ArticlesDWI-FLAIR mismatch for the identification of patients with acute ischaemic stroke within 4·5 h of symptom onset (PRE-FLAIR): a multicentre observational study
Introduction
International guidelines and approval criteria for use of intravenous thrombolysis exclude patients with stroke whose time of symptom onset is unknown.1, 2 However, many patients who have had a stroke present with unknown time of symptom onset (eg, if symptoms occur during sleep and are recognised only when the patient awakes) and, as with myocardial infarction, patients are more likely to have a stroke in the morning.3, 4 An estimated 25% of ischaemic strokes occur during sleep,4, 5, 6 which means that this large group of patients are precluded from thrombolysis—the only approved and effective specific treatment for acute ischaemic stroke. This dissatisfying situation has raised great interest in surrogate markers of lesion age. Moreover, clinical and imaging findings in patients waking with stroke symptoms were reported to be much the same as in patients with symptom onset known to be within 3–6 h,5, 6, 7 indicating that some of these patients might benefit from thrombolysis.
Multiparametric MRI has been suggested as a means to identify patients who are likely to benefit from thrombolysis, by use of different sequences that are sensitive to different aspects of tissue pathophysiology in acute cerebral ischaemia.6, 8, 9 Alterations of water diffusion can be detected with diffusion-weighted imaging (DWI) within 3 min from onset of ischaemia,10 whereas a net increase of water can be detected as an increase of T2 signal within 1–4 h from onset of ischaemia.11, 12 Fluid-attenuated inversion recovery (FLAIR) imaging is a T2-weighted imaging sequence13 and an integral part of common multiparametric stroke MRI protocols. FLAIR imaging is highly sensitive to subacute ischaemic brain lesions,14 but typically cannot be used to detect ischaemic lesions within the first few hours.15, 16
The mismatch of visibility of an acute ischaemic lesion between DWI and FLAIR has been suggested to enable identification of patients with acute ischaemic stroke who are likely to be within the 3-h time window. In a single centre study,9 a DWI-FLAIR mismatch identified patients within 3 h of symptom onset with high specificity (93%) and positive predictive value (PPV; 94%), although sensitivity and negative predictive value (NPV) were low. These findings have been reproduced in studies from other groups.17, 18, 19 All these studies showed a clear time dependency, with acute ischaemic lesion visibility on FLAIR imaging increasing during the first 6 h after stroke onset and approaching nearly 100% thereafter, a finding that is well known from experimental stroke.11 Taken together, the results from these studies are promising with regard to the use of DWI-FLAIR mismatch to estimate lesion age in patients with unknown time of stroke onset.
In some centres, multiparametric MRI incorporating the absence of a lesion on FLAIR imaging is already used to decide whether thrombolytic drugs should be given to patients who wake up with symptoms of stroke.8, 20, 21 However, neither safety nor efficacy of MRI-based thrombolysis in patients who awake with stroke have been shown. Thus, more data on the benefits of DWI-FLAIR mismatch in clinical practice are needed. Four reliable but retrospective single-centre studies lend support to such use of DWI-FLAIR mismatch,6, 17, 18, 19 but several questions remain unanswered—eg, reproducibility in a multicentre setting and the effect of potential confounding factors such as leukoaraiosis, lesion volume, and infarct location. Before DWI-FLAIR mismatch can be used to guide use of thrombolysis in patients with unknown time of symptom onset, its safety and efficacy should be assessed in a randomised controlled trial, and before such a trial, the diagnostic accuracy of this approach needs to be confirmed in a large multicentre study. In the PREdictive value of FLAIR and DWI for the identification of patients with acute ischaemic stroke within 4·5 h of symptom onset (PRE-FLAIR) study, we aimed to test whether DWI-FLAIR mismatch can be used to identify patients within 4·5 h of symptom onset with high specificity and PPV, and thus to assess the possibility of using DWI-FLAIR mismatch in a future randomised controlled trial of MRI-based thrombolysis.
Section snippets
Study design and patients
The PRE-FLAIR study was a multicentre observational study of patients with acute ischaemic stroke with known time of symptom onset who had undergone MRI with DWI and FLAIR within 12 h of symptom onset. We tested the predefined hypothesis that DWI-FLAIR mismatch can be used to identify patients within 4·5 h of symptom onset with high specificity in a retrospective cohort of patients studied within 12 h of symptom onset, presenting between Jan 1, 2001, and May 31, 2009. The study was done by an
Results
Figure 1 shows the study profile. Figure 2 shows examples of DWI and FLAIR images, including poor quality MRI scans for which some patients were excluded. At least one observer judged imaging to be of poor quality on DWI in 26 cases (4·0%), on FLAIR in 29 cases (5%), and on both DWI and FLAIR in 7 cases (1%). Table 1 shows baseline characteristics of the study sample.
Interobserver agreement for the detection of acute DWI was 93·9% with a κ of 0·506 (95% CI 0·361–0·651). Interobserver agreement
Discussion
Our assessment of a large multicentre dataset yielded three main findings. First, we showed that the DWI-FLAIR mismatch can be used to identify patients within 4·5 h of symptom onset with high specificity and high PPV. This finding substantiates those from some of the previous smaller single-centre studies,9, 17, 19 lending support to the use of DWI-FLAIR mismatch as a surrogate marker to identify patients with acute stroke who are eligible for intravenous thrombolysis (panel). Second, the
References (37)
- et al.
FLAIR can estimate the onset time in acute ischemic stroke patients
J Neurol Sci
(2010) - et al.
Effects of alteplase beyond 3 h after stroke in the Echoplanar Imaging Thrombolytic Evaluation Trial (EPITHET): a placebo-controlled randomised trial
Lancet Neurol
(2008) - et al.
Time to treatment with intravenous alteplase and outcome in stroke: an updated pooled analysis of ECASS, ATLANTIS, NINDS, and EPITHET trials
Lancet
(2010) Guidelines for management of ischaemic stroke and transient ischaemic attack 2008
Cerebrovasc Dis
(2008)- et al.
Guidelines for the early management of adults with ischemic stroke: a guideline from the American Heart Association/American Stroke Association Stroke Council, Clinical Cardiology Council, Cardiovascular Radiology and Intervention Council, and the Atherosclerotic Peripheral Vascular Disease and Quality of Care Outcomes in Research Interdisciplinary Working Groups—the American Academy of Neurology affirms the value of this guideline as an educational tool for neurologists
Stroke
(2007) Circadian variation in the timing of stroke onset: a meta-analysis
Stroke
(1998)- et al.
Circadian variation in acute ischemic stroke: a hospital-based study
Stroke
(1998) - et al.
The stroke patient who woke up: clinical and radiological features, including diffusion and perfusion MRI
Stroke
(2002) - et al.
Stroke on awakening: looking for a more rational management
Cerebrovasc Dis
(2003) - et al.
Early CT findings in unknown-onset and wake-up strokes
Cerebrovasc Dis
(2006)
Safety and efficacy of MRI-based thrombolysis in unclear-onset stroke: a preliminary report
Cerebrovasc Dis
Negative fluid-attenuated inversion recovery imaging identifies acute ischemic stroke at 3 hours or less
Ann Neurol
Diffusion-weighted MR imaging of acute stroke: correlation with T2-weighted and magnetic susceptibility-enhanced MR imaging in cats
AJNR Am J Neuroradiol
Changes of relaxation times (T1, T2) and apparent diffusion coefficient after permanent middle cerebral artery occlusion in the rat: temporal evolution, regional extent, and comparison with histology
Magn Reson Med
Absolute measurements of water content using magnetic resonance imaging: preliminary findings in an in vivo focal ischemic rat model
Magn Reson Med
Use of fluid attenuated inversion recovery (FLAIR) pulse sequences in MRI of the brain
J Comput Assist Tomogr
Fluid-attenuated inversion recovery (FLAIR) for assessment of cerebral infarction. Initial clinical experience in 50 patients
Stroke
Clinical utility of diffusion-weighted magnetic resonance imaging in the assessment of ischemic stroke
Ann Neurol
Cited by (435)
Deep and infratentorial cerebral microbleeds are related to wake-up stroke by large-artery atherosclerosis and cardioembolism
2024, Journal of the Neurological SciencesCharacteristics and outcomes of unknown onset stroke: The Japan Stroke Data Bank
2023, Journal of the Neurological SciencesCollaborative multi-modal deep learning and radiomic features for classification of strokes within 6 h
2023, Expert Systems with ApplicationsImaging predictors of clinical outcomes after endovascular treatment in MRI-selected patients with acute basilar artery occlusion
2023, Clinical Neurology and Neurosurgery