Background Clinical response immediately after revascularisation therapy differs among patients. Although reperfusion is the deciding factor with respect to this dramatic response to revascularisation therapy, the influence of pre- and post-treatment diffusion–perfusion status on the speed and degree of recovery are unknown.
Methods Consecutive stroke patients who were eligible for revascularisation therapy underwent serial diffusion–perfusion MRI. Tmax perfusion maps were generated, and stroke severity and recovery were assessed up to day 90. The relationship of diffusion and perfusion lesion indices with the speed and degree of recovery were evaluated.
Results 69 patients (42 men; aged 66.3±15.9 years) were included; National Institutes of Health Stroke Scale (NIHSS) score was 13.3±6.4 points. 19 received intravenous tissue plasminogen activator (tPA) and 50 received endovascular therapy with/without intravenous tPA. Early dramatic improvement (NIHSS score reduction of ≥40% within 24 h) was observed in 24 (34.8%) patients. Among the other 45 patients, 18 (40%) showed good outcomes (modified Rankin score 0–2 at day 90), suggesting delayed recovery. The volume of post-treatment perfusion delay was similar between the early and delayed recovery groups (p=0.329) but smaller than in the non-responders group (p<0.05). Multivariate testing revealed that smaller post-treatment perfusion delay volumes were independently associated with both early dramatic improvement and delayed recovery. In addition, initial diffusion weighted imaging lesion volume was smaller in the former than in the latter (p=0.029) and was independently associated with early dramatic recovery.
Conclusions A significant proportion of patients with a lack of early dramatic improvement (40%) showed delayed recovery. Both pretreatment infarct volume and post-treatment reperfusion correlated with the degree and speed of recovery.
Statistics from Altmetric.com
Arterial revascularisation to restore perfusion to the ischaemic territory remains the principal therapeutic approach for acute ischaemic stroke. However, not all patients show dramatic improvement after revascularisation therapy, and the treatment response is usually unpredictable in clinical practice. Alexandrov et al reported that a significant proportion of patients with a lack of early clinical improvement after 24 h of thrombolysis still achieved good outcomes at 3 months, suggesting the possibility of a ‘stunned brain syndrome’ with delayed recovery.1
Most previous studies have employed serial transcranial Doppler examination and focused on the influence of revascularisation on early and delayed recovery. Recently, multiparametric MRI, including diffusion (DWI) and perfusion weighted imaging (PWI) has increasingly been used in clinical practice to select patients for revascularisation therapy.2–5 However, relatively little attention has been devoted to the impact of pre- and post-treatment diffusion–perfusion status on the speed of recovery after revascularisation therapy in patients who suffer from acute ischaemic stroke.
Therefore, in the present study, we evaluated the relationship between pre- and post-treatment MRI parameters and the speed and degree of recovery after revascularisation therapy using prospectively registered data from two centres.
Patients and methods
We analysed demographic, clinical, laboratory and radiographic data collected prospectively from consecutive patients who were eligible for revascularisation therapy for acute cerebral ischaemia at a North American centre (Los Angeles, California, USA) from September 2005 to April 2007, and at an Asian centre (Seoul, Korea) from May 2006 to June 2008. Patients were included in this study if: (1) they presented with symptoms of acute cerebral ischaemia within the middle cerebral artery territory, (2) they were eligible for revascularisation therapy and (3) pre- and post-treatment DWI and PWI were performed. The local institutional review boards approved the study.
MRI methods and image analysis
All patients underwent MRI using a 1.5 T MRI machine (Siemens Medical Systems, Iselin, New Jersey, USA) at a North American centre or a 3.0 T MRI machine (Achieva, Philips Medical Systems, Best, The Netherlands) at an Asian centre. In both centres the typical stroke MRI protocol consisted of DWI, gradient recalled echo, fluid attenuated inversion recovery, T2* perfusion weighted imaging and magnetic resonance angiography (MRA) of the cervical and intracranial vessels (three-dimensional time of flight MRA and contrast enhanced MRA, including extracranial internal carotid and vertebral artery). DWI was obtained at two levels of diffusion sensitisation (b values of 0 and 1000 s/mm2; 5–7 mm slice thickness; and no gap). PWI was conducted using a timed contrast bolus passage technique (0.1 mg/kg contrast administered into an antecubital vein with a power injector at a rate of 5 cm3/s); PWI of 1.5 T performed with the following parameters: TR 2000 ms; TE 60 ms; flip angle 90°; matrix 128×128, FOV 24 cm; section thickness 5–7 mm; and intersection gap 2 mm, while parameters of 3T PWI were as follows: TR 1500 ms; TE 35 ms; flip angle 40°; matrix 128×128, FOV 24 cm; section thickness 5 mm; and intersection gap 2 mm.
Perfusion delay was defined based on the perfusion parameter, Tmax, which is the time to the peak MR signal intensity change after deconvolution. Based on previous studies, Tmax perfusion lesion maps were generated by deconvolution of an arterial input function and tissue concentration curves.6 Data analysis was performed using the Stroke Cerebral Analysis 2 (SCAN 2) software package, which was developed inhouse. This software used the Interactive Data Language produced by ITT Visual Systems (Boulder, Colorado, USA). MRI volume measurements were made by one of the authors who was blinded to the clinical information associated with the patient. For each patient, DWI and PWI lesion volumes were automatically outlined with subsequent manual corrections. We measured DWI lesion volume using visual inspection, and ADC thresholding was not used in this study. Volumes were calculated using a computer assisted volumetric analysis program (Medical Image Processing, Analysis and Visualisation, V.2.1, CIT, NIH).
Patients were evaluated in standardised fashion based on demographic data, medical history, vascular risk factors and the National Institutes of Health Stroke Scale (NIHSS) score. All patients underwent routine blood tests, electrocardiography, cardiac telemetry for at least 24 h and echocardiography. The stroke mechanism categories were classified using a modified TOAST classification.7
All patients received revascularisation therapy (intravenous or intra-arterial thrombolytic therapy, endovascular mechanical clot retrieval and/or angioplasty with/without stenting). Patients who presented within 3 h of the onset of stroke symptoms received intravenous tissue plasminogen activator (tPA). Endovascular treatment was performed in (a) patients who were ineligible for treatment with tPA, such as those who presented later than 3 h after the onset of symptoms or those who were prone to bleed, and (b) when the attending stroke neurologists and interventional neuroradiologists felt that endovascular treatment was needed after treatment with intravenous tPA.
The NIHSS score was serially evaluated at baseline and at 1, 2, 3, 5 and 7 days after treatment. Patients were subsequently evaluated in the outpatient clinic. The outcome was investigated by determining the modified Rankin score (mRS) at 90 days after stroke onset and was defined as good if the mRS value was 0–2.8 9 Dramatic improvement was defined as an NIHSS score reduction of ≥40%.10 Patients were defined as having early dramatic recovery if dramatic improvement was observed within 24 h of stroke onset. Correspondingly, those patients who did not show early dramatic recovery yet showed good outcomes at day 90 were classified as having delayed recovery.
We analysed the differences between groups using Pearson's χ2 test, the Student t test or one-way analysis of variance (ANOVA) for means of normally distributed variables. The Mann–Whitney U test or Kruskall–Wallis test was used to analyse the medians. Independent factors for patients that showed early dramatic recovery or delayed recovery were evaluated using logistic regression. Age, NIHSS at admission, atrial fibrillation, pretreatment DWI lesion and perfusion delay (volume of Tmax ≥2 s and Tmax ≥8 s), and immediate post-treatment perfusion delay (Tmax ≥2 s) were entered into a stepwise logistic regression model with clinical outcomes as the dependent variable, in which entry was set at a univariate association with a probability value of ≤0.2. All statistical analyses were performed using commercially available software (SPSS for Windows, V.13.0; SPSS Inc). A p value <0.05 was considered statistically significant.
Of 151 patients who underwent revascularisation for acute cerebral ischaemia within the middle cerebral artery MCA territory during the study period, 69 were included in this study. The 82 patients that were not included were excluded for the following reasons: (a) MRI was contraindicated (n=12), (b) pre- or post-treatment PWI sequences were not performed (generally due to patient movement) (n=29), (c) failure of MRI post-processing due to excessive patient motion or absence of an identifiable, technically adequate arterial input function (n=7), (d) serial NIHSS or mRS at day 90 was not assessed (n=24) and (e) premorbid mRS>1, recurrent stroke during the follow-up period or undergoing craniectomy (n=10). The patients evaluated in the study included 42 South Koreans (mean age 62.2; 64.3% male) and 27 Southern Californians (mean age 72.8; 55.6% male; 20 whites). Nineteen patients were atherosclerotic, 36 cardioembolic and 14 other or undetermined causes. Pretreatment MRI was conducted at 3.0±1.4 h (range 0.7–8.1 h) and revascularisation therapy was performed at 4.3±2.1 h (range 1.0–9.9 h) after the last known well time. Nineteen patients only received intravenous thrombolysis while 27 received endovascular therapy only and 23 received a combination of thrombolysis and endovascular therapy. Post-treatment MRI was conducted in all patients at 10.7±9.2 h (range 1.2–29.2 h) after revascularisation therapy.
Among the 69 patients included in the study, early dramatic recovery was observed in 24 (34.8%). An example of serial diffusion–perfusion MRI is shown in figure 1. Of the remaining 45 patients, 27 (60%) exhibited poor outcomes on day 90 (mRS ≥3) while 18 (40%) exhibited good outcomes, suggesting that delayed recovery without early dramatic improvement had occurred. In patients with delayed recovery, the NIHSS reduction of ≥40% was achieved throughout the follow-up period up to day 90. Specifically, recovery occurred in six patients between 24 h and 72 h, in seven patients between 72 h and 7 days and in five patients between day 7 and day 90.
The differences in patient characteristics among treatment response groups are shown in table 1 and figure 2. Clinical characteristics including age, gender and risk factor profiles did not differ between patients who did and did not show early dramatic improvement. The NIHSS score on admission was lower in patients that showed early dramatic improvement than in those that did not (p=0.005). Additionally, the treatment modality differed among groups, with endovascular treatment being performed more frequently in patients that did not undergo early dramatic recovery than in those that did (p=0.015). The time interval between the onset of symptoms and start of revascularisation therapy did not differ among the groups. The volume of the pretreatment DWI lesion was lower in patients that showed early dramatic improvement (5.4±6.0 ml) than in those that did not (35.1±41.5 ml) (p<0.001). Utilisation of receiver operating characteristic curves (not shown) revealed that the pretreatment DWI lesion volumes of >13.8 ml forecast a high probability of the absence of early dramatic improvement after revascularisation therapy, yielding a high sensitivity of 95.8% (95% CI 78.8 to 99.3%) but a low specificity of 48.9% (95% CI 33.7 to 64.2%), probably due to a small number of cohort. The volumes of pretreatment perfusion delay did not differ among groups (figure 2B). Serial MRI studies revealed that the volume of post-treatment Tmax >2 s perfusion delay was smaller in patients that showed early dramatic improvement (28.9±39.2 ml) than in patients that did not (81.0±84.1 ml) (p=0.001).
Among patients who did not show early dramatic improvement, age and NIHSS score on admission were lower in the delayed recovery group than in the no recovery group (p<0.01 in both cases). Additionally, both the volumes of the initial DWI lesions and post-treatment Tmax >2 s perfusion delay were smaller in the delayed recovery group than in the no recovery group (p<0.05 in both cases) (table 1, figure 2).
Multivariate testing showed that a smaller pretreatment DWI lesion and post-treatment Tmax >2 s perfusion delay was independently associated with both early dramatic improvement and delayed recovery with lack of early dramatic improvement (p<0.05 in all cases) (table 2). In addition, atrial fibrillation was independently associated with early dramatic improvement. As shown in figure 3, the NIHSS reduction ratio was highly correlated with the post-treatment perfusion delay volume (r=0.544, p<0.001). Moreover, there was a significant correlation between time to >40% reduction in the NIHSS score and the initial DWI lesion volumes (r=0.549, p<0.001).
Although a lack of improvement at 24 h after thrombolytic therapy is associated with poor outcome and death at 3 months after stroke,9 a significant proportion of patients still have a chance to achieve a favourable outcome. Indeed, at least one-third of non-responders still achieved good outcomes at 3 months after successful early arterial revascularisation with intravenous tPA.1 This raises the possibility of stunned brain syndrome or delayed recovery with a lack of early clinical improvement.1 Our novel findings demonstrated that delayed recovery after lack of early dramatic improvement is relatively common (40%), which may be manifest throughout the follow-up period.
Several mechanisms may potentially account for stunned brain syndrome.1 Factors that have been reported to be associated with this phenomenon include the degree and speed of revascularisation,11–15 oedema formation,16 distal occlusion or proximal reocclusion,17 and reperfusion injury.18 Most studies that have been conducted to date have used transcranial Doppler and focused on the time and degree of revascularisation of antegrade flow after intravenous application of tPA.9 11 12 Christou et al reported that the timing of revascularisation was inversely correlated with early improvement in the NIHSS scores at 24 h, and suggested that there was a 5 h window to achieve early complete recovery.14 Conversely, Wunderlich et al reported that both full and partial revascularisation not only in the early course of up to 6 h after symptom onset but also between 6 h and 24 h after stroke were associated with early improvement and improved functional outcome 30 days after stroke.19 In the present study, the time interval from symptom onset to the start of revascularisation varied, and early dramatic improvement was observed in patients who received revascularisation therapy up to 8.1 h after the onset of symptoms. Additionally, no correlation between the time from the onset of symptoms to the start of revascularisation therapy and the speed and degree of recovery was observed. However, serial transcranial Doppler monitoring was not performed in the present study and the exact time of revascularisation could not be assessed.
Multimodal MRI may provide information regarding the influence of the ischaemic zone on the degree and speed of recovery after revascularisation. The intensity of hypoperfusion within an oligaemic field has been reported to be largely independent of the size of the oligaemic region.20 Thus we assessed the severity and extent of perfusion delay on pretreatment MRI and found that they were not correlated with the speed of recovery. Conversely, a larger volume of pretreatment DWI and post-treatment perfusion delay resulted in a lower chance of early dramatic improvement. Overall, the results of this study suggest that there are three different types of treatment response depending on serial MRI findings: (a) early dramatic improvement, if reperfusion occurs after revascularisation therapy and baseline smaller DWI lesions are present, (b) delayed recovery with a lack of early improvement (stunned brain syndrome), which is indicated by reperfusion and larger DWI lesions, and (c) non-responders, if reperfusion did not occur after revascularisation therapy. These findings suggest that recovery after reperfusion may be delayed in patients with a large established infarct, even though the initial clinical deficits in patients with delayed recovery were similar to those in patients with early dramatic improvement. Additionally, post-treatment reperfusion was associated with both early and delayed recovery, which suggests that, among patients who lacked early dramatic improvement after revascularisation therapy, delayed recovery may still occur when post-treatment reperfusion is observed on post-treatment PWI.
Our diffusion–perfusion MRI results suggest that the pathophysiological mechanism of stunned brain syndrome may not be explained by ischaemic penumbra, as defined by Tmax delay lesions. In the present study, the speed of recovery was not associated with the extent and severity of perfusion delay. Thus it is conceivable that factors other than the severity of ischaemic injury measured by delayed tissue perfusion may play an important role in recovery. For example, the capacity of functional restoration, such as neuroplasticity, which is related to the extent of infarct, may be related to the speed of recovery after stroke. Stroke induces axonal sprouting in the neighbouring, partially damaged, cortex within the first 3 weeks after the infarct.21 The degree and speed of this process may be associated with the pretreatment infarct volumes. The recovery process after reperfusion may be delayed in patients with extensive infarcts, even though the initial clinical deficits did not differ between the early and delayed recovery group. However, further studies using functional imaging techniques are needed to demonstrate the mechanisms of delayed recovery.
The strengths of this study include the prospective recruitment of patients from two centres and a serial MRI based study that included evaluation of the pre- and post-treatment perfusion status. However, the results of this study should be interpreted with some caution due to the limited sample size and various treatment modalities used. In addition, in this study, dramatic improvement was defined as an NIHSS score reduction of ≥40%, which has been reported to best predict revascularisation after therapy.10 Similar results were observed when we used the definition of early dramatic recovery as a decrease in NIHSS score of ≥4 points at 24 h after stroke onset (see supplementary table available online).1 9 Finally, in the present study, immediate post-treatment PWI was performed to evaluate the reperfusion status rather than measurement of the vascular recanalisation, and the time and degree of antegrade recanalisation was not considered in this study. There may be discordance between revascularisation and reperfusion after revascularisation. However, post-treatment perfusion status and tissue fate after stroke may depend on other factors in addition to revascularisation of antegrade flow, such as baseline perfusion status and collateral flow.22 Additionally, there was a good relationship between the volume of post-treatment perfusion delay and the degree of revascularisation and pretreatment collateral grade among patients who underwent conventional angiography (see supplementary figure available online).
Our data clearly show that a significant proportion of patients lacking early dramatic improvement (40%) exhibit delayed recovery, which indicate that there may be an opportunity for delayed recovery for patients who exhibit reperfusion after revascularisation therapy. In addition, both pretreatment infarct volume and post-treatment reperfusion were associated with the degree and speed of recovery. Serial diffusion–perfusion MRI may identify patients with ‘stunned brain syndrome’.
The UCLA-Samsung Stroke Collaborators: Oh Young Bang; Suk Jae Kim; Ji Won Kim; Jung Jae Lee; Gyeong-Moon Kim; Keon Ha Kim; Pyoung Jeon; Chin-Sang Chung; Kwang Ho Lee; Jeffry R Alger; Sidney Starkman; Bruce Ovbiagele; Doojin Kim; Latisha K Ali; Samir H Shah; Paul M Vespa; Reza Jahan; Noriko Salamon; Gary R Duckwiler; J Pablo Villablanca; Fernando Viñuela; David S Liebeskind; and Jeffrey L Saver.
Funding This study was supported by a grant from the Korean Healthcare Technology R&D Project, Ministry of Health and Welfare (A080044), the Samsung Medical Center Clinical Research Development Program grant CRS110-13-1 and IN-SUNG Foundation for Medical Research.
Ethics approval The local institutional review boards approved the study.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.