Neuroimaging in childhood arterial ischaemic stroke: evaluation of aetiologies and imaging modalities

• Published in: Neuropediatrics
• Main Authors: Buerki, S, Roellin, K, Remonda, L, Gubser-Mercati, D, Jeannet, PY, Keller, E, Lütschg, J, Menache, C, Ramelli, GP, Schmitt-Mechelke, T, Weissert, M, Boltshauser, E, Steinlin, M
• Format: Conference Proceeding
• Language: English
• Published: 18.06.2008
• ISSN: 0174-304X; 1439-1899
• DOI: 10.1055/s-2008-1079486

Introduction: Considering the differences of paediatric to adult stroke neuroimaging pattern in different aetiologies of childhood stroke is of interest. We describe neuroimaging after arterial ischaemic stroke (AIS) and put them in relation to the classification by Ganesan (Dev Med Child Neurol 2005, 47:252–256). Methods: From 95 children registered prospectively by the Swiss Neuropaediatric Stroke Registry from 2000–2006 neuroimaging was reevaluated in 80 patients by a neuroradiologist and a neuropaediatrician (blinded to the clinical findings). The images were analysed for localisation of stroke, involved vessels, volume of stroke and findings of MR angiography (MRA). Available acute imaging (AI) (<14 days after stroke) was compared to follow up imaging (FI) (>14 days after stroke). Results: From 80 children (49 male) mean age at stroke was 6.7 years (0.1–16.2 years) AI was CT alone in 16 children, CT and MR in 25 children, MR in 35 and MRA in 47 children. From 65 cases with infarction in medial cerebral artery (MCA) territory, lesion was cortical/subcortical (cs) in 27, with involvement of basal ganglia (BG) in 29, BG alone in 9 patients. Infarction within posterior cerebral artery (PCA) territory showed cs infarction in 7, cs and thalamus infarction in 9 and isolated thalamic lesion in only 2 children. Anterior and posterior circulation was involved in 8 patients. Aetiology according classification of Ganesan was: no sickle cell, 10 cardioembolic, 5 moyamoya, 2 arterial dissection, 25 steno-occlusive arteriopathy, 16 other determined aetiology, 8 with multiple and 4 with undetermined aetiology, 4/8 undetermined due to lack of information (no MRA). CT was normal in AI in 5 children. Diffusion weighted imaging (DWI) in AI (n=44) was abnormal in all, but accompanying T2 weighted images (T2w) were normal in 4 cases. DWI was always accurate to predict size of lesion on FI (n=18). Acute MRA was abnormal in 37/47 patients. In 22 follow up MRA (4 days to 7 months) 12 patients had persistent findings, 3 partial and 5 patients complete revascularisation. Conclusion: Involvement of unilateral anterior circulation alone was present in 71%, posterior circulation in 19% and bilateral infarction anterior and/or posterior in 15%. Involvement of cs structures without BG or thalamus respectively was about equal to involvement with BG or thalamus, but isolated BG/thalamus involvement was markedly less frequent. Stroke pattern was not related to aetiology. DWI was best in early detection of lesion, CT and T2w MR can be normal within the first day. Acute MRA showed vasculopathy in 79%, normalisation could be detected within 4 days. Thus MR imaging including DWI and MRA are best for investigation of children with AIS.