Abstract 54: Utility of Automated Perfusion-diffusion Magnetic Resonance Imaging in Acute Childhood Arterial Ischaemic Stroke

Abstract

Background and Objectives: Recent studies have used automated perfusion imaging software to identify adults most likely to benefit from reperfusion therapies in extended time windows. The time course of penumbral evolution in childhood arterial ischemic stroke (AIS) is poorly characterised. We investigated the utility of automated perfusion imaging software in childhood AIS and explored the relationship between time to imaging and perfusion-diffusion mismatch. Methods: Convenience population of children with acute AIS, presenting to the Royal Children’s Hospital Melbourne from 2005-2014, where diffusion and dynamic susceptibility contrast MRI were performed <48 hours of symptom onset. Perfusion-diffusion mismatch was estimated using RAPID (iSchemaView). Core was defined as ADC<620х10 -6 mm2/s and hypoperfusion as Tmax>6s. Favourable mismatch profile was defined as core <70mls, mismatch volume ≥15mls and ratio ≥1.8. Results: Twenty-nine children (median age 8, IQR 4.4-14.6) met eligibility criteria (26 unilateral MCA and 3 unilateral cerebellar infarcts). Median PedNIHSS was 4.5. Etiologies included focal cerebral arteriopathy (n=9), cryptogenic (n=12), cardioembolic (n=5), other (n=3). Median time from onset to imaging was 13.7 hours (IQR 7.5-25.3). The visible diffusion lesion was not below ADC<620 threshold in 19 (34% of cases) (median time to imaging 21 hours). Two children with subcortical/cortical lesions, imaged at 3.75 and 11 hours had favourable mismatch profile. Conclusions: RAPID failed to segment the ischaemic core in cases with delayed imaging. Favourable mismatch profiles persisted beyond the standard 4.5 hours window for thrombolysis. Further work is required to investigate the effect of time-delay and aetiology on mismatch characteristics in childhood AIS. Figure: 22-month child with acute left MCA occlusion. Diffusion (left) and Tmax perfusion (right) MRI at 3.75 hours post-onset demonstrating a favourable mismatch profile.