Advanced imaging improves prediction of hemorrhage after stroke thrombolysis

OBJECTIVE

Very low cerebral blood volume (VLCBV), diffusion, and hypoperfusion lesion volumes have been proposed as predictors of hemorrhagic transformation following stroke thrombolysis. We aimed to compare these parameters, validate VLCBV in an independent cohort using DEFUSE study data, and investigate the interaction of VLCBV with regional reperfusion.

METHODS

The EPITHET and DEFUSE studies obtained diffusion and perfusion magnetic resonance imaging (MRI) in patients 3 to 6 hours from onset of ischemic stroke. EPITHET randomized patients to tissue plasminogen activator (tPA) or placebo, and all DEFUSE patients received tPA. VLCBV was defined as cerebral blood volume<2.5th percentile of brain contralateral to the infarct. Parenchymal hematoma (PH) was defined using European Cooperative Acute Stroke Study criteria. Reperfusion was assessed using subacute perfusion MRI coregistered to baseline imaging.

RESULTS

In DEFUSE, 69 patients were analyzed, including 9 who developed PH. The >2 ml VLCBV threshold defined in EPITHET predicted PH with 100% sensitivity, 72% specificity, 35% positive predictive value, and 100% negative predictive value. Pooling EPITHET and DEFUSE (163 patients, including 23 with PH), regression models using VLCBV (p<0.001) and tPA (p=0.02) predicted PH independent of clinical factors better than models using diffusion or time to maximum>8 seconds lesion volumes. Excluding VLCBV in regions without reperfusion improved specificity from 61 to 78% in the pooled analysis.

INTERPRETATION

VLCBV predicts PH after stroke thrombolysis and appears to be a more powerful predictor than baseline diffusion or hypoperfusion lesion volumes. Reperfusion of regions of VLCBV is strongly associated with post-thrombolysis PH. VLCBV may be clinically useful to identify patients at significant risk of hemorrhage following reperfusion.

Abstract WMP53: Vitamin D Levels Are Lower In Acute Ischemic Stroke Patients Compared To Matched Controls

Introduction: Vitamin D deficiency is increasingly recognized as a global pandemic. In longitudinal healthy population studies, vitamin D deficiency is associated with increased incidence of ischemic stroke. We compared serum 25-OH vitamin D levels between Asian ischemic stroke patients within 1 week of stroke onset with age and gender matched healthy individuals.

Methods: We prospectively recruited 133 consecutive Asian acute ischemic stroke patients admitted to the Singapore General Hospital. Blood samples were collected within seven days of stroke onset. Stroke patients were matched for age and gender to individuals with no history of stroke from a database of healthy Asian controls. Serum 25-OH vitamin D was measured using Roche competitive electrochemiluminescence immunoassay for both stroke patients and healthy controls in the same laboratory. Vitamin D deficiency was defined as serum 25-OH vitamin D <20 μg/L and insufficiency as 20-40 μg/L. Matched statistical analyses were performed using Wilcoxon and McNemar tests.

Results: Among the 133 matched pairs of ischemic stroke patients and healthy controls, median age was 60 years and 75% are males. Median serum 25-OH vitamin D level was lower among stroke patients (23.2 IQR 16.65-29.60 μg/L) compared to healthy controls (28.9 IQR 21.75-35.50 μg/L) (p<0.0001). Stroke patients had a higher prevalence of vitamin D deficiency (39%) compared to healthy controls (20%) (p<0.0001). A high proportion of stroke patients had insufficient vitamin D compared to healthy controls (95% vs 84%, p=0.007). Corrected serum calcium was lower in stroke patients (median 2.26 mmol/L IQR 2.21-2.35) than controls (median 2.34 mmol/L IQR 2.28- 2.41) (p<0.0001). There was no difference in serum phosphate (p=0.266) and parathyroid hormone levels (p=0.807).

Conclusion: This cross-sectional study of 25-OH vitamin D levels in Asian patients within 1 week of ischemic stroke onset contributes to the growing evidence that vitamin D insufficiency/ deficiency is a stroke risk factor. Vitamin D measurement within 1 week of stroke onset likely reflects pre-stroke levels and is not influenced by post-stroke effects. Our findings provide impetus for future studies to investigate if vitamin D supplementation reduces stroke incidence.

Abstract TP47: Infarct Growth Does Not Predict Functional Outcome For Small Vessel Stroke

Introduction: Infarct growth has been shown to be a predictor of clinical outcome following ischemic stroke and has been used as a surrogate in reperfusion studies. There are no data whether the association of infarct growth and clinical outcome depends on stroke etiology. Furthermore, there is uncertainty on how to handle small volume infarcts with some studies excluding them from analyses. We studied the relationship between infarct growth and functional outcome in stroke patients with small volume infarcts for subgroups with small and non-small vessel etiologies.

Methods: We studied 37 patients in the MRI substudy of the Efficacy of Nitric Oxide in Stroke (ENOS) trial with baseline infarct volumes of ≤5ml. None of the patients were treated with reperfusion strategies. Brain MRI was performed serially at baseline within 48 hours of onset, on days 7 and 90. Infarct growth was measured as the volume difference between final T2 and baseline DWI lesions. Good functional outcome was defined as day 90 modified Rankin score ≤2 and poor as >2.

Results: Among the 26 patients with underlying small vessel etiology, there was no difference between those with good and poor outcomes in terms of absolute [median 0.0 (IQR -0.4 to 1.1) vs 0.1 (-0.7 to 0.2) mL, p=0.802] and relative infarct growth [median -3 (-27 to 81) vs 8 (-46 to 10) %, p=0.802]. However, in the 11 patients with etiologies other than small vessel disease, patients with good outcome had less absolute [median -0.9 (-1.6 to 0.7) vs 3.2 (0.2 to 9.1) mL, p=0.033] and relative infarct growth [median -59 (-81 to 9) vs 154 (35 to 292) %, p=0.019] compared to those with poor outcome.

Discussion: In small volume strokes, infarct growth was not associated with functional outcome for small vessel stroke although there was a significant association for other stroke etiologies. This novel finding in this small sample should be confirmed in larger studies. Our findings are expected as small vessel stroke involves occlusion of penetrating arterioles which supply small, limited yet strategically important brain regions. Thus, while infarct growth may be a suitable surrogate for clinical outcomes for small volume infarcts due to non-small vessel disease, but it may not be for small vessel strokes.

The Effects of Alteplase 3 to 6 Hours After Stroke in the EPITHET–DEFUSE Combined Dataset

Background and Purpose—

Two phase 2 studies of alteplase in acute ischemic stroke 3 to 6 hours after onset, Echoplanar Imaging Thrombolytic Evaluation Trial (EPITHET; a randomized, controlled, double-blinded trial), and Diffusion and Perfusion Imaging Evaluation for Understanding Stroke Evolution Study (DEFUSE; open-label, treatment only) using MR imaging-based outcomes have been conducted. We have pooled individual patient data from these to assess the response to alteplase. The primary hypothesis was that alteplase would significantly attenuate infarct growth compared with placebo in mismatch-selected patients using coregistration techniques.

Methods—

The EPITHET–DEFUSE study datasets were pooled while retaining the original inclusion and exclusion criteria. Significant hypoperfusion was defined as a Tmax delay >6 seconds), and coregistration techniques were used to define MR diffusion-weighted imaging/perfusion-weighted imaging mismatch. Neuroimaging, parameters including reperfusion, recanalization, symptomatic intracerebral hemorrhage, and clinical outcomes were assessed. Alteplase and placebo groups were compared for the primary outcome of infarct growth as well for secondary outcome measures.

Results—

From 165 patients with adequate MR scans in the EPITHET–DEFUSE pooled data, 121 patients (73.3%) were found to have mismatch. For the primary outcome analysis, 60 patients received alteplase and 41 placebo. Mismatch patients receiving alteplase had significantly attenuated infarct growth compared with placebo ( P =0.025). The reperfusion rate was also increased (62.7% vs 31.7%; P =0.003). Mortality and clinical outcomes were not different between groups.

Conclusions—

The data provide further evidence that alteplase significantly attenuates infarct growth and increases reperfusion compared with placebo in the 3- to 6- hour time window in patients selected based on MR penumbral imaging.

Clinical Trial Registration—

URL: http://www.clinicaltrials.gov . Unique identifier: NCT00238537

Multiparametric MRI and CT models of infarct core and favorable penumbral imaging patterns in acute ischemic stroke

BACKGROUND AND PURPOSE

Objective imaging methods to identify optimal candidates for late recanalization therapies are needed. The study goals were (1) to develop magnetic resonance imaging (MRI) and computed tomography (CT) multiparametric, voxel-based predictive models of infarct core and penumbra in acute ischemic stroke patients, and (2) to develop patient-level imaging criteria for favorable penumbral pattern based on good clinical outcome in response to successful recanalization.

METHODS

An analysis of imaging and clinical data was performed on 2 cohorts of patients (one screened with CT, the other with MRI) who underwent successful treatment for large vessel, anterior circulation stroke. Subjects were divided 2:1 into derivation and validation cohorts. Pretreatment imaging parameters independently predicting final tissue infarct and final clinical outcome were identified.

RESULTS

The MRI and CT models were developed and validated from 34 and 32 patients, using 943 320 and 1 236 917 voxels, respectively. The derivation MRI and 2-branch CT models had an overall accuracy of 74% and 80%, respectively, and were independently validated with an accuracy of 71% and 79%, respectively. The imaging criteria of (1) predicted infarct core ≤90 mL and (2) ratio of predicted infarct tissue within the at-risk region ≤70% identified patients as having a favorable penumbral pattern with 78% to 100% accuracy.

CONCLUSIONS

Multiparametric voxel-based MRI and CT models were developed to predict the extent of infarct core and overall penumbral pattern status in patients with acute ischemic stroke who may be candidates for late recanalization therapies. These models provide an alternative approach to mismatch in predicting ultimate tissue fate.

Differing Associations of White Matter Lesions and Lacunar Infarction with Retinal Microvascular Signs

Background

White matter lesions (WML) and lacunar infarcts (LI) are believed to have microvascular etiologies but the exact microvascular changes occurring in each is unclear. Aim Using the retina as a proxy, we assessed retinal microvascular changes in WML and LI.

Methods

We prospectively recruited 1211 acute stroke patients. Four subgroups were identified from neuroimaging: WML alone, LI alone, both WML and LI, neither WML nor LI. Masked retinal photographs identified retinopathy and retinal arteriolar wall signs and measured retinal vascular caliber.

Results

Compared with 448 controls with neither WML nor LI, 384 patients with only WML were more likely to have retinopathy [odds ratio (OR) 1·5, 95% confidence interval (CI) 1·1 to 2·1] and enhanced arteriolar light reflex (OR 1·6, 95% CI 1·1 to 2·3); 200 patients with only LI were more likely to have arteriolar narrowing (OR 1·6, 95% CI 1·1 to 2·3) and enhanced arteriolar light reflex (OR 1·6, 95% CI 1·0 to 2·4); and 179 patients with both WML and LI were more likely to have arteriovenous nicking (OR 1·7, 95% CI 1·1 to 2·6), enhanced arteriolar light reflex (OR 2·0, 95% CI 1·3 to 3·2) and wider venules (OR 2·3, 95% CI 1·4 to 3·6). All analyses were adjusted for age, gender, study site and cardiovascular risk factors.

Conclusion

Both WML and LI were associated with retinal microvascular signs, supporting a microvascular etiology. Differing patterns of association suggest different mechanisms may predominate, e.g. greater endothelial permeability in WML, and ischemia associated with arteriolar wall disease in LI.

Abstract 61: Reperfusion Beyond 4.5 Hours Reduces Infarct Growth And Improves Clinical Outcome

Background and purpose: The ECASS 3 study demonstrated efficacy of intravenous thrombolysis up to 4.5h after stroke onset. It has been hypothesized that some patients have tissue at risk and an acceptably low hemorrhage risk beyond 4.5h. Imaging based selection may help identify these patients for late reperfusion therapies. No randomized data have shown efficacy of tPA or reperfusion later than 4.5h after onset. We analysed the Echoplanar Imaging Thrombolytic Evaluation Trial (EPITHET) data to assess the effect of treatment and reperfusion on attenuation of infarct growth in the 4.5 to 6 hour time window.

Methods: Patients were randomized to placebo or tissue plasminogen activator (tPA) between 4.5-6h from stroke onset (without using imaging selection criteria). Pre-treatment DWI and day 90 T2-weighted lesion volumes (average of manually outlined lesions by 2 independent raters) were compared to assess the influence of tPA and reperfusion on absolute and relative infarct growth. Day 3 volume was used when day 90 data was missing. The effect of tPA on reperfusion was also assessed. Good clinical outcome was defined as a National Institute of Health Stroke Scale (NIHSS) at day 90 0-1 or improvement ≥ 8 from baseline. Good functional outcome was defined as modified Rankin Scale (mRS) 0-2.

Results: Of 69 patients treated 4.5-6hrs hours after stroke onset, infarct growth could be assessed in 63. The median relative growth was significantly lower in the tPA group compared to placebo (0.94 vs 1.68, p=0.025). There was a nonsignificant trend towards lower absolute growth (-0.17mL vs 9.56mL, p=0.069). Reperfusion markedly reduced relative (0.80 vs 1.89, p<0.001) and absolute infarct growth (-2.49mL vs 39.50mL, p<0.001). Reperfusion was more likely in the tPA group (57.7 vs 25.0% p=0.026) and was associated with better clinical and functional outcomes (86.4% vs 28.1% p<0.001 and 72.7 vs 34.4% p=0.012).

Conclusion: Thrombolysis after 4.5 hours reduced infarct growth and increased the rate of reperfusion. The strong positive effect of reperfusion on clinical and functional outcomes in this later time window is evidence of persisting salvageable ischemic penumbra. This supports continuing efforts to extend the treatment window for reperfusion therapies.

Abstract 92: MRI Patient Selection In Acute Stroke Trials: Implications For Sample Size

Background: MRI selection of patients with acute stroke may reduce the required sample size for randomized controlled trials of interventions aimed at restoring blood flow. The Target Mismatch pattern has been proposed as a MRI-marker that can differentiate patients who are likely to benefit from reperfusion from those who will have no effect from or may be harmed by reperfusion. The Target Mismatch pattern is present in approximately 50% of stroke patients with a middle cerebral artery occlusion who present in the 3-6 hour time-window. The pattern is based on the following PWI and DWI criteria: ratio of Tmax>6s over DWI volume >1.2, difference between Tmax>6s and DWI volume >10ml, and DWI volume <100ml. The aim of this study was to compare the required sample size of acute stroke studies that use MRI selection to that of studies which do not use MRI selection.

Methods: All sample size calculations were based on an alpha of 0.05 in a two-sided test and a desired power of 0.8. MRI patient selection was assumed to be according to Target Mismatch criteria. The primary study outcome was assumed to be good functional outcome, defined as a modified Rankin Scale of 0-2 at 90 days. Other assumptions for the sample size calculations were based on data from the literature and data from the pooled EPITHET-DEFUSE database regarding patients with middle cerebral artery (MCA) occlusions treated up to six hours after symptom onset. The spontaneous reperfusion rate in this population was assumed to be 0.22. For MRI-selected mismatch patients the rate of good functional outcome was assumed to be 0.73 with reperfusion and 0.30 without reperfusion. For unselected patients, the rate of favorable clinical outcome was assumed to be 0.67 with reperfusion and 0.41 without reperfusion.

Results: A trial of an intervention that leads to reperfusion of the MCA in 45% of patients would require 396 patients per group if MRI selection criteria were applied and 1096 patients per group without MRI selection criteria. A trial of an intervention that has a 70% reperfusion rate would require 92 patients in each arm if MRI selection criteria were applied versus 251 patients per arm if patients were enrolled without MRI selection. (see figure )

Conclusion: MRI selection can markedly reduce the required sample size of randomized controlled stroke trials in the delayed time-window. This advantage needs to be balanced against the potential drawbacks of using a MRI-marker as an inclusion criterion for a randomized controlled trial.

Abstract 95: Regional Very Low Cerebral Blood Volume with Subsequent Local Reperfusion Predicts Hemorrhagic Transformation in Acute Ischemic Stroke

Background and Purpose Regions of very low cerebral blood volume (VLCBV) on MR perfusion imaging have been shown to predict hemorrhagic transformation (HT) following stroke thrombolysis. We tested the hypothesis that local reperfusion in a region of VLCBV is a pre-requisite for hemorrhagic transformation using pooled imaging data from the EPITHET and DEFUSE studies.

Methods Standard CBV maps were calculated and smoothed (Gaussian) to reduce noise. The volume of VLCBV was calculated within the acute Tmax>4sec perfusion lesion using fully automated techniques and a range of VLCBV thresholds relative to CBV values in the non-stroke hemisphere. Receiver operating characteristic (ROC) analysis was used to determine the optimal definition and threshold of VLCBV to predict parenchymal hematoma (PH, ECASS definition). Regional reperfusion was assessed using co-registered subacute Tmax perfusion images (DEFUSE 3-6hrs post thrombolysis, EPITHET 3-5 days post thrombolysis/placebo). The risk of PH associated with VLCBV was assessed with and without exclusion of regions of VLCBV within persistently hypoperfused regions.

Results Of 145 patients with baseline perfusion imaging, 22 (15.2%) had PH (13 PH1, 9 PH2). A VLCBV definition of either <2.5 th percentile of the contralateral CBV distribution (VLCBV<2.5pctile) or <15% of the mean contralateral CBV (VLCBV<15%) had similar performance in predicting PH (AUC 0.73 for both). To achieve sensitivity of 95% required a VLCBV<2.5pctile threshold of >2mL (specificity 47%) or a VLCBV<15% threshold of >0.5mL (specificity 41%). There were 130 patients with subacute perfusion imaging, at which time 15 (11.5%) had developed PH. A further 3 patients (without reperfusion at subacute MRI) later developed PH and were excluded as reperfusion status at the time of PH was unknown. In the remaining 127 patients, the AUC for PH increased from 0.77 to 0.92 (p<0.001, VLCBV<2.5pctile definition) when regions of VLCBV without reperfusion on subacute imaging were excluded. The specificity of the >2mL threshold (VLCBV<2.5pctile) increased from 46 to 75%, positive predictive value increased from 20 to 35%, likelihood ratio for PH increased from 1.9 to 4.0 (sensitivity and negative predictive value were both 100% in these 127 patients). No patient developed PH at the time of subacute imaging in the absence of local reperfusion, including one patient where reperfusion of basal ganglia infarction had occurred (with CBV normalisation) prior to thrombolysis.

Conclusions Local reperfusion is a critical factor in determining the risk of HT associated with regional VLCBV. This is consistent with the hypothesis that the severe ischemia represented by VLCBV is associated with focal blood-brain-barrier disruption and potential HT should reperfusion subsequently occur. Assessment of VLCBV can be automated and may be useful in clinical risk-benefit decisions regarding thrombolysis.

Greater effect of stroke thrombolysis in the presence of arterial obstruction

OBJECTIVE

Recanalization of arterial obstruction is associated with improved clinical outcomes. There are no controlled data demonstrating whether arterial obstruction status predicts the treatment effect of intravenous (IV) tissue plasminogen activator (tPA). We aimed to determine if the presence of arterial obstruction improves the treatment effect of IV tPA over placebo in attenuating infarct growth.

METHODS

We analyzed 175 ischemic stroke patients treated in the 3-6 hour time window from the Echoplanar Imaging Thrombolytic Evaluation Trial (EPITHET) trial (randomized to IV tPA or placebo) and Diffusion and perfusion imaging Evaluation For Understanding Stroke Evolution (DEFUSE) study (all treated with IV tPA). Infarct growth was calculated as the difference between baseline diffusion-weighted imaging (DWI) and final T2 lesion volumes. Baseline arterial obstruction of large intracranial arteries was graded on magnetic resonance angiography (MRA).

RESULTS

Among the 116 patients with adequate baseline MRA and final lesion assessment, 72 had arterial obstruction (48 tPA, 24 placebo) and 44 no arterial obstruction (33 tPA, 11 placebo). Infarct growth was lower in the tPA than placebo group (median difference 26ml, 95% confidence interval [CI], 1-50) in patients with arterial obstruction, but was similar in patients with no arterial obstruction (median difference 5ml, 95%CI, -3 to 9). Infarct growth attenuation with tPA over placebo treatment was greater among patients with arterial obstruction than those without arterial obstruction by a median of 32ml (95%CI, 21-43, p < 0.001).

INTERPRETATION

The treatment effect of IV tPA over placebo was greater with baseline arterial obstruction, supporting arterial obstruction status as a consideration in selecting patients more likely to benefit from IV thrombolysis.

The infarct core is well represented by the acute diffusion lesion: sustained reversal is infrequent

Diffusion-weighted imaging (DWI) is commonly used to assess irreversibly infarcted tissue but its accuracy is challenged by reports of diffusion lesion reversal (DLR). We investigated the frequency and implications for mismatch classification of DLR using imaging from the EPITHET (Echoplanar Imaging Thrombolytic Evaluation Trial) and DEFUSE (Diffusion and Perfusion Imaging Evaluation for Understanding Stroke Evolution) studies. In 119 patients (83 treated with IV tissue plasminogen activator), follow-up images were coregistered to acute diffusion images and the lesions manually outlined to their maximal visual extent in diffusion space. Diffusion lesion reversal was defined as voxels of acute diffusion lesion that corresponded to normal brain at follow-up (i.e., final infarct, leukoaraiosis, and cerebrospinal fluid (CSF) voxels were excluded from consideration). The appearance of DLR was visually checked for artifacts, the volume calculated, and the impact of adjusting baseline diffusion lesion volume for DLR volume on perfusion-diffusion mismatch analyzed. Median DLR volume reduced from 4.4 to 1.5 mL after excluding CSF/leukoaraiosis. Visual inspection verified 8/119 (6.7%) with true DLR, median volume 2.33 mL. Subtracting DLR from acute diffusion volume altered perfusion-diffusion mismatch (T(max)>6 seconds, ratio>1.2) in 3/119 (2.5%) patients. Diffusion lesion reversal between baseline and 3 to 6 hours DWI was also uncommon (7/65, 11%) and often transient. Clinically relevant DLR is uncommon and rarely alters perfusion-diffusion mismatch. The acute diffusion lesion is generally a reliable signature of the infarct core.

Fluid-attenuated inversion recovery hyperintensity in acute ischemic stroke may not predict hemorrhagic transformation

BACKGROUND

Fluid-attenuated inversion recovery (FLAIR) hyperintensity within an acute cerebral infarct may reflect delayed onset time and increased risk of hemorrhage after thrombolysis. Given the important implications for clinical practice, we examined the prevalence of FLAIR hyperintensity in patients 3-6 h from stroke onset and its relationship to parenchymal hematoma (PH).

METHODS

Baseline DWI and FLAIR imaging with subsequent hemorrhage detection (ECASS criteria) were prospectively obtained in patients 3-6 h after stroke onset from the pooled EPITHET and DEFUSE trials. FLAIR hyperintensity within the region of the acute DWI lesion was rated qualitatively (dichotomized as visually obvious or subtle (i.e. only visible after careful windowing)) and quantitatively (using relative signal intensity (RSI)). The association of FLAIR hyperintensity with hemorrhage was then tested alongside established predictors (very low cerebral blood volume (VLCBV) and diffusion (DWI) lesion volume) in logistic regression analysis.

RESULTS

There were 49 patients with pre-treatment FLAIR imaging (38 received tissue plasminogen activator (tPA), 5 developed PH). FLAIR hyperintensity within the region of acute DWI lesion occurred in 48/49 (98%) patients, was obvious in 18/49 (37%) and subtle in 30/49 (61%). Inter-rater agreement was 92% (κ = 0.82). The prevalence of obvious FLAIR hyperintensity did not differ between studies obtained in the 3-4.5 h and 4.5-6 h time periods (40% vs. 33%, p = 0.77). PH was poorly predicted by obvious FLAIR hyperintensity (sensitivity 40%, specificity 64%, positive predictive value 11%). In univariate logistic regression, VLCBV (p = 0.02) and DWI lesion volume (p = 0.03) predicted PH but FLAIR lesion volume (p = 0.87) and RSI (p = 0.11) did not. In ordinal logistic regression for hemorrhage grade adjusted for age and baseline stroke severity (NIHSS), increased VLCBV (p = 0.002) and DWI lesion volume (p = 0.003) were associated with hemorrhage but FLAIR lesion volume (p = 0.66) and RSI (p = 0.35) were not.

CONCLUSIONS

Visible FLAIR hyperintensity is almost universal 3-6 h after stroke onset and did not predict subsequent hemorrhage in this dataset. Our findings question the value of excluding patients with FLAIR hyperintensity from reperfusion therapies. Larger studies are required to clarify what implications FLAIR-positive lesions have for patient selection.

RAPID automated patient selection for reperfusion therapy: a pooled analysis of the Echoplanar Imaging Thrombolytic Evaluation Trial (EPITHET) and the Diffusion and Perfusion Imaging Evaluation for Understanding Stroke Evolution (DEFUSE) Study

BACKGROUND AND PURPOSE

The aim of this study was to determine if automated MRI analysis software (RAPID) can be used to identify patients with stroke in whom reperfusion is associated with an increased chance of good outcome.

METHODS

Baseline diffusion- and perfusion-weighted MRI scans from the Diffusion and Perfusion Imaging Evaluation for Understanding Stroke Evolution study (DEFUSE; n=74) and the Echoplanar Imaging Thrombolytic Evaluation Trial (EPITHET; n=100) were reprocessed with RAPID. Based on RAPID-generated diffusion-weighted imaging and perfusion-weighted imaging lesion volumes, patients were categorized according to 3 prespecified MRI profiles that were hypothesized to predict benefit (Target Mismatch), harm (Malignant), and no effect (No Mismatch) from reperfusion. Favorable clinical response was defined as a National Institutes of Health Stroke Scale score of 0 to 1 or a ≥ 8-point improvement on the National Institutes of Health Stroke Scale score at Day 90.

RESULTS

In Target Mismatch patients, reperfusion was strongly associated with a favorable clinical response (OR, 5.6; 95% CI, 2.1 to 15.3) and attenuation of infarct growth (10 ± 23 mL with reperfusion versus 40 ± 44 mL without reperfusion; P<0.001). In Malignant profile patients, reperfusion was not associated with a favorable clinical response (OR, 0.74; 95% CI, 0.1 to 5.8) or attenuation of infarct growth (85 ± 74 mL with reperfusion versus 95 ± 79 mL without reperfusion; P=0.7). Reperfusion was also not associated with a favorable clinical response (OR, 1.05; 95% CI, 0.1 to 9.4) or attenuation of lesion growth (10 ± 15 mL with reperfusion versus 17 ± 30 mL without reperfusion; P=0.9) in No Mismatch patients.

CONCLUSIONS

MRI profiles that are associated with a differential response to reperfusion can be identified with RAPID. This supports the use of automated image analysis software such as RAPID for patient selection in acute stroke trials.

Refining the definition of the malignant profile: insights from the DEFUSE-EPITHET pooled data set

BACKGROUND AND PURPOSE

To refine the definition of the malignant magnetic resonance imaging profile in acute stroke patients using baseline diffusion-weighted magnetic resonance imaging (DWI) and perfusion-weighted magnetic resonance imaging (PWI) findings from the pooled DEFUSE/EPITHET database.

METHODS

Patients presenting with acute stroke within 3 to 6 hours from symptom onset were treated with tissue plasminogen activator or placebo. Baseline and follow-up DWI and PWI images from both studies were reprocessed using the same software program. A receiver operating characteristic curve analysis was used to identify Tmax and DWI volumes that optimally predicted poor outcomes (modified Rankin Scale 5-6) at 90 days in patients who achieved reperfusion.

RESULTS

Sixty-five patients achieved reperfusion and 46 did not reperfuse. Receiver operating characteristic analysis identified a PWI (Tmax>8 s) volume of >85 mL as the optimal definition of the malignant profile. Eighty-nine percent of malignant profile patients had poor outcome with reperfusion versus 39% of patients without reperfusion (P=0.02). Parenchymal hematomas occurred more frequently in malignant profile patients who experienced reperfusion versus no reperfusion (67% versus 11%, P<0.01). DWI analysis identified a volume of 80 mL as the best DWI threshold, but this definition was less sensitive than were PWI-based definitions.

CONCLUSIONS

Stroke patients likely to suffer parenchymal hemorrhages and poor outcomes following reperfusion can be identified from baseline magnetic resonance imaging findings. The current analysis demonstrates that a PWI threshold (Tmax>8 s) of approximately 100 mL is appropriate for identifying these patients. Exclusion of malignant profile patients from reperfusion therapies may substantially improve the efficacy and safety of reperfusion therapies. Clinical Trial Registration Information- URL: http://www.clinicaltrials.gov. Unique identifier: NCT00238537.

Associations of retinal microvascular signs and intracranial large artery disease

BACKGROUND AND PURPOSE

Intracranial large artery disease (ICLAD) is a major cause of ischemic stroke. Retinal microvascular changes are associated with stroke, including small vessel cerebral disease and extracranial carotid disease. We examined the relationship between ICLAD and retinal microvascular changes.

METHODS

This is a prospective cohort of 802 acute ischemic stroke patients. Retinal changes were assessed from photographs by graders masked to clinical data. ICLAD was evaluated using prespecified criteria.

RESULTS

ICLAD was not associated with ipsilateral retinal arteriolar/venular caliber, focal arteriolar narrowing, or arteriovenous nicking. Severe enhanced arteriolar light reflex was independently associated with any ICLAD (P=0.006) and severe ICLAD (P<0.001).

CONCLUSIONS

Enhanced arteriolar light reflex, but not retinal vessel caliber, was related to ICLAD. These data suggest that retinal microvascular signs have specific associations with large cerebral vessel disease.

EPITHET: Positive Result After Reanalysis Using Baseline Diffusion-Weighted Imaging/Perfusion-Weighted Imaging Co-Registration

BACKGROUND AND PURPOSE

the Echoplanar Imaging Thrombolytic Evaluation Trial (EPITHET) was a prospective, randomized, double-blinded, placebo-controlled, phase II trial of alteplase between 3 and 6 hours after stroke onset. The primary outcome of infarct growth attenuation on MRI with alteplase in mismatch patients was negative when mismatch volumes were assessed volumetrically, without coregistration, which underestimates mismatch volumes. We hypothesized that assessing the extent of mismatch by coregistration of perfusion and diffusion MRI maps may more accurately allow the effects of alteplase vs placebo to be evaluated.

METHODS

patients were classified as having mismatch if perfusion-weighted imaging divided by coregistered diffusion-weighted imaging volume ratio was >1.2 and total coregistered mismatch volume was ≥ 10 mL. The primary outcome was a comparison of infarct growth in alteplase vs placebo patients with coregistered mismatch.

RESULTS

of 99 patients with baseline diffusion-weighted imaging and perfusion-weighted imaging, coregistration of both images was possible in 95 patients. Coregistered mismatch was present in 93% (88/95) compared to 85% (81/95) with standard volumetric mismatch. In the coregistered mismatch patients, of whom 45 received alteplase and 43 received placebo, the primary outcome measure of geometric mean infarct growth was significantly attenuated by a ratio of 0.58 with alteplase compared to placebo (1.02 vs 1.77; 95% CI, 0.33-0.99; P=0.0459).

CONCLUSIONS

when using coregistration techniques to determine the presence of mismatch at study entry, alteplase significantly attenuated infarct growth. This highlights the necessity for a randomized, placebo-controlled, phase III clinical trial of alteplase using penumbral selection beyond 3 hours.

The benefits of intravenous thrombolysis relate to the site of baseline arterial occlusion in the Echoplanar Imaging Thrombolytic Evaluation Trial (EPITHET)

BACKGROUND AND PURPOSE

In ischemic stroke, the site of arterial obstruction has been shown to influence recanalization and clinical outcomes. However, this has not been studied in randomized controlled trials, nor has the impact of arterial obstruction site on reperfusion and infarct growth been assessed. We studied the influence of site and degree of arterial obstruction patients enrolled in the Echoplanar Imaging Thrombolytic Evaluation Trial (EPITHET).

METHODS

EPITHET was a prospective, randomized, placebo-controlled trial of intravenous tissue plasminogen activator (tPA) in the 3- to 6-hour time window. Arterial obstruction site and degree were rated on magnetic resonance angiography blinded to treatment allocation and outcomes.

RESULTS

In 101 EPITHET patients, 87 had adequate quality magnetic resonance angiography, of whom 54 had baseline arterial obstruction. Infarct growth attenuation was greater in those with tPA treatment compared to placebo among patients with middle cerebral artery (MCA) obstruction (P=0.037). The treatment benefit of tPA over placebo in attenuating infarct growth was greater for MCA than internal carotid artery (ICA) obstruction (P=0.060). With tPA treatment, good clinical outcome was more likely with MCA than with ICA obstruction (P=0.005). Most patients with ICA obstruction did not achieve good clinical outcome, whether treated with tPA (100%) or placebo (77%). The study was underpowered to prove any treatment benefit of tPA among patients with any or severe degree of arterial obstruction.

CONCLUSIONS

Arterial obstruction site strongly predicts outcomes. ICA obstruction carries a uniformly poor prognosis, whereas good outcomes with MCA obstruction are associated with tPA therapy.

Regional very low cerebral blood volume predicts hemorrhagic transformation better than diffusion-weighted imaging volume and thresholded apparent diffusion coefficient in acute ischemic stroke

BACKGROUND AND PURPOSE

Currently, diffusion-weighted imaging (DWI) lesion volume is the most useful magnetic resonance imaging predictor of hemorrhagic transformation (HT). Preliminary studies have suggested that very low cerebral blood volume (VLCBV) predicts HT. We compared HT prediction by VLCBV and DWI using data from the EPITHET study.

METHODS

Normal-percentile CBV values were calculated from the nonstroke hemisphere. Whole-brain masks with CBV thresholds of the <0, 2.5, 5, and 10th percentiles were created. The volume of tissue with VLCBV was calculated within the acute DWI ischemic lesion. HT was graded as per ECASS criteria.

RESULTS

HT occurred in 44 of 91 patients. Parenchymal hematoma (PH) occurred in 13 (4 symptomatic) and asymptomatic hemorrhagic infarction (HI) in 31. The median volume of VLCBV was significantly higher in cases with PH. VLCBV predicted HT better than DWI lesion volume and thresholded apparent diffusion coefficient lesion volume in receiver operating characteristic analysis and logistic regression. A cutpoint at 2 mL VLCBV with the <2.5th percentile had 100% sensitivity for PH and, in patients treated with tissue plasminogen activator, defined a population with a 43% risk of PH (95% CI, 23% to 66%, likelihood ratio=16). VLCBV remained an independent predictor of PH in multivariate analysis with traditional clinical risk factors for HT.

CONCLUSIONS

VLCBV predicted HT after thrombolysis better than did DWI or apparent diffusion coefficient volume in this large patient cohort. The advantage was greatest in patients with smaller DWI volumes. Prediction was better in patients who recanalized. If validated in an independent cohort, the addition of VLCBV to prethrombolysis decision making may reduce the incidence of HT.