RESUMEN
AIM: To compare frequency and distribution of deep nuclei involvement in isolated basal ganglia and ventrolateral thalamus (BGT) versus combined BGT and watershed (BGT-WS) hypoxic-ischaemic injury (HII). MATERIALS AND METHODS: A retrospective review was undertaken of the magnetic resonance imaging (MRI) reports of children (0-18 years) with isolated BGT or combined BGT-WS HII. The location and extent of deep nuclear injuries were compared between groups using Fisher's exact test. RESULTS: Of 762 MRI reports, 435 (57%) had isolated BGT and 327 (43%) combined BGT-WS. Isolated BGT showed basal ganglia involvement in 85.1% (n=370) versus 49.8% (n=163) for combined BGT-WS (p<0.01). Sole putamen lesions were more common in isolated BGT (70.3%; 306) versus combined (19.3%; 63; p<0.01). Thalamic involvement was similar between isolated BGT (93.8%; 408) and combined BGT-WS (96.9%; 317; p>0.05). Sole ventrolateral nucleus involvement was more common in isolated BGT (66.6%; 291) while sole pulvinar lesions (25.1%; 82) and whole thalamus lesions (41.6%; 136) were more common in combined BGT-WS (p<0.01). Putamen and ventrolateral nucleus was the most frequent BGT lesion combination in isolated BGT (55.4%) but not in combined BGT-WS (8.6%; p<0.01). CONCLUSION: Variations in the frequency of deep nuclear lesions between groups may reflect different underlying pathogenetic mechanisms. Therefore, combined BGT-WS patterns may not necessarily indicate a superimposed profound on partial prolonged HII, as other causes such as neonatal hypoglycaemia may cause these.