Prediction of early childhood outcome after neonatal encephalopathy using magnetic resonance imaging, proton spectroscopy, and clinical neurological assessment: a prospective observational study

SummaryBackground Intrapartum-related neonatal encephalopathy (NE) is a leading cause of child mortality and disability. Magnetic resonance imaging, proton spectroscopy (MRI/MRS), and neurological assessments can predict outcomes, however sub-Saharan African data are lacking. We examined feasibility, findings, and prognostic accuracy of MRI/MRS and clinical assessments amongst NE infants in Uganda. Methods Neonates (≥36 weeks) with NE underwent MRI/MRS when clinically stable. MRI findings (Rutherford, NICHD criteria) and MRS metabolite ratios (lactate, N-acetyl aspartate, creatine, choline) were reported. General Movements Assessment (GMA) and Hammersmith Infant Neurological Examination (HINE) were conducted at 3 and 6 months, respectively. Prediction of adverse outcome (18–24 month death/neurodevelopmental impairment (NDI)) were analysed using receiver operating characteristic curves. Findings Amongst 51 neonates, 13 died before imaging and 6 were not imaged due to COVID-19 restrictions; 27/32 underwent MRI including 24 MRS (median 11 days, all diagnostic quality). Outcome was known for 21 infants, with 29% adverse (1 death, 5 NDI). MRI abnormalities occurred in 93%; 15% with moderate-severe BGT and 15% severe white matter injury, most (85%) consistent with hypoxia-ischaemia. Sensitivity/specificity for outcome were: MRI Rutherford (sensitivity 83%, specificity 93%); MRI NICHD (100%, 73%); MRS (83%, 92%); GMA (100%, 92%); and HINE (83%, 100%). Interpretation MRI, MRS, and neurological assessments were feasible in this Ugandan research setting and showed good prognostic accuracy for early childhood outcome after NE. Most had evidence of hypoxia-ischaemia on MRI, however injury patterns and MRS cutpoints differed from high-income country cohorts, possibly due to high mortality prior to imaging in this setting without access to intensive care. Funding The Gates Foundation grant OPP1210890.