Presented at the Neonatal Society 2007 Summer Meeting.

Thayyil S¹, Omar R², Cady E³, Robertson N⁴

¹ UCL Institute of Child Health, London, UK
² Statistical science department, UCL, London, UK
³ Medical Physics and Bioengineering, UCL, London, UK
⁴ UCL EGA Institute for Women’s Health, London, UK

Background: Although conventional magnetic resonance (MR) imaging performed between 5 and 14 days after birth provides prognostic information in neonatal encephalopathy (NE), earlier & more objective and quantitative tools are required. Whilst diffusion weighted imaging has been a useful prognostic tool in adult stroke, its utility in NE has been disappointing (1). Cerebral (¹H) magnetic resonance spectroscopy (MRS) peak area ratios are quantitative and objective measures of brain injury severity, however there has been no agreement as to which peak area metabolite at which time is most robust. As brain lactate increases and N-acetyl aspartate (a neuronal marker) decreases with injury, we hypothesised that thalamic Lac/NAA peak area ratio is a robust biomarker.

Objective: To perform a systematic review of the diagnostic accuracy of thalamic Lac/NAA acquired during the neonatal period for predicting adverse neurodevelopmental outcome.

Methods: We searched Medline and Embase libraries between 1980 to April 2007. We reviewed all prospective or retrospective studies on newborn infants (born at > 35 weeks) with NE comparing ¹H MRS metabolite peak area ratios with neurodevelopmental outcome at 1 year or more. We used QUADAS criteria to examine study quality. We sub grouped and re analysed the data according to post natal age (<4 days, 5-14 days, 15-30 days), metabolite ratio and voxel location and calculated diagnostic indices.

Results: We screened 726 papers and extracted individual patient or 2×2 table data from 23 that reported various peak area ratios (NAA/Creatine (Cr), NAA/Choline (Cho), Cho/Lac, NAA/Cr, Lac/Cr, Lac/NAA and inositol/Cr). Eight studies reported Lac/NAA ratios which showed good diagnostic utility at all postnatal ages (Chart 1). The sensitivity and specificity for Lac/NAA for predicting adversity at < 4 days, 5-14 days and 15-50 days from study 3 is shown in Chart 2.

Conclusion: Basal ganglia Lac/NAA peak area ratios within 1 month after birth in NE is a good biomarker of adversity.