Imaging-based treatment selection for intravenous and intra-arterial stroke therapies: a comprehensive review

Abstract

Reperfusion therapy is the only approved treatment for acute ischemic stroke. The current approach to patient selection is primarily based on the time from stroke symptom onset. However, this algorithm sharply restricts the eligible patient population, and neglects large variations in collateral circulation that ultimately determine the therapeutic time window in individual patients. Time alone is unlikely to remain the dominant parameter. Alternative approaches to patient selection involve advanced neuroimaging methods including MRI diffusion-weighted imaging, magnetic resonance and computed tomography perfusion imaging and noninvasive angiography that provide potentially valuable information regarding the state of the brain parenchyma and the neurovasculature. These techniques have now been used extensively, and there is emerging evidence on how specific imaging data may result in improved clinical outcomes. This article will review the major studies that have investigated the role of imaging in patient selection for both intravenous and intra-arterial therapies.

Figure 1. Examples of patients who have large infarcts despite early presentation, and patients with negligible infarcts at later time points

(A–C) A 66-year-old female with left middle cerebral artery M1 occlusion (arrow) demonstrated on (A) axial maximum intensity projection (MIP) and small infarct (14-ml volume) on (B) diffusion-weighted image (DWI) and (C) apparent diffusion coefficient (ADC) map. MRI was performed 8 h, 58 min post-ictus. The 3-month Modified Rankin Scale (mRS) score was 2. (D–F) A 52-year-old male with left middle cerebral artery M1 occlusion (arrow) demonstrated on (D) axial MIP and extensive infarct (146 ml volume) on (E) DWI and (F) ADC map. MRI was performed 4 h, 16 min post-ictus. The 3-month mRS score was 6. Reprinted with permission from Massachusetts General Hospital (MA, USA).

Figure 2. Poor sensitivity of noncontrast computed tomography in acute ischemia

(A) Noncontrast computed tomography (NCCT) in a 45-year-old male demonstrates subtle hypodensity in the putamen and possible effacement of sulci in the frontal operculum. (B) Diffusion-weighted image (DWI) and (C) apparent diffusion coefficient (ADC) map at the same level show more extensive acute infarct involving the basal ganglia, insula and cortical regions of the frontal and temporal lobe. (D) At a different level in the same patient, NCCT demonstrates subtle hypodensity in the deep white matter. (E) DWI and (F) ADC map reveal a larger region of acute infarction in the gray and white matter of the frontal and parietal lobes that is difficult to identify in the CT scan. Time between computed tomography and MRI was 45 min. Reprinted with permission from Massachusetts General Hospital (MA, USA).

Figure 3. Computed tomography angiography thick slab maximum intensity projection to identify arterial occlusions

Left middle cerebral artery (MCA) M2 occlusion visualized on (A) axial and (B) coronal maximum intensity projection images in a 72-year-old female with admission NIH Stroke Scale of 9 who was imaged approximately 7 h after stroke onset. The 3-months modified Rankin score was 1. Identification of proximal artery occlusions, particularly in the second-order branches such as the M2 (MCA) branch shown here, is facilitated by thick slab (20–30 mm) maximum intensity projection images, which can be constructed immediately at the CT scanner console and along three orthogonal planes. Reprinted with permission from Massachusetts General Hospital (MA, USA).

Figure 4. Massachusetts General Hospital imaging algorithm for deciding which patients to treat with endovascular stroke therapy

CT: Computed tomography; CTA: Computed tomography angiography; DWI: Diffusion-weighted imaging; IA: Intra-arterial; MR: Magnetic resonance; NCCT: Noncontrast computed tomography. Reprinted with permission from Massachusetts General Hospital (MA, USA).