vs 29%; 3 month mRS 1) with rtPA when compared with placebo. Among patients with more extensive EIC (ASPECTS 3 to 7), improved outcome was observed (36% vs 23%; 3 month mRS 1) When the baseline CT scan revealed very extensive EIC (ASPECTS <3) no benefit from rtPA could be demonstrated (Table 1). Two of the 5 deaths in the treated group with ASPECTS <3 were associated with symptomatic ICH compared with none in the placebo group (Table 2).
An exploratory analysis using the most powerful measures of effectiveness (NIHSS 2 at 24 hours or NIHSS 15-point improvement from baseline to 24 hours) of rtPA in the NINDS rtPA Stroke Study14 revealed no evidence of a multiplicative treatment by baseline CT ASPECTS interaction when the baseline CT ASPECTS was dichotomized at 7/>7.
There was a negative but nonlinear correlation between baseline ASPECTS and final infarct volume in both the placebo and rtPA groups (r=–0.33). Final infarct volumes were different across the 3 categories of ASPECTS (P=0.002) but not according to treatment assignment (Table 3).
Discussion
The results from the present study do not support the concept that quantifying EIC using a standardized validated scoring system on the baseline noncontrast CT scan are critical to intravenous thrombolysis decision-making in the first 3 hours from acute stroke symptom onset. The thrombolytic treatment effect is robust. In this regard, our study is consistent with the recently published reevaluation of the NINDS rtPA Stroke Study for EIC on CT scans using the one third MCA rule, which found benefit from tPA even if one third MCA territory is involved.15
However, this study did reveal value in EIC detection because careful evaluation of the CT scan has useful in predicting likelihood of benefit. Those patients with ASPECTS 8 to 10 had 12% mortality in the rtPA arm and 18% in the placebo arm, suggesting a subgroup of patients with a trend to reduced mortality if rtPA administered. No mortality reduction was seen in the tPA-treated patients in the ASPECTS 3 to 7 subgroup (Figure 1). The extent of ischemic change (ASPECTS score) also predicts the likelihood of benefit with a trend to lower number needed to treat with ASPECTS >7 (ARR 19%, NNT=5) than ASPECTS 3 to 7 (ARR 13%, NNT=8). Patients with a baseline CT ASPECTS <3 had a nonsignificantly higher rate of mortality and symptomatic ICH with rtPA. This observation must be tempered by the small numbers (3% of patients enrolled in the trial), which had such extensive EIC. This study does not provide sufficient evidence to recommend that patients be excluded from rtPA on the basis of extensive EIC (ASPECTS <3).
The European Co-operative Acute Stroke Study (ECASS-1) trial16 first identified the possible importance of EIC. They identified a trend to increased mortality with tPA treatment of patients with EIC visible in greater than one third of the MCA territory in patients treated up to 6 hours after symptom onset. This led to an assumption of a qualitative interaction between treatment effect and EIC and the recommendation that patients should not receive thrombolysis if CT demonstrates early changes of a recent major infarction such as sulcal effacement, mass effect, or edema. A clear statistical interaction effect was not demonstrated, however. In fact, this study demonstrated that rtPA was associated with an increased risk for fatal brain hemorrhage both in the >one third MCA territory EIC group and normal CT scan group (without EIC).17 Based on this lack of statistical significance, the original guidelines for thrombolytic therapy for stroke did not stipulate that >one third MCA territory EIC involvement was a contraindication to rtPA.18 Other difficulties with the one third MCA rule in practice include: (1) extrapolating from the 3- to 6-hour time window to the under 3-hour time window, (2) the lack of confirmation of >one third MCA rule importance in ECASS-2 trial in which no association with increased mortality was seen with rtPA,19 and (3) demonstration by multiple groups that the one third MCA rule has poor interrater reliability.20–23 Methods to improve the one third MCA rule have resulted in better reliability but have not been tested among less experienced readers.6
ASPECTS5 was devised to provide a simple, systematic approach of assessing and improving reliability of EIC in regions of the MCA territory. ASPECTS provides a reliable semiquantitative, localization-weighted estimation of ischemic tissue volume within the MCA territory, which correlates well with functional outcome after rtPA.5,24,25 Improved EIC detection was demonstrated in this study when compared with the one third MCA territory rule. EIC on NCCT with ASPECTS also correlates well with diffusion-weighted imaging abnormalities and DWI-ASPECTS value.26 ASPECTS was recently shown to have treatment-modifying effect in patients with stroke with MCA occlusion receiving intraarterial therapy an average 5.3 hours from symptom onset. Patients who received prourokinase intraarterial treatment did better than controls only when they had a baseline ASPECTS >7. The magnitude of benefit for these patients was large with a number needed to treat of 3 to yield functional independence at 90 days. There was a consistency of treatment effect at higher ASPECTS.27 The main difference between PROACT-2 and the NINDS rtPA Stroke Study was the preselection of MCA occlusions in PROACT-2, which ensures perfusion abnormalities and penumbra in some patients.28 We have previously shown that ASPECTS value of 10 is only associated with occlusion in 15% of cases so good CT scan patients may not be ideal targets for thrombolysis when intracranial occlusion absent (no penumbra).29 In PROACT-2, all high ASPECTS value patients (small regions of core) had such MCA occlusion (large penumbral areas), thus explaining the robust effect of intraarterial prourokinase at high ASPECTS values.27
Another very plausible explanation for the lack of effect modification in the current study is that EIC may be a less critical factor for treatment decisions in the early stages of ischemia when time is the most critical factor influencing benefit from rtPA.2,30 In PROACT-2, patients were treated 3 hours later than NINDS rtPA at a time when EIC may represent a more irreversible injury.28
The relative importance of the individual characteristics of EIC (sulcal effacement, parenchymal hypoattenuation, and distortion of gray white differentiation) on CT is not known. The frequently seen subtle signs of ischemia on CT scan may not always represent irreversible injury.31 Regions of ischemic brain showing definite hypoattenuation as a manifestation of tissue edema are likely but not always irreversibly damaged (core of ischemia and therefore not salvageable with thrombolysis), whereas EIC manifested as cortical sulcal effacement or very subtle hypoattenuation as a result of hypoperfusion on baseline CT scan more likely can reverse. Recent work has suggested that sulcal effacement without hypodensity is the result of increased cerebral blood volume and reperfused areas of brain. These areas have not consistently resulted in infarction on follow-up imaging.32 Modification of ASPECTS methodology may be required to ignore the potentially reversible EIC finding of sulcal effacement. CT technology and CT imaging techniques used at the time in the NINDS rtPA Stroke Study may have limited our observations. We have previously suggested 5-mm axial cuts should be used for acute stroke CT imaging to allow visualization of all ASPECTS regions on more than one slice. This allows differentiation of hypoattenuation as a result of volume averaging of adjacent sulci rather than real EIC. The NINDS rtPA Stroke Study used 10-mm axial cuts at baseline to shorten CT imaging time, which in the early 1990s was relatively slow compared with current high-speed CT scanners (S. Patel, personal communication, 2003). Despite this limitation, the ASPECTS readers considered most available scans to be of good to excellent quality. Adequate scanning power, good window width, and window leveling for filming were used in a majority of scans.
Poor outcome still occurred in some patients despite a high ASPECTS score. In some cases, this may be because ASPECTS does not address ischemic change in other vascular territories such as the anterior cerebral artery, posterior cerebral artery, or vertebrobasilar system. Alternately, ASPECTS may not fully reflect the core of ischemia. Newer contrast-enhanced CT techniques such as blood-pool analysis from CT angiography source images33,34 or CT perfusion studies35–37 are now available using multislice CT scanners and may more accurately identify "core" and "penumbra" than conventional CT and provide better identification of salvageable ischemic brain. Similarly, multimodal magnetic resonance imaging techniques are also available as alternatives for thrombolysis decision-making.38,39
The results of the current study suggest that EIC evaluation by ASPECTS improves early ischemic change detection. Higher ASPECTS values are associated with a greater magnitude of benefit from rtPA and a trend to reduced mortality. ASPECTS does not, however, identify patients who should or should not receive thrombolytics. CT ischemic change may be more relevant for thrombolytic decision-making in the 3- to 6-hour time window after stroke onset given the results of the ECASS-1 trial and PROACT-2 ASPECTS analysis.27 Regions of CT ischemic change may become irreversibly damaged if treatment is delayed. Early treatment appears more critical to thrombolysis and overwhelms the significance of early CT changes.
References
The National Institute of Neurological Disorders and Stroke rt-PA Stroke Study Group. Tissue plasminogen activator for acute ischemic stroke. N Engl J Med. 1995; 333: 1581–1587.
Marler JR, Tilley BC, Lu M, Brott TG, Lyden PC, Grotta JC, et al. Early stroke treatment associated with better outcome: the NINDS rt-PA stroke study. Neurology. 2000; 55: 1649–1655.
The NINDS tPA Stroke Study Group. Generalized efficacy of tPA for Acute Stroke: subgroup analysis of the NINDS tPA Stroke Trial. Stroke. 1997; 28: 2119–2125.
Hacke W, Kaste M, Fieschi C, Toni D, Lesaffre E, von Kummer R, et al, for the ECASS Study Group. Intravenous thrombolysis with recombinant tissue plasminogen activator for acute hemispheric stroke: The European Cooperative Acute Stroke Study (ECASS). JAMA. 1995; 274: 1017–1025.
Barber PA, Demchuk AM, Zhang J, Buchan AM, for the ASPECTS Study Group. Validity and reliability of a quantitative computed tomography score in predicting outcome in hyperacute stroke before thrombolytic therapy. Lancet. 2000; 355: 1670–1674.
Silver B, Demaerschalk B, Merino JG, Wong E, Tamayo A, Devasenapathy A, et al. Improved outcomes in stroke thrombolysis with pre-specified imaging criteria. Can J Neurol Sci. 2001; 28: 113–119.
von Kummer R, Bourquain H, Bastianello S, Bozzao L, Manelfe C, Meier D, et al. Early prediction of irreversible brain damage after ischemic stroke at CT. Radiology. 2001; 219: 95–100.
Gilligan A, Markus R, Read S, Srikanth V, Hirano T, Fitt G, et al, Australian Streptokinase Trial Investigators. Baseline blood pressure but not early computed tomography changes predict major hemorrhage after streptokinase in acute ischemic stroke. Stroke. 2002; 33: 2236–2242.
The NINDS t-PA Stroke Study Group. Intracerebral hemorrhage after intravenous t-PA therapy for ischemic stroke. Stroke. 1997; 28: 2109–2118.
The National Institute of Neurological Disorders and Stroke (NINDS) rt-PA Stroke Study Group. Effect of intravenous recombinant tissue plasminogen activator on ischemic stroke lesion size measured by computed tomography. Stroke. 2000; 31: 2912–2919.
Hsieh FY. Sample size tables for logistic regression. Stat Med. 1989; 8: 795–802.
Rothman KJ, Greenland S. Modern Epidemiology, 2nd ed. Philadelphia: Lippincott Raven; 1998.
Altman DG. Confidence intervals for the number needed to treat. BMJ. 1998; 317: 1309–1312.
Broderick JP, Lu M, Kothari R, Levine SR, Lyden PD, Haley EC, et al. Finding the most powerful measures of the effectiveness of tissue plasminogen activator in the NINDS tPA stroke trial. Stroke. 2000; 31: 2335–2341.
Patel SC, Levine SR, Tilley BC, Grotta JC, Lu M, Frankel M, et al. Lack of clinical significance of early ischemic changes on computed tomography in acute stroke. JAMA. 2001; 286: 2830–2838.
Hacke W, Kaste M, Fieschi C, et al, for the ECASS Study Group. Intravenous thrombolysis with recombinant tissue plasminogen activator for acute hemispheric stroke: the European Cooperative Acute Stroke Study (ECASS). JAMA. 1995; 274: 1017–1025.
von Kummer R, Allen KL, Holle R, Bozzao L, Bastianello S, Manelfe C, et al. Acute stroke: usefulness of early CT findings before thrombolytic therapy. Radiology. 1997; 205: 327–333.
Adams HP Jr, Brott TG, Furlan AJ, Gomez CR, Grotta J, Helgason CM, et al. Guidelines for thrombolytic therapy for acute stroke: a supplement to the guidelines for the management of patients with acute ischemic stroke. A statement for healthcare professionals from a Special Writing Group of the Stroke Council, American Heart Association. Circulation. 1996; 94: 1167–1174.
Hacke W, Kaste M, Fieschi C, von Kummer R, Davalos A, Meier D, et al. Randomised double blind placebo-controlled trial of thrombolytic therapy with intravenous alteplase in acute ischaemic stroke (ECASS II). Lancet. 1998; 352: 1245–1251.
Dippel DWJ, Du RY, van Beest Holle M, van Kooten F, et al. The validity and reliability of early infarct signs on CT in acute ischaemic stroke. Cerebrovas Dis. 1997; 7 (suppl 4): 15.
Schriger D, Kalafut M, Starkman S, et al. Cranial computed tomography interpretation in acute stroke. Physicians accuracy in determining eligibility for thrombolytic therapy. JAMA. 1998; 279: 1293–1297.
Grotta JC, Chiu D, Lu M, Patel S, Levine SR, Tilley BC, et al. Agreement and variability in the interpretation of early CT changes in stroke patients qualifying for intravenous rtPA therapy. Stroke. 1999; 30: 1528–1533.
Wardlaw JM, Dorman PJ, Lewis SC, Sandercock PAG. Can stroke physicians and neurologists identify signs of early cerebral infarction on CT J Neurol Neurosurg Psychiatry. 1999; 67: 651–653.
Pexman JHW, Barber PA, Hill MD, Sevick RJ, Demchuk AM, Hudon M, et al. Use of the Alberta Stroke Program Early CT Score (ASPECTS) for assessing CT scans in patients with acute stroke. AJNR Am J Neuroradiol. 2001; 22: 1534–1542.
Coutts SB, Demchuk AM, Barber PA, Hu WY, Simon JE, Buchan AM, et al, VISION Study Group. Interobserver variation of ASPECTS in real time. Stroke. 2004; 35: e103–105.
Barber PA, Hill MD, Eliasziw M, Demchuk AM, Pexman W, Hudon M, et al. A comparative analysis of CT with magnetic resonance diffusion weighted imaging in hyperacute stroke. J Neurol Neurosurg Psychiatry. In press.
Hill MD, Rowley HA, Adler F, Eliasziw M, Furlan A, Higashida RT, et al, for the PROACT-II Investigators. Selection of acute ischemic stroke patients for intra-arterial thrombolysis with pro-urokinase by using ASPECTS. Stroke. 2003; 34: 1925–1931.
Furlan A, Higashida R, Wechsler L, Gent M, Rowley H, Kase C, et al. Intra-arterial prourokinase for acute ischemic stroke. The PROACT II study: a randomized controlled trial. Prolyse in acute cerebral thromboembolism. JAMA. 1999; 282: 2003–2011.
Barber PA, Demchuk AM, Hill MD, Pexman JH, Hudon ME, Frayne R, et al. The probability of middle cerebral artery MRA flow signal abnormality with quantified CT ischaemic change: targets for future therapeutic studies. J Neurol Neurosurg Psychiatry. 2004; 75: 1426–1430.
Hacke W, Donnan G, Fieschi C, Kaste M, von Kummer R, Broderick JP, et al, ATLANTIS Trials Investigators, ECASS Trials Investigators, NINDS rt-PA Study Group Investigators. Association of outcome with early stroke treatment: pooled analysis of ATLANTIS, ECASS, and NINDS rt-PA stroke trials. Lancet. 2004; 363: 768–774.
Jaillard A, Hommel M, Baird AE, Linfante I, Llinas RH, Caplan LR, et al. Significance of early CT signs in acute stroke: a CT scan-diffusion MRI study. Cerebrovasc Dis. 2002; 13: 47–56.
Butcher KS, Lee SB, Parsons M, Fink J, Levi C, Allport L, et al, for the EPITHET Investigators. Increased blood volume maintains viability in tissue with isolated focal swelling on CT in acute stroke . Stroke. 2005; 36: 418.
Ezzeddine MA, Lev MH, McDonald CT, Rordorf G, Oliveira-Filho J, Aksoy FG, et al. CT angiography with whole brain perfused blood volume imaging: added clinical value in the assessment of acute stroke. Stroke. 2002; 33: 959–966.
Coutts SB, Lev MH, Eliasziw M, Roccatagliata L, Hill MD, Schwamm LH, et al. ASPECTS on CTA source images versus unenhanced CT: added value in predicting final infarct extent and clinical outcome. Stroke. 2004; 35: 2472–2476.
Wintermark M, Reichhart M, Thiran JP, Maeder P, Chalaron M, Schnyder P, et al. Prognostic accuracy of cerebral blood flow measurement by perfusion computed tomography, at the time of emergency room admission, in acute stroke patients. Ann Neurol. 2002; 51: 417–432.
Koenig M, Kraus M, Theek C, Klotz E, Gehlen W, Heuser L. Quantitative assessment of the ischemic brain by means of perfusion-related parameters derived from perfusion CT. Stroke. 2001; 32: 431–437.
Wintermark M, Fischbein NJ, Smith WS, Ko NU, Quist M, Dillon WP. Accuracy of dynamic perfusion CT with deconvolution in detecting acute hemispheric stroke. AJNR Am J Neuroradiol. 2005; 26: 104–112.
Schellinger PD, Fiebach JB, Hacke W. Imaging-based decision making in thrombolytic therapy for ischemic stroke: present status. Stroke. 2003; 34: 575–583.
Chalela JA, Kang DW, Luby M, Ezzeddine M, Latour LL, Todd JW, et al. Early magnetic resonance imaging findings in patients receiving tissue plasminogen activator predict outcome: insights into the pathophysiology of acute stroke in the thrombolysis era. Ann Neurol. 2004; 55: 105–112.