Multimodal computed tomography in patient selection for acute ischemic stroke intervention

Background: Multimodal computed tomographic (CT) imaging for acute ischemic stroke patients being considered for therapy has been the focus of an increasing body of research over the past decade. The impact of this research has been an increasing use of multimodal CT for individual patient management in acute ischemic stroke by, for example, by excluding hemorrhages from therapy, assisting the treatment decision making process for neurologists or providing useful information to prediction patient outcomes and inform patient family discussions. However, numerous important research questions remain unanswered in this field. The body of work presented in this thesis attempts to answer a number of clinical questions in the application of multimodal CT imaging in acute stroke. Computed tomography (CT), including non-contrast CT (NCCT), CT perfusion (CTP), and CT angiography (CTA) is being increasingly applied in acute stroke imaging because its wide availability, safety and ability to be performed rapidly as an imaging technique when compared to conventional magnetic resonance imaging (MRI). Furthermore, the endovascular trials published in 2015 have shown that multimodal CT selection of patients with M1-segment of the middle cerebral artery, or internal carotid artery occlusions who undergo thrombectomy therapy have significantly better clinical outcomes compared to the best medical care intravenous thrombolysis. Additionally, the recent positive endovascular stroke trials extending the treatment time window from 6 to 24 hours required to use of multimodal CT including CTP imaging to identify patients with reperfusion-responsive lesions. As a result, the utility of CTP imaging and the angiographic modalities, including static CTA, multiphase CTA, and dynamic CTA have become more important for the evaluation of the cerebral vasculature. This thesis aimed to 1) assess the ‘treatment effect’ of alteplase in patients without a baseline vessel occlusion (Chapter IV); 2a) test the vessel occlusion location frequency and clinical and perfusion imaging profiles of patients presenting with differing occlusion locations; 2b) understand if occlusion location was an independent predictor of outcome; 2c) test the relationship between occlusion location and baseline ischemic core, a known predictor of modified Rankin Scale (mRS) at 90-day; 2d) assess the influences of excluding patients with a proximal large vessel occlusion from thrombolysis trials on the power of the study (Chapter V); 3a) test the inter-rater reliability of collateral scores assessed on dynamic CTA and optimized multiphase CTA; 3b) test the relationships of collaterals assessed on dynamic CTA and multiphase CTA with CTP tissue compartments (Chapter VI). Methods: Acute ischemic stroke patients included in this thesis were sourced from the INternational Stroke Perfusion Imaging REgistry (INSPIRE). All included patients underwent multimodal CT imaging before treatment. The baseline and follow-up CTP images were post-processed using MIStar (Apollo Medical Imaging Technology, Melbourne, Australia); vessel occlusion status and the location of vessel occlusions was assessed on static CTA; collateral status was assessed using Miteff collateral scores on static CTA, or ASPECTS collaterals (Alberta Stroke Program Early CT Score collateral system, a 6-point grading system, with Score 0 for poor collaterals and Score 5 for normal collaterals) and ASITN/SIR collateral (American Society of Interventional and Therapeutic Neuroradiology/Society of Interventional Radiology, a 5-point grading system, with Score 0 for poor collaterals and Score 4 for normal collaterals) systems on optimized multiphase CTA and dynamic CTA images derived from CTP source images using MIStar. Patient baseline and 24-hour stroke severity were assessed using National Institutes of Health Stroke Scale (NIHSS), and patient functional outcomes at 90-day were assessed using modified Rankin Score (mRS). Results: 1) alteplase treated patients without a vessel occlusion were less likely to have an excellent outcome (mRS 0-1; 56%) than untreated (78.8%, OR, 0.42, 95% confidence interval, 0.24-0.75, P=0.003); 2a) the favorable patient outcomes at 90-day were higher in patients with a distal vessel occlusion (M2-, M3-segment of the middle cerebral artery, anterior cerebral artery and posterior cerebral artery) compared to patients with an M1 occlusion (mRS 0-2, 73% versus 50%, P<0.001); 2b) Vessel occlusion location was an independent predictor of outcome; 2c) the location was not a strong predictor of outcomes compared to baseline ischemic core (area under the curve, mRS 0-1, 0.64 versus 0.83; mRS 0-2 0.70 versus 0.86, P<0.001); 2d) the power for detecting any ‘treatment effect’ in patients with a distal vessel occlusion in was increased when using surrogate outcomes such as imaging biomarkers or early clinical outcomes (35-51% for mRS 0-2 compared to 72% for penumbral salvage); 3a) The collateral assessment using ASPECTS collaterals on dynamic CTA has a similar inter-rater agreement (K-alpha: 0.71) compared to the collaterals assessed using optimized multiphase CTA (K-alpha: 0.69); 3b) the agreement between dynamic CTA and CTP in classifying patients with target mismatch was higher compared to optimized multiphase CTA (Kappa, dynamic CTA: 0.81; optimized multiphase CTA: 0.64). Conclusion: The studies in this thesis support the application of advanced stroke imaging in both research and clinical care. However, the results indicate the need for careful and through examination of CT perfusion tissue compartments, vessel occlusion status and collateral status when advanced CT is applied for decision assistance or for clinical trials eligibility. The results suggest specifically that CTP can 1) aid in the assessment of the probability of the patient gaining benefit from alteplase treatment; 2) inform new trial designs for intravenous thrombolysis and provide important information for neuro-interventionalists who are using thrombolysis as an accompanying therapy during mechanical thrombectomy; 3) increase the power of the study when using imaging biomarker as an outcome measure, an important element in trial design; 4) provide collateral vessel assessments on time-resolved CTAs, and post-processed CT perfusion (a reliable approach compared to user-dependent collateral assessments on other CTA modalities) as adjunctive information to the standard “target mismatch” approach; 5) finally, CTP approaches can be fully automated in comparison to the more standard CTA techniques that are user and rater dependent. Overall, these results indicate that CT perfusion is a modality that has a number of advantages over standard CT angiographic techniques.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.