Does Arterial Recanalization Improve Outcome in Carotid Territory Stroke?
Background and Purpose We sought to determine whether early (<8 hours) or delayed (8 to 24 hours) recanalization after stroke may be an independent variable in the improvement of clinical outcome in patients with occlusion of the middle cerebral artery.
Methods We prospectively studied 77 patients by combined Scandinavian Stroke Scale score at admission, repeated computed tomography and angiography before and after thrombolytic treatment at <8 hours after stroke onset, and transcranial Doppler ultrasound 24 hours later. We tested an association between clinical and neuroradiological baseline characteristics, recanalization, and outcome as assessed by the modified Rankin Scale 4 weeks after stroke and determined the effect of recanalization on mortality and good outcome (Rankin Scale grades 0 to 3) by multiple logistic regression analyses.
Results Recanalization rates at 8 and 24 hours after stroke correlated with sites of occlusion (middle cerebral artery branch, 73% and 73%; trunk, 27% and 38%, respectively; intracranial internal carotid artery bifurcation, 14% and 14%; P=.002), collaterals (good, 43% and 51%, respectively; scarce, 17% and 19%, respectively; P=.01), and Scandinavian Stroke Scale score at admission (P=.002). Six of 7 patients with delayed recanalization had good outcomes. Recanalization at <8 hours after symptom onset had no independent predictive value for good outcome (P=.69). Recanalization at 24 hours increased the proportion of good outcomes from 23% to 75% in a subgroup of patients. Recanalization did not independently affect mortality (P>.15).
Conclusions Even if delayed, arterial recanalization may improve clinical outcome in a subgroup of patients with middle cerebral artery occlusion.
The final goal of thrombolytic therapy in acute stroke is the rescue of brain tissue from ischemic damage to preserve neuronal function. In theory, the most important step in this therapeutic strategy is recanalization of occluded intracranial arteries to restore cerebral blood flow (CBF). However, arterial recanalization is not at all times associated with neuronal recovery. First, arterial recanalization does not mean tissue reperfusion in each patient and for each part of ischemic brain tissue volume. Migration of emboli, secondary thrombosis, or swelling of the intima may cause a no-reflow phenomenon.1 On the other hand, reperfusion of ischemic brain tissue may involve the risk of reperfusion injury.2 Enhancement of focal CBF by recanalization may rescue brain tissue at risk3 4 5 but is ineffective in areas where severe ischemia has already destroyed brain tissue. Moreover, thrombolytic agents themselves or recanalization could provoke cerebral bleeding. All these events could counteract the theoretical clinical benefit of recanalization. However, meta-analysis of previous experience has suggested that thrombolytic therapy in acute stroke may be associated with a significant reduction in the odds of death and deterioration and does not excessively enhance the risk of hemorrhage or severe edema.6 Recanalization of arterial occlusion in the carotid territory is associated with a reduction in mortality from 34% to 13% and an increase in patients with favorable outcome from 13% to 65%.7
Thus far the association between thrombolytic therapy and prognosis has been studied in small and mostly uncontrolled trials that were not able to control baseline prognostic differences.8 9 10 11 12 13 14 The relation of arterial recanalization per se to clinical improvement was not presented in those reports. Recent experience has suggested that recanalization is easier to achieve in smaller arteries and in the presence of good leptomeningeal collateral blood flow.15 16 17 Both conditions will affect clinical outcome independently.18 Therefore, it is still not clear whether arterial recanalization can independently influence clinical course after stroke. To answer this question we prospectively studied acute stroke patients with repeated computed tomography (CT), angiography, and transcranial Doppler ultrasound (TCD) before and after thrombolytic therapy and performed multiple logistic regression analyses to determine the impact of recanalization on clinical outcome compared with other variables. In this report we present our observations on patients with occlusion of the middle cerebral artery (MCA).
Subjects and Methods
From February 1988 to October 1993, we recruited 77 consecutive patients (30 women, 47 men; age range, 28 to 78 years; mean age, 57 years) with acute hemispheric stroke for thrombolytic therapy. This cohort includes 32 patients reported on earlier in this journal17 and 53 patients whose diagnostic CT scans were analyzed regarding their prognostic value.19
All patients were admitted to our hospital within 4 hours after the onset of hemiparesis. By following a standardized protocol, we documented each patient’s clinical status by neurological examination at admission and after 4 weeks using all items of the combined Scandinavian Stroke Scale (SSS)20 (maximum [best] score=58) and the modified Rankin Scale (RS)21 (death score=6). Inclusion criteria for this study were the following: (1) a precisely defined and witnessed onset of hemiparesis within the previous 6 hours before thrombolytic treatment; (2) a technically satisfactory CT scan before thrombolytic treatment without evidence of cerebral hemorrhage; (3) occlusion of the intracranial internal carotid artery (ICA) bifurcation including the proximal MCA, of the MCA trunk, or of one major MCA branch; and (4) informed consent from the patient or his or her relatives for angiography and thrombolytic treatment.
The first CT study was performed 32 to 311 minutes (median, 120 minutes; mean, 136±64 minutes) after symptom onset. All CT scans were unenhanced and obtained on a Picker 1200 SX (Picker International) scanner with a slice thickness of 8 mm throughout the brain. The CT scans were examined prospectively by a neuroradiologist who was aware of the clinical status of the patient. The interpreter determined regions of parenchymal radiolucency with an arterial distribution and focal brain swelling (effacement of sulci, compression of ventricle). The size of a hypodense area was categorized as small if it covered <50% of the presumed MCA territory and large if it covered ≥50%. A second CT scan was obtained 1 to 8 days after therapy to show the size of the infarct and/or hyperdense lesion consistent with intracerebral hemorrhage. The presence of hemorrhage was defined as hemorrhagic infarction (HI) or parenchymal hematoma (PH) according to criteria described by Pessin et al.22
Within 6 hours after the onset of symptoms an arterial digital subtraction angiogram of the symptomatic and contralateral ICA was performed. The site and extent of arterial occlusion and collateral blood supply were assessed semiquantitatively by a neuroradiologist, who graded the collateral circulation as “scarce” when no leptomeningeal collaterals or only few collaterals with slow flow were seen and “good” in the case of many leptomeningeal collaterals from the anterior cerebral artery. A second angiogram performed immediately after thrombolytic treatment (<8 hours after symptom onset) determined recanalization of the MCA. We used Thrombolysis in Myocardial Infarction (TIMI) flow grades23 to assess recanalization and grouped together patients with no perfusion and penetration with minimal perfusion (TIMI grades 0 and 1) of the MCA in the no-recanalization group and patients displaying partial or complete perfusion (TIMI grades 2 and 3) of the MCA in the recanalization group.
Patients were treated with intra-arterial urokinase (n=8; 0.5 to 1.5 MIU), intra-arterial recombinant tissue plasminogen activator (rTPA) (n=8; 22 to 100 mg), or intravenous rTPA (n=61; 100 mg) over 90 minutes or less starting within 6 hours after symptom onset. Simultaneous with the thrombolytic agent, we injected a bolus of heparin 5000 IU IV and continued heparinization by infusion of 1000 to 1500 IU/h, aiming to double the activated partial thromboplastin time in each patient. Heparinization was discontinued when the second CT revealed any evidence of parenchymal hemorrhage.
Recanalization was again assessed by TCD 24 hours later. Clinical outcome was defined as poor in patients with an RS score of 4 or 5 determined 4 weeks after stroke and as good in patients with an RS score of 0 to 3. Death rates were assessed separately. We were not able to blind the clinical examiners to the results of angiography before and after thrombolytic therapy.
The association between clinical and neuroradiological baseline characteristics and arterial recanalization was tested by χ2 test, Fisher’s exact test, unpaired t test, Mann-Whitney U test, Kruskal-Wallis rank test, and Spearman’s rank correlation. Two multiple logistic regression analyses were performed to select minimum sets of baseline variables that enabled one to identify patients with good (versus poor) and with fatal (versus other) clinical outcome. Based on these regression models, two simple prognostic scores were defined and then used to analyze and describe the additional prognostic effect of recanalization.
Table 1⇓ shows clinical and neuroradiological baseline characteristics.
Diagnostic CT was normal in 15 patients (19%) and showed small and large areas of parenchymal hypodensity in 39 patients (51%) and 23 patients (30%), respectively. All normal CT scans were performed within the first 132 minutes after the onset of symptoms. With good collaterals, no or small parenchymal hypodensity was more often seen than large hypodensity (P<.005). The extent of parenchymal hypodensity did not correlate significantly with the site of arterial occlusion (P=.46, Fisher’s exact test).
Twenty-two patients (29%) had partial or complete recanalization of the MCA at <8 hours and 26 patients (34%) at 8 to 24 hours after symptom onset. The comparison between patients in different recanalization groups showed the following (Table 1⇑): Patients with recanalization (at either 8 or 24 hours) had a significantly better SSS score at admission, more often showed MCA branch occlusions than MCA trunk or ICA/MCA occlusions, and more often had good than scarce collaterals. Recanalization at <8 hours seemed unrelated to age, sex, interval to treatment, or extent of parenchymal hypodensity or brain swelling as evidenced by diagnostic CT. Recanalization at 8 to 24 hours was more often observed with earlier treatment, in women, and with no or small parenchymal hypodensity as shown by first CT scan.
TCD detected reocclusion in 3 patients and delayed recanalization between 8 and 24 hours after symptom onset in 7 patients. Table 2⇓ lists the conditions associated with reocclusion and delayed recanalization. Interestingly, only 1 of 7 patients with delayed recanalization showed brain hemorrhage (HI), whereas all 3 patients with reocclusion had brain hemorrhage (two PH, one HI). Clinical outcome was favorable in most of the patients with delayed recanalization.
Follow-up CT showed HI more often in patients with recanalization at <8 hours (P=.011) and at 8 to 24 hours (P=.042) after stroke compared with patients without recanalization at those time points (Table 1⇑). No positive correlation was found between recanalization and PH. Clinical outcome was unaffected by HI (P=.96, U test) but seemed worse with PH (P=.008, U test) (Table 3⇓).
Clinical outcome did not correlate significantly with age and sex of the patients and was unaffected by the time interval between onset of symptoms and start of therapy in this group of patients. Clinical outcome was better in patients with better clinical condition at admission as measured by the SSS, in patients showing no or small parenchymal hypodensity and no brain swelling by diagnostic CT, in patients with occlusion of smaller arteries and with good collaterals, and in patients showing recanalization at 8 and 24 hours after stroke (Table 3⇑).
The extent of parenchymal hypodensity shown on CT scans within the first 6 hours after the onset of symptoms was the most important predictor of good clinical outcome: No single patient with a hypodense area ≥50% of the MCA territory on the initial CT reached an RS score ≤3 at 4 weeks after stroke. From the other variables examined, multiple logistic regression analysis selected SSS score at admission >25 (P=.018), age younger than 51 years (P=.006), and site of occlusion (P=.029) as the most significant variables to identify patients with good outcome (RS score 0 to 3). Based on the regression model, a simple score was defined by assessing 2 points for MCA branch occlusion and 1 point for MCA trunk occlusion, age younger than 51 years, SSS score at admission >25, and parenchymal hypodensity <50% of MCA territory. These points were summed for the total score value, which was, however, set to 0 for hypodensity ≥50% of MCA territory. We found that all patients scoring 3 to 5 had an 88% chance to be only moderately or minimally disabled 4 weeks after stroke (Fig 1⇓), whereas patients scoring 0 to 2 had only a 19% chance for good outcome. The sensitivity of a score of 3 to 5 for good clinical outcome was .71, and the specificity was .92.
Recanalization at <8 hours had no independent prognostic value for good clinical outcome (P=.64) in our logistic regression model. In contrast, recanalization at 8 to 24 hours after symptom onset as assessed by TCD provided significant additional information regarding the chance of good outcome (P=.012). The effect of recanalization 8 to 24 hours after symptom onset is shown in Fig 2⇓. In patients scoring 2 and 3, the chance for favorable outcome at 4 weeks after stroke increased with recanalization from 23% to 75%. Patients with low scores of 0 and 1 had a low probability of both recanalization at 24 hours after symptom onset and good outcome. Patients with high scores of 4 and 5 had a high probability of both recanalization and subsequent good outcome.
Multiple logistic regression analysis selected parenchymal hypodensity ≥50% of MCA territory (P=.0002), scarce collaterals (P=.009), and site of occlusion (P=.017) as the most significant variables to identify patients with fatal outcome. The effect of recanalization at <8 hours and 8 to 24 hours on mortality was not significant (P=.51 and P=.16, respectively). Similarly, we defined a second score to predict mortality within 4 weeks of stroke by assessing 2 points for MCA branch occlusion and <50% parenchymal hypodensity and 1 point for MCA trunk occlusion and good collaterals. We found that patients with a sum score of 0 to 2 had a mortality of 57% or greater (Fig 3⇓). The sensitivity of this finding was .92, the specificity was .85, the positive predictive value was .96, and the negative predictive value was .74. All patients with ICA/MCA occlusion, scarce collaterals, and large parenchymal hypodensity on the first CT (sum score of 0) died within the first 4 weeks after stroke.
This study did not primarily address the question of by which means arterial recanalization can best be achieved. Only controlled randomized trials can answer this question.24 We learned, however, that the frequency of recanalization was higher with smaller arteries and good collaterals, confirming observations by others.9 15 16 Consequently, we found that patients displaying recanalization of the MCA at 8 or 24 hours after the onset of symptoms were in better clinical condition at admission than patients without recanalization, and patients with large parenchymal hypodensity as shown by first CT had a smaller chance for arterial recanalization at 24 hours after stroke. SSS score at admission, extent of parenchymal hypodensity and brain swelling on diagnostic CT, site of occlusion, and state of collaterals correlated significantly with clinical outcome 4 weeks after stroke. It is therefore questionable whether the relation between arterial recanalization and good clinical outcome is causal or epiphenomenal.
Based on a prospective protocol of thrombolytic treatment in acute hemispheric stroke with different agents and modes of application, we performed this post hoc analysis to determine the individual prognostic value of variables other than recanalization and to determine under which conditions arterial recanalization—if achieved after thrombolysis—has an independent effect on clinical outcome. We thought that such an analysis could be helpful in designing protocols of future randomized trials of thrombolytic treatment in stroke.
We assessed clinical outcome by the modified RS 4 weeks after stroke; this scale allows determination of the patient’s independence with satisfactory interobserver agreement.21 However, this assessment could not take into account events that occurred after 4 weeks and that affected long-term clinical outcome and was probably biased by the examiner’s awareness of the result of angiography. In view of the severity of the clinical presentation at admission (mean±SD SSS score, 20±9), we subdivided the cohort into deceased patients and patients who were or were not able to walk without assistance 4 weeks after stroke. These three groups had significantly different mean SSS scores at this time. In our opinion these assessments were adequate to describe the probable clinical effects of arterial recanalization after MCA stroke.
Multiple logistic regression analyses assessed the independent prognostic value of arterial recanalization compared with other variables examined. Recanalization at <8 hours after symptom onset did not significantly affect mortality and the proportion of patients with good outcome 4 weeks after stroke. In contrast, patients with recanalization at 8 to 24 hours after symptom onset had a 33% to 54% better chance for good outcome than patients without recanalization if their sum score of important variables (parenchymal hypodensity, SSS score at admission, age, site of occlusion) was 2 or 3. However, recanalization at 8 to 24 hours was not associated with reduced mortality.
These analyses suggest that on the basis of our data it was not possible to assess the effect of thrombolytic recanalization on good outcome in those subgroups of patients who had certain unfavorable or favorable preconditions. No more than 6 of 34 patients (18%) with a sum score of 0 or 1 for good clinical outcome (44% of all patients studied) showed recanalization, and only 1 patient of this group who reached an RS score <4 after 4 weeks did not show recanalization. However, there was a nonsignificant trend toward lower mortality (33% versus 68%) for patients with recanalization in this group. In contrast, only 1 of 9 patients with prognostic score values of 4 or 5 (12% of all patients studied) showed no recanalization but nevertheless good outcome, similar to the other 8 patients. These observations suggest that patients of the subgroup with unfavorable preconditions may not profit from thrombolytic therapy in terms of morbidity but probably will profit in terms of mortality. For those patients with good prognostic score values, recanalization and subsequent good outcome was almost always achieved. Therefore, it is difficult to determine whether recanalization was a precondition or epiphenomenon for favorable prognosis.
The negative correlation between artery size and frequency of recanalization could be due to systemic and not local application of the thrombolytic agent. We treated most of our patients (79%) intravenously. The frequencies of recanalization as assessed by angiography within 8 hours of symptom onset were within the range of other studies that used intravenous treatment13 16 25 26 but lower than studies that used intra-arterial application of the thrombolytic agent in the carotid territory.8 9 10 11 12 14 With intra-arterial thrombolytic treatment in our study, the recanalization rate at 8 hours after symptom onset was only 19% (3/16). This relatively low recanalization rate may be explained by the high proportion (13/16) of patients with combined ICA/MCA occlusion we selected for this mode of application in an effort to increase the recanalization rate.
The size of an occluded artery reflects the size of thromboembolus and the volume of ischemic tissue. We presume that the size of a thromboembolus is related to its composition in most instances. It is likely that nonorganized and noncalcified (ie, fibrin-rich, less rigid, and probably already fragmented) emboli will have a better chance to pass the intracranial ICA bifurcation and even the MCA trunk and will have a better chance to resolve spontaneously or by applied agents. Leptomeningeal anastomoses can allow retrograde blood flow into the distal part of occluded arteries, thereby reducing the tissue volume of severe ischemia27 28 and enabling the thrombolytic agent, if given systemically, to reach the surface of the clot at both ends, which explains the higher frequency of recanalization.15 Both the size of the occluded artery and the state of collaterals may be important preconditions for the clinical condition of the patient, the clinical outcome, and arterial recanalization, and these two factors seem to be primarily responsible for significant correlations between recanalization and clinical outcome.
Reocclusion occurred in only three patients (4%) and was associated with hemorrhagic transformation in all patients and with fatal or poor outcome in two patients. A remarkable and somewhat surprising finding was that delayed recanalization at 8 to 24 hours after the onset of symptoms was associated with good outcome in six of seven patients (86%). This subgroup of patients contributed significantly to the observation that recanalization at 24 hours after symptom onset seemed to be more relevant for clinical outcome than recanalization at <8 hours. We presume that collateral blood supply was sufficient >8 hours after MCA occlusion to keep considerable amounts of brain tissue viable. HI occurred in only one patient of this group. Thus, we did not get the impression that the theoretical concept of reperfusion injury2 is valid under clinical conditions even in the case of delayed recanalization after thrombolytic treatment in MCA occlusion. We confirmed that HI in contrast to PH is not associated with clinical deterioration, ie, it may be possible to differentiate between clinically relevant and irrelevant types of postischemic brain hemorrhages by CT criteria.22
In conclusion, our observations challenge the theory of a 6-hour time window for therapy in ischemic stroke. We observed that arterial recanalization at <8 hours after symptom onset was hard to achieve under the adverse conditions of scarce collaterals and occlusion of the intracranial internal carotid bifurcation. The independent effect of recanalization within this period on clinical outcome could not be assessed, probably because the site of occlusion and collateral blood supply primarily determined the extent of severe ischemia. In this group of patients, the effect of recanalization on clinical outcome could be studied only if higher frequencies of recanalization were achieved. We did not get the impression that intra-arterial application of the thrombolytic agent was superior to systemic application in this respect. We also showed that arterial recanalization improved clinical outcome under more favorable conditions even when it occurred >8 hours after symptom onset. That means that in patients with sufficient collateral blood supply of viable ischemic brain tissue, recanalization and restoration of orthograde flow may help brain tissue at risk to survive and to regain neuronal function.
From this post hoc analysis, we hypothesize that the therapeutic window in ischemic stroke has two sides. One side is determined by the extent of primary ischemia and may already be closed when the patient first presents. The second side is open for an uncertain period. Collateral blood flow and cerebral oxygen supply determine this period, which is jeopardized by each drop in perfusion pressure or decrease in arterial oxygen content but can persist for hours (or perhaps days) under stable conditions. Regarding the primarily unstable conditions in the cerebral circulation after cerebral artery occlusion, the second side of the therapeutic time window should be used for therapeutic approaches as soon as possible. The site of arterial occlusion, state of collaterals, and extent of parenchymal hypodensity by CT should be considered when designing trials to assess the effect of recanalizing therapies in acute stroke.
We gratefully acknowledge the advice and assistance of Klaus Sartor, MD, who is responsible for the favorable conditions under which we could perform this study. We thank our friend Gregory J. del Zoppo for many discussions and for his careful review of this manuscript. We would also like to thank our colleagues in the Departments of Neurology and Neuroradiology, who cared for our acute stroke patients.
Reprint requests to Prof Dr R. von Kummer, Abt Klinische Neuroradiologie, Im Neuenheimer Feld 400, D-69120 Heidelberg, Germany.
- Received November 29, 1994.
- Revision received January 20, 1995.
- Accepted January 21, 1995.
- Copyright © 1995 by American Heart Association
Crowell RM, Olsson Y. Impaired microvascular filling after focal cerebral ischemia in the monkey: modification by treatment. Neurology. 1972;22:710-719.
Bednar MM, McAuliffe T, Raymond S, Gross CE. Tissue plasminogen activator reduces brain injury in a rabbit model of thromboembolic stroke. Stroke. 1992;21:1705-1709.
Wardlaw JM, Warlow CP. Thrombolysis in acute ischemic stroke: does it work? Stroke. 1992;23:1826-1839.
del Zoppo GJ, Ferbert A, Otis S, Brückmann H, Hacke W, Zyroff J, Harker LA, Zeumer H. Local intra-arterial fibrinolytic therapy in acute carotid territory stroke. Stroke. 1988;19:307-313.
Mori E, Tabuchi M, Yoshida T, Yamadori A. Intracarotid urokinase with thromboembolic occlusion of the middle cerebral artery. Stroke. 1988;19:802-812.
Theron J, Courtheoux P, Casaco A, Alachkar F, Notan F, Ganem F, Maiza D. Local intraarterial fibrinolysis in the carotid territory. AJNR Am J Neuroradiol. 1989;10:753-765.
Matsumoto K, Satoh K. Topical intraarterial urokinase infusion for acute stroke. In: Hacke W, del Zoppo GJ, Hirschberg M, eds. Thrombolytic Therapy in Acute Ischemic Stroke. Heidelberg, Germany: Springer; 1991:207-212.
Siepmann G, Müller-Jensen M, Goossens H, Lachenmeyer L, Zeumer H. Local intra-arterial fibrinolysis in acute middle cerebral artery occlusion. Neuroradiology. 1991;33:69-71.
Mori E, Yoneda Y, Tabuchi M, Yoshida T, Ohkawa S, Ohsumi Y, Kitano K, Tsutsumi A, Yamadori A. Intravenous recombinant tissue plasminogen activator in acute carotid artery territory stroke. Neurology. 1992;42:976-982.
Ringelstein EB, Biniek R, Weiller C, Ammeling B, Nolte PN, Thron A. Type and extent of hemispheric brain infarctions and clinical outcome in early and delayed middle cerebral artery recanalization. Neurology. 1992;42:289-298.
del Zoppo GJ, Poeck K, Pessin MS, Wolpert SM, Furlan AJ, Ferbert A, Alberts MJ, Zivin J, Wechsler L, Busse O, Greenlee R, Brass L, Mohr JP, Feldmann E, Hacke W, Kase CS, Biller J, Gress D, Otis SM. Recombinant tissue plasminogen activator in acute thrombotic and embolic stroke. Ann Neurol. 1992;32:78-86.
von Kummer R, Hacke W. Safety and efficacy of intravenous tissue plasminogen activator and heparin in acute middle cerebral artery stroke. Stroke. 1992;23:646-652.
Furlan AJ. Natural history of atherothrombotic occlusion of cerebral arteries: carotid versus vertebrobasilar territories. In: Hacke W, del Zoppo GJ, Hirschberg M, eds. Thrombolytic Therapy in Acute Ischemic Stroke. Heidelberg, Germany: Springer; 1991:3-8.
von Kummer R, Meyding-Lamadé U, Forsting M, Rosin L, Rieke K, Hacke W, Sartor K. Sensitivity and prognostic value of early computed tomography in middle cerebral artery trunk occlusion. AJNR Am J Neuroradiol. 1994;15:9-15.
Scandinavian Stroke Study Group. Multicenter trial of hemodilution in ischemic stroke: background and study protocol. Stroke. 1985;16:885-890.
van Swieten JC, Koudstaal PJ, Visser MC, Schouten HJA, van Gijn J. Interobserver agreement for the assessment of handicap in stroke patients. Stroke. 1988;19:604-607.
del Zoppo GJ, Higashida RT, Furlan AJ. The case for a phase III trial of cerebral intraarterial fibrinolysis. AJNR Am J Neuroradiol. 1994;15:1217-1222.
Yamaguchi T, Hayakawa T, Kikichi H, Abe T. Thrombolytic therapy in embolic and thrombotic cerebral infarction: a cooperative study. In: Hacke W, del Zoppo GJ, Hirschberg M, eds. Thrombolytic Therapy in Acute Ischemic Stroke. Heidelberg, Germany: Springer; 1991:168-174.
Yamaguchi T, Hayakawa T, Kikichi H, and the Japanese Thrombolysis Study Group. Intravenous tissue plasminogen activator in acute thromboembolic stroke: a placebo-controlled, double blind trial. In: del Zoppo GJ, Mori E, Hacke W, eds. Thrombolytic Therapy in Acute Ischemic Stroke II. Heidelberg, Germany: Springer; 1993:59-65.
Saito I, Segawa H, Shiokawa Y, Taniguchi M, Tsutsumi K. Middle cerebral artery occlusion: correlation of computed tomography with clinical outcome. Stroke. 1987;18:863-868.
Bozzao L, Fantozzi LM, Bastianello S, Bozzao A, Fieschi C. Early collateral blood supply and late parenchymal brain damage in patients with middle cerebral artery occlusion. Stroke. 1989;20:735-740.