CT Perfusion to Predict Response to Recanalization in Ischemic Stroke (CRISP2)

Stroke is the number 1 cause of disability in the United States and a global public health problem. Globally, approximately 15 million people suffer a stroke each year, leading to 5 million deaths and another 5 million patients who suffer permanent disability from their stroke. Strokes that are caused by blockage of large blood vessels supplying blood to the brain are typically the most disabling.  Over the last 5 years, there have been major breakthroughs in acute stroke therapy for this type of stroke.  In 2015, multiple trials demonstrated a profound benefit from endovascular stroke therapy for patients with a large vessel ischemic stroke who present in the early time-window (within 6 hours after symptom onset).  This success was followed by the DEFUSE 3 and DAWN trials showing a very strong benefit from endovascular therapy in the delayed time-window (6-24 hours after symptom onset).  These therapies have now become standard of care at specialty hospitals (comprehensive stroke centers) across the country. Despite this new highly effective therapy, stroke-related disability continues to be substantial for patients with large-vessel occlusions, because of brain damage that occurs in the time-period before patients receive endovascular therapy that restores blood flow to the brain. The time period before blood flow is restored can be long (several hours) especially if patients need to be transferred from a community hospital that does not have the capability to provide endovascular therapy to a comprehensive stroke center that does. Because of this delay, significant brain damage can occur between the time that a patient initially presents to a hospital and the time that blood flow is restored. To address this problem, we need to conduct clinical trials of therapies that protect the brain in the crucial time-period before blood flow is restored. In order to conduct such trials, we first need tools to identify patients who are most likely to benefit from treatments that protect the brain and radiological tools that can be used to determine if the treatments are effective. The overall aim of this project is to develop these tools. We will conduct a two-part study.  The first part involves collection and analysis of historical data on stroke patients who have previously been transferred for endovascular therapy from referring hospitals.  Using the neuroimaging and clinical data abstracted from medical records of patients included in this historical cohort study, we will improve our methods to segment the acute stroke on CT imaging.  We will also develop models that predict infarct growth during transfer.  The second part of the study involves prospective data collection on patients who are transferred for endovascular therapy from referring hospitals.  Using the neuroimaging and clinical data abstracted from medical records of patients included in this prospective cohort study, we will validate the infarct growth prediction models that were developed in part 1.