The Benefit of Thrombectomy in Patients With Low ASPECTS Is a Matter of Shades of Gray—What Current Trials May Have Missed

Randomized trials supporting the benefit of endovascular treatment in acute ischemic stroke patients with a large early infarction are not yet available. Few retrospective studies exist that suggest a potential positive treatment effect on functional outcome, as well as procedural safety. However, potential benefit or harm of MT in patients with low initial ASPECTS is still a subject of current debate, and in particular, how to select these patients for treatment. The purpose of this pilot study was to evaluate how early tissue water uptake in acute ischemic brain might determine lesion fate and functional outcome in low ASPECTS patients undergoing MT. We observed that the degree of early water uptake measured by quantitative NWU was significantly associated with functional outcome in low ASPECTS patients, yielding a higher diagnostic power compared to other parameters such as ASPECTS, age, or NIHSS. No conclusive evidence of a beneficial effect of successful reperfusion was observed in patients with low ASPECTS and high NWU, which highlights the potential of NWU as a tool to specify patient selection.

Abstract WP83: Validation of a Relative Non-Contrast CT Map to Detect Early Ischemic Changes in Acute Stroke

Background: Discerning signs of early infarct on the non-contrast CT (NCCT) can be difficult. To facilitate interpretation of the NCCT we previously developed a technique to generate symmetry ratio maps of the NCCT (rNCCT maps) on which subtle (≥5%) differences in density between symmetric brain regions are enhanced. We sought to validate the rNCCT map against other measures of early infarction in a large cohort. Methods: rNCCT maps were generated for 146 ischemic stroke patients. We assessed how often a neurologist’s interpretation of the NCCT was changed when provided with the rNCCT map. The neurologist was blinded to CTP and DWI but was given the infarct hemisphere. In addition, using the 24-hour DWI as the gold standard, we assessed the sensitivity, specificity and volumetric accuracy of the rNCCT-defined infarct core and compared this to the test characteristics of CTP-defined infarct core (CBF<38% threshold). Results: Addition of rNCCT overlay map changed clinician’s initial read 64.4% of the time (95% CI 56-72%); the rNCCT identified new areas of ischemia not appreciated on blinded review 86.2% of the time (95% CI 78-92%) and in 35.1% helped rule out early ischemia where the reader was unsure of its presence (95% CI 26-45%). In the 53 patients with reperfusion and follow-up MRI, specificity of rNCCT for final lesion volume was 99.5% for rNCCT [98.5-99.8%] vs. 99.8% [IQR 98.8-99.9%] for CTP (P=0.08). Sensitivity for rNCCT was 19.9% [7.1-28.1%] vs. 17.5% [4.7-32.2%] for CTP (P=0.56). Conclusions: This study validates the rNCCT map for detection of early ischemic changes. It is more quantitative and objective than a clinician’s read of the NCCT alone. The sensitivity and specificity for detecting early ischemic changes on rNCCT were comparable to those achieved with CTP. This indicates that the rNCCT could be a valuable tool in the evaluation of acute stroke patients.

Abstract WP37: Evidence of Late Hemorrhage in Ischemic Stroke Patients Treated With Intravenous Alteplase: A Post Hoc Analysis of the Multicenter MR WITNESS Trial (NCT01282242)

Introduction: Although early hemorrhagic transformation (HT) in acute ischemic stroke has been studied, less is known about patients who develop hemorrhage after the acute phase. We hypothesized that patients with late hemorrhage (LH) would have more severe strokes than those without, and tested this hypothesis in a cohort of thrombolysed patients from the MR WITNESS trial. Methods: Subjects were recruited from 10 sites between Jan 2011-Oct 2015. MR WITNESS enrolled subjects if they were last seen between 4.5 and 24 hours prior to evaluation, but otherwise qualified for IV tPA in the 3-4.5 hr window per AHA guidelines, and if their brain MRI findings indicated very early infarction: either no FLAIR hyperintensity or subtle hyperintensity, ie signal increase <15% compared to the contralateral hemisphere. Patients with early (≤48 hours) HT (defined per ECASS criteria) were excluded from analysis. Late hemorrhage was defined as imaging manifestation of hemorrhage on 30 days MRI in patient without manifestation of HT at 48h. Good outcome was pre-specified as modified Rankin Scale (mRS) 0-1 at 90 d. Univariate comparisons utilized Fisher’s exact test and Wilcoxon Rank Sums 2-sample exact test for categorical and continuous variables, respectively. Results: Among the 80 patients included in the MR WITNESS cohort, 53 met our inclusion criteria and were analysed. When compared to those with no HT, patients with LH had larger baseline infarct volumes and perfusion defects, as well as more frequent proximal vessel occlusion at baseline (all p <0.01, See Table). Patients with LH also demonstrated worse functional outcome at 90 days (mRS, Median [IQR], 3 [1.75-4] vs 1 [0-2], p=0.006, Table). Conclusion: Patients with LH demonstrate a more severe imaging profile at baseline and a worse functional outcome at 90 days when compared to patients without hemorrhage. Understanding the underlying pathophysiology of LH may shed light on to the mechanisms of acute and subacute brain injury.

Early vs Delayed Cerebral Infarction After Aneurysm Repair After Subarachnoid Hemorrhage

BACKGROUND: Cerebral infarction is a major contributor to poor outcome after subarachnoid hemorrhage (SAH). Although usually considered a complication of delayed cerebral ischemia, infarcts may also occur early, in relation to initial brain injury or aneurysm-securing procedures. OBJECTIVE: We analyzed the relative frequency and volume of early vs delayed infarcts after SAH and their relationship to hospital outcome. METHODS: Retrospective review of consecutive patients admitted with aneurysmal SAH over 4 years who had follow-up brain imaging 7 days or later after admission. Imaging 24 to 48-hours after aneurysm-securing procedures was reviewed to classify infarcts seen on final imaging as early or delayed. Infarct volumes were measured by perimeter tracing and infarct burden calculated for each patient. RESULTS: Of 250 eligible patients, 205 had follow-up imaging; infarcts were present in 61 patients. Of these, 29 had early infarcts, 16 had delayed infarcts, and 5 had both early and delayed infarcts. Eleven patients with infarcts did not undergo postprocedure computed tomography; these were presumptively classified as having late infarcts. Early and delayed infarcts contributed equally to infarct burden. Early infarcts were associated with aneurysm clipping (odds ratio: 4.2, 95% confidence interval: 1.8-9.5 compared with coiling), whereas delayed infarcts were almost always seen in association with angiographic vasospasm (odds ratio: 3.3, 95% confidence interval: 1.5-7.3). Patients with early as well as late infarcts, especially those with infarct burden more than 30 cm3 had worse hospital discharge disposition. CONCLUSION: Early infarction occurs frequently after SAH and contributes as much as delayed cerebral ischemia to infarct burden and hospital outcome. Efforts to better understand and modify contributors to early infarction appear warranted.