Establishing the Diagnosis of Acute Ischemic Stroke
Ruling Out Intracranial Hemorrhage
The initial step in the evaluation of patients with symptoms of acute stroke is to differentiate between hemorrhagic and ischemic stroke. Since intracranial hemorrhage is an absolute contraindication for reperfusion therapies, most stroke protocols begin with noncontrast head CT (NCHCT). NCHCT has been widely accepted as the standard method for the detection of acute intracranial hemorrhage since early reports describing its accuracy with early-generation CT scanners. However, its sensitivity and specificity in detecting intracranial hemorrhage have not been formally studied by comparing to actual pathological/histological specimens.
Although CT is the standard method, gradient T2*-weighted MRI sequences (including gradient-recalled echo [GRE] and susceptibility-weighted imaging [SWI] sequences) are equally—if not more—sensitive for the detection of acute intracranial hemorrhage. The accuracy of MR imaging techniques in the detection of intracranial hemorrhage in acute stroke setting (within 6 hours) was reported as likely equivalent to NCHCT. Furthermore, T2*-weighted sequences have superior accuracy in the detection of small hemosiderin deposits from chronic microhemorrhages, which are often undetected on NCHCT due to insufficient signal contrast and limited spatial resolution. However, the clinical significance of microhemorrhages is uncertain and remains an area of intense investigation. While a meta-analysis of patients who received IV thrombolysis with a small number of chronic hemorrhages (fewer than 5) concluded that there is no significantly increased risk of hemorrhage, the outcome when numerous microhemorrhages (5 or more) are present has not been studied.
Diagnosing Ischemic Stroke
Early ischemic changes in NCHCT include loss of gray-white distinction, indistinct insular cortex and obscured basal ganglia, and hyperattenuated clot in the proximal vessels (Fig. 1). The hyperdense vessel sign is most specific but has low sensitivity. Abnormalities on NCHCT are present in 40%–50% of acute ischemic strokes. When obvious signs are present, NCHCT allows rapid diagnosis and correlation with presenting symptoms. However, acute ischemic changes are often subtle, with intra- and interobserver variability. Multiple classification systems based on imaging features have been developed to offer reliable, reproducible grading of the extent of ischemic changes on NCHCT. One of them is the Alberta Stroke Program Early CT Score (ASPECTS), a 10-point scoring system (Fig. 2; Table 1). Studies have shown that baseline ASPECTS correlates inversely with severity as assessed by the NIHSS within the first 3 hours of middle cerebral artery (MCA) stroke onset. ASPECTS of 7 or lower has been shown to predict poor functional outcome (78% sensitivity and 96% specificity) and symptomatic hemorrhage (90% sensitivity and 62% specificity) (Fig. 3).
(Enlarge Image)
Figure 1.
Early ischemic changes on NCHCT. Axial NCHCT images demonstrating an indistinct right insular cortex (arrow,A) and obscured right basal ganglia, with loss of gray-white distinction (open arrow, B) and hyperdense right MCA (arrow, B and C).
(Enlarge Image)
Figure 2.
Admission NCHCT of patient with ASPECTS 4, calculated at the level of the basal ganglia and at the supraganglionic level. In this scoring system, the MCA territory is divided into 10 regions: the caudate (C), lentiform (L), internal capsule (IC), and 7 cortical regions (insular cortex [I], M1, M2, M3, M4, M5, and M6), based on data from Barber et al.
(Enlarge Image)
Figure 3.
Type PH2 hemorrhagic transformation. Axial NCHCT images obtained 14 hours after initial presentation in a patient with a left MCA infarct, demonstrating hemorrhagic transformation with confluent hematoma (arrows) and surrounding edema causing mass effect (left-to-right midline shift, subfalcine herniation, right lateral ventricle entrapment).
Diffusion-weighted imaging (DWI) is approximately 4 to 5 times more sensitive in detecting acute stroke than NCHCT(Fig. 4). Its sensitivity in detecting ischemia is reported as 99% with a high specificity of 92%. Within minutes of vessel occlusion, failure of the sodium potassium pump leads to an influx of water from the extracellular space to the intracellular space. Cytotoxic edema restricts diffusion of water molecules and appears as increased signal intensity on DWI. About 95% of hyperacute infarcts are positive on DWI. DWI can detect acute brain infarction within 1 to 2 hours, while NCHCT may be negative for the first 24 to 36 hours. DWI can distinguish acute from chronic ischemia, thereby delineating new lesions even when located in proximity to prior ischemic injury (Table 2). Obscure lesions indiscernible on CT scans, such as lacunar infarcts, particularly those located in the posterior fossa, are better visualized on DWI.
(Enlarge Image)
Figure 4.
Superior sensitivity of DWI in the detection of acute infarcts. A: Axial NCHCT image obtained at admission showing subtle obscuration of the right insular cortex (arrow). B: Axial DW image demonstrating hyperintense lesions in the right MCA territory—right lateral frontal cortex and right basal ganglia—consistent with acute infarction (arrow). The central hypodensities likely represent petechial hemorrhage. C: Axial CTA maximum intensity projection confirms occlusion of the M1 segment of the right MCA (arrow).
Despite strong evidence supporting DWI as superior to NCHCT for confirming diagnosis of acute stroke within the first 24 hours, logistical issues limits its use in emergent setting. Most institutions find it challenging to obtain emergent MRI without delaying treatment.
Of note, restricted diffusion is not exclusively observed in acute ischemic stroke, but can also be seen in some nonischemic entities. These include seizure, encephalitis, abscesses, metabolic derangement (hypoglycemia), Creutzfeldt-Jakob disease, lymphoma, and mucinous adenocarcinoma metastases. Under the guidance of clinical presentation, these entities can be readily differentiated when studies are reviewed in combination with studies using other imaging modalities such as FLAIR. Occasionally, clinical differentiation of seizures from acute stroke may be difficult. Discerning DWI lesions confined to a major vascular territory, a distinctive feature of acute stroke lesions, may be helpful.
Determining Eligibility for IV tPA
There is strong evidence supporting the use of IV tPA as a recanalization therapy to improve clinical outcomes when patients present within the standard 0- to 3-hour time window and the extended 3- to 4.5-hour time window. There is also strong evidence supporting the timely use of imaging to exclude hemorrhage in stroke patients before initiating IV thrombolytic therapy. Admission NCHCT is recommended prior to thrombolysis, because intracranial hemorrhage is an absolute contraindication to IV thrombolysis. Ischemia involving more than one-third of the MCA territory on images obtained within the 0- to 6-hour window constitutes a relative contraindication to IV thrombolysis. Of note, there is disagreement regarding the significance of extensive ischemic signs on admission NCHCT. Early on, the European Cooperative Acute Stroke Studies (ECASS) reported poor outcome with increased incidence of hemorrhage following thrombolysis in patients with early extensive infarctions. However, the National Institute of Neurological Disorders and Stroke tPA Stroke Trial contested that early extensive infarctions were associated with symptom severity but not with adverse outcome after thrombolysis. Results from recent studies recommend withholding IV thrombolysis when more than one-third of the MCA territory is involved.