Magnetic resonance imaging and magnetic resonance angiographic views of a patient (patient 1) with a right-sided middle cerebral artery (MCA) territory infarction (A-D) and a patient (patient 2) with a right-sided subcortical infarction (E-H). Patient 1: The MCA territory infarction (A) had a trace of longer and more dominant signal intensity of the ipsilateral posterior cerebral artery than the contralateral one (arrow) (B). Despite thrombolysis treatment, the lesion grew considerably (C) and the occluded MCA was not recanalized (D). Patient 2: The patient with right-sided subcortical infarction (E) did not have a dominant ipsilateral posterior cerebral artery (F). After thrombolysis, the lesion grew only a little (G), and the occluded MCA was recanalized (arrow) (H).
Kwon J, Kwon SU, Lee JH, Choi CG, Suh DC, Kim JS. Factors Affecting the Angiographic Recanalization and Early Clinical Improvement in Middle Cerebral Artery Territory Infarction After Thrombolysis. Arch Neurol. 2004;61(11):1682-1686. doi:10.1001/archneur.61.11.1682
Factors affecting the angiographic recanalization (AR) and clinical improvement (CI) still remain unclear in patients receiving thrombolytic therapy.
To elucidate factors related to AR and early CI in patients with middle cerebral artery (MCA) or internal carotid artery (ICA) occlusion.
Department of Neurology, Asan Medical Center, Seoul, South Korea.
We studied 42 patients who (1) underwent diffusion-weighted magnetic resonance (MR) imaging and MR angiography within 6 hours after onset, (2) had MCA territory infarction, (3) had nonvisualization of the MCA or the ICA on initial MR angiography, (4) were treated with thrombolytics, and (5) underwent follow-up MR imaging and MR angiography at day 2 or 3.
Successful AR and CI were achieved in 31 and 16 patients, respectively. Angiographic recanalization was related to CI (P<.01), lower follow-up National Institutes of Health Stroke Scale scores (P<.05), the absence of a dominant ipsilateral posterior cerebral artery (P<.01) on initial MR angiography, and the sparing of the internal capsule on both initial (P<.05) and follow-up (P<.01) MR imaging. Clinical improvement was associated with the absence of ICA (vs MCA) flow signals (P<.05), the sparing of the internal capsule (P<.01), and marginally, with the infarct volume change (P = .06).
In patients with MCA or ICA occlusion, CI after thrombolysis is related to the AR and the sparing of the critical motor pathway. The presence of a dominant ipsilateral posterior cerebral artery may predict poor AR after thrombolysis.
Diffusion-weighted magnetic resonance imaging (DWI), perfusion-weighted imaging, and magnetic resonance angiography (MRA) help us to better predict the angiographic recanalization (AR) or clinical improvement (CI) in patients receiving thrombolytic therapy.1,2 Small initial DWI lesions,3 occlusion of distal vessels,4,5 and the presence of good collateral flow6 have been shown to correlate with AR or CI. However, factors related to AR and CI still remain insufficiently identified, especially in patients having occlusion of either the proximal middle cerebral artery (MCA) or the internal carotid artery (ICA).5,7
From the stroke registry of the Asan Medical Center, Seoul, South Korea, we retrospectively analyzed all consecutive patients with MCA territory infarction who underwent thrombolytic therapy in the acute stage (ie, within <6 hours of symptom onset) between January 15, 2000, and June 22, 2002. However, 10 patients who were treated between 6 and 8 hours for various reasons were also included.
Other inclusion criteria were the patients (1) whose National Institutes of Health Stroke Scale (NIHSS) scores were higher than 4 points, (2) who underwent DWI and MRA less than 6 hours after symptom onset, (3) whose initial MRA showed nonvisualization of the MCA or ICA signals that explains the patient’s clinical symptoms, and (4) who underwent follow-up magnetic resonance imaging (MRI)/MRA at day 2 or 3. We excluded the patients who (1) had preservation of MCA flow via collateral channels despite the occlusion of the ICA, (2) had CI before the initiation of thrombolytic therapy, (3) showed a DWI-identified lesion greater than two thirds of the total MCA territory, and (4) had occlusion of the M2 or M3 segment of the MCA.
All MRI and MRAs were obtained with a 1.5-T, whole-body MR scanner equipped with echo planar imaging capability (Signa Cvi; GE Medical Systems, Milwaukee, Wis). The Asan Medical Center protocol for patients with a hyperacute stroke included an axial T2-weighted, fast-spin echo sequence (repetition time/echo time, 4500/118 ms; 20 segments covering the whole brain; a 5-mm segment thickness; and a 2-mm intersegment gap, with a 25-cm field of view), diffusion-weighted trace sequence (single-shot, echo-planar spin-echo sequence, b = 1000 s/mm2, repetition time/echo time, 5000/139 ms; 20 segments covering the whole brain; a 5-mm segment thickness; and a 2-mm intersegment gap, with a 25-cm field of view), and a time-of-flight MRA covering the circle of Willis (repetition time/echo time, 35/6.9 ms; 25° flip angle; 50-mm single slab, with 36 partitions). Follow-up MRI/MRA on day 2 or 3 included gadolinium-enhanced 3-dimensional MRA covering neck vessels (repetition time/echo time, 3.2/1.2 ms; 30° flip angle; a 60-mm single slab, with 36 partitions; and a 26-cm field of view) in addition to a series of MRIs identical to the initial ones.
Of 137 patients who were treated with thrombolytics, 91 had MCA territory infarction. Fifty patients were excluded who did not undergo baseline or follow-up MRI/MRA; 29 initially underwent brain computed tomography, and 20 did not undergo follow-up MRI/MRA because of their poor clinical condition (n = 11) or the patients’ or relatives’ reluctance (n = 9). Therefore, the remaining 42 patients became the subjects of this study.
We made it a rule to give urokinase (UK) intra-arterially (IA) via a femoral artery approach in patients with ICA or M1 occlusion whenever conventional angiography was immediately available. When it was not, intravenous (IV) recombinant tissue plasminogen activator (r-tPA) (0.9 mg/kg) was given if the patients were initially seen within less than 3 hours of symptom onset. However, IV r-tPA was occasionally used in selected patients who arrived at 3 to 6 hours of symptom onset.
The initial dose of IA urokinase was either 100 000 or 200 000 U, and repeat doses up to 1 million units were given to achieve AR. Whenever possible, mechanical disruption of the clot with the microwire and microcatheter was attempted before and/or after administration of the thrombolytic agent. After confirming the absence of hemorrhage using computed tomography, we started heparinization shortly afterward in patients receiving IA urokinase, and 24 hours afterward in those treated with IV r-tPA. Written consent was obtained from all the patients or relatives before thrombolytic therapy was started.
Angiographic recanalization was defined as a restoration of flow past the area of occlusion on a follow-up MRA. Definite CI was defined as a decrease of 4-points on the patient’s NIHSS score at 48 to 72 hours compared with the baseline NIHSS score. Factors affecting the AR and CI were assessed, which included age, sex, the presence of vascular risk factors, the presence of atherosclerotic vessels other than the target vessels, initial blood pressure, initial serum glucose level, interval between stroke onset and treatment, the laterality of the lesion, the involvement of the internal capsule on initial and follow-up MRIs, and the presence of a dominant ipsilateral posterior cerebral artery (DIPCA) on the initial MRA.
The dominant ipsilateral posterior cerebral artery was defined as a trace of longer and more prominent signal intensity of the ipsilateral posterior cerebral artery than that shown on the contralateral side (Figure). The presence of DIPCA and the involvement of the internal capsule were assessed by 2 neurologists (J.-H.K. and S.U.K.), who were blinded to the patients’ clinical information. If there was a disagreement, a third neurologist (J.S.K.) made a final decision. To measure the infarct volume, the margin of the lesion shown by DWI was outlined and the volume was calculated using an Asan Medical Center picture-archiving computerized system by a person blinded to the clinical information. The difference in lesion volume between the lesion shown in the follow-up DWI and that shown in the initial DWI was designated as “volume difference.”
Using SPSS version 10.0 software (SPSS Inc, Chicago, Ill), the Fisher exact and χ2 tests were used for categorical variables, and the Mann-Whitney test was used for continuous parameters when data were deviated from normal distribution. All statistical tests were 2-tailed; P<.05 was considered statistically significant.
The initial MRI/MRA was done at a mean of 2.9 hours after symptom onset. Treatment modalities included IV r-tPA in 22 patients (r-tPA only in 20 patients, r-tPA and stent placement in 2 patients) and IA UK in 20 patients (IA UK only in 12 patients; IA UK preceded by IV r-tPA, 0.6 mg/kg, in 2 patients; and IA UK and additional stent or angioplasty in 6 patients). Thrombolytic therapy was initiated at a mean [SD] of 4.6 hours (IV r-tPA, 3.4 [0.9] hours; IA UK, 5.2 [2.3] hours; P = .009). There was no difference in the baseline NIHSS score between the patients treated with IV r-tPA (median, 12.5; range, 3-20) and the patients treated with IA UK (median, 12; range, 4-24). The initial mean (SD) lesion volume was not different between the 2 groups, either 28.6 (36.3) cm3 or 29.2 (43.2) cm3, respectively.
In follow-up MRA, 31 patients showed AR whereas 11 did not (Table 1). Patients with AR had significantly lower follow-up NIHSS scores, less often had DIPCA and internal capsule involvement, and more often had CI than those without AR. Regardless of AR, all patients showed increased lesion volumes in the follow-up MRI.
Sixteen patients had CI while 26 patients had either a stationary (n = 22) or an aggravated clinical course (n = 4, defined as an increase of >2 points on the NIHSS score at 72 hours) (Table 2). Patients with CI significantly more often had AR, less often had MCA occlusion (vs ICA occlusion), and less often had internal capsule involvement than those without CI. Patients with CI had a tendency to have a smaller lesion volume difference than those without CI (P = .06).
Because of differences in therapeutic strategies (ie, additional mechanical intervention) and the timing of the initiation of therapy, we cannot compare the therapeutic efficacy between IV r-tPA and IA UK. Nevertheless, similar to previous results,1,2,5,6 our data clearly showed that AR after thrombolysis is related to CI. However, the volume difference between the final volume and the initial one was only marginally associated with CI and was unrelated to AR. Interestingly, both AR and CI were significantly related to the sparing of the internal capsule in the initial as well as the follow-up MRI. Thus, initial involvement of the critical motor pathway seems to be one of the predictors for poor prognosis in patients receiving thrombolysis.
We also found that the presence of a DIPCA was negatively related to AR. Considering that vessel signals on time-of-flight MRA are volume dependent,8 the DIPCA may reflect an asymmetrically increased flow of the ipsilateral posterior cerebral artery to compensate for the shortage of MCA blood flows. Although the DIPCA may be caused by an acute compensation secondary to the occlusion of the carotid system, the persistent presence of DIPCA on a follow-up MRA (data not shown) illustrates that DIPCA had been present before the occurrence of stroke. Therefore, the presence of a DIPCA suggests that the pathogenic mechanism of infarction may be either artery-to-artery embolism or in situ MCA atherothrombotic occlusion in the setting of previously existing severe ICA or MCA diseases. The latter possibility seems to be particularly important in Asian countries where the frequency of intrinsic intracranial atherosclerosis is relatively high.9 Because in situ atherothrombosis or artery-to-artery embolism is less susceptible for thrombolysis than emboli from the heart,10 the presence of a DIPCA in an initial MRA may be used as one of the markers for an unfavorable sign for AR.
Unexpectedly, we found that ICA occlusion (vs MCA) was a favorable sign for CI, an observation at odds with the previous observation that proximal vessel occlusion is less often associated with AR or CI than distal artery occlusion.1,4- 6 This may be explained by our exclusion of the patients with M2 or M3 segment occlusion who generally have shown an excellent clinical outcome after thrombolysis.4 Alternatively, relatively frequent mechanical intervention may have improved the outcome of the patients with proximal ICA occlusion. However, our data may be an artifact caused by the exclusion of 11 patients who did not undergo follow-up MRI/MRA owing to poor clinical outcome. Among them, 6 had ICA occlusion while 5 had MCA occlusion. Inclusion of these patients as a “no CI group” and reanalysis of the final 53 patients yielded that the site of occlusion was unrelated to the prognosis (P = .13). Thus, the relationship between the site of vascular occlusion and the CI must again be investigated in the future study using larger samples.
There are limitations in our study. First, we excluded the patients who did not undergo a follow-up MRI/MRA because AR could not be assessed in them. Therefore, our results, especially the AR rate cannot be applied to the general stroke population. Second, because the presence of a DIPCA was assessed with MRA but not conventional angiography, collateral circulation of the patients with a DIPCA could not be accurately assessed. Third, some of the results of our study, such as those related to the DIPCA, may not be applicable in non-Asian populations where intrinsic MCA diseases are uncommon.
Despite these limitations, our study showed that CI after thrombolysis is related to the AR and the sparing of the critical motor pathway. The presence of a DIPCA and involvement of the internal capsule in the early MRI/MRA could be used as unfavorable signs against successful thrombolysis.
Correspondence: Jong S. Kim, MD, Department of Neurology, Asan Medical Center, Song-Pa PO Box 145, Seoul 138-600, South Korea (firstname.lastname@example.org).
Accepted for Publication: May 20, 2004.
Author Contributions:Study concept and design: Kim. Acquisition of data: J.-H. Kwon, S. U. Kwon, Lee, Choi, and Suh. Drafting of the manuscript: J.-H. Kwon, Lee, Choi, and Suh. Critical revision of the manuscript for important intellectual content: S. U. Kwon, and Kim. Statistical analysis: S. U. Kwon, and Lee. Administrative, technical, and material support: J.-H. Kwon, Choi, Suh, and Kim.
Acknowledgment: This study was supported by research fund 03-PJ1-PG1-CH06-0001 from the Korean Ministry of Health and Welfare, Seoul, South Korea.