Magnetic Resonance Biomarkers in Neonatal Encephalopathy

Overview

N-acetylaspartate (NAA) is a surrogate neuronal marker and its proton magnetic resonance spectroscopy (1H MRS) signal decreases with increasing neuronal mortality associated with cerebral hypoxia-ischaemia. The MRS lactate (Lac)/NAA peak-area ratio increases during and after severe cerebral hypoxia-ischaemia reflecting mitochondrial injury and impaired oxidative phosphorylation. Aims: (1) To establish normative ranges for thalamic 1H MRS NAA concentration and Lac/NAA in healthy newborn infants (2) To examine the accuracies of thalamic 1H MRS NAA concentration and Lac/NAA for predicting adverse neurodevelopmental outcome in neonatal encephalopathy (NE) Design: Prospective observational study Methods: Year 1: Following 1H MRS methodology optimisation 40 healthy control infants will be recruited to collect normative data. Year 2 to 3: 115 infants with NE, undergoing therapeutic hypothermia will be recruited. MRS will be performed aged less than 4 days and 7 to 14 days and thalamic NAA levels and Lac/NAA will be quantified; Qualitative interviews to evaluate parental understanding of this biomarker. Year 4, 5: Outcome assessment by BSID III at 18 months. Outcomes: Mean thalamic NAA levels and Lac/NAA and appropriate confidence intervals in normal infants, and thalamic NAA levels and Lac/NAA in infants with NE according to neurodevelopmental outcome. Areas under curves for thalamic NAA and Lac/NAA will be examined separately for early \& late MRS. Accuracy of early MRS will inform utility of this investigation in decisions about withdrawal of life support; late MRS will inform about efficacy as a surrogate end point in clinical trials. Qualitative interviews will be thematically analysed and reported.

2\. 1 PRIMARY OBJECTIVE To examine the accuracy of 1H MRS thalamic \[NAA\] for predicting adverse neurodevelopment in infants with NE undergoing therapeutic hypothermia 2.2 SECONDARY OBJECTIVE 2.2.1 To examine inter and intra-center variability of thalamic \[NAA\] levels in healthy adult volunteers 2.2.2 To further develop normative ranges for 1H MRS thalamic \[NAA\] and other quantitative measures of potential utility in healthy term newborn infants 2.2.3 To compare the prognostic accuracy of 1H MRS thalamic \[NAA\] and other MRS quantities with that of the Lactate/NAA peak-area ratio acquired early (aged ≤ 4 days) and with that acquired late (aged 5 to 14 days) 2.2.4 To examine parental attitudes towards MR use in decisions about withdrawal of life support and counselling about long term outcomes 3\. METHODS Prospective observational study; the index test (1H MRS) and gold-standard test (neurological outcome at 18 months) will be performed independently and blinded to each other. 3.3 Recruitment Control infants will be recruited from the postnatal wards at the participating hospitals. The MR scans will be performed before the infant is first discharged from hospital as far as possible (for example following a Caesarean delivery) i.e. within the first week of life. A lead research midwife will be identified in each of the participating hospitals who will approach the parents for recruitment and informed consent. NE infants will be recruited from neonatal intensive care units. MR scans will be performed twice after informed parental consent. The first scan at less than 4 days of age (if clinically feasible) and the second scan between 5 to 14 days of age. The first scan will be performed while the infant is still in the intensive care unit at the participating centre. At the time of the second MR scan, it is likely most babies would be off the ventilator. Sedation would be used if required in these babies. 3.4 Neurodevelopmental evaluation Neurodevelopmental outcomes will be assessed at 18 months of age using Bayley Scale of Infant Development® (Version 3) blinded to the MRS information. Favourable outcome will be defined as survival without disability, no disability being defined as all of: developmental quotient (DQ) \>84 (BSID III), no neuromotor impairment (neurological exam score ≥73 and Gross Motor Function Classification System (GMFCS) ≤I) \[19\], and normal vision and hearing on clinical examination. Unfavourable outcome will be defined as death or, in survivors, moderate or severe disability. 3.5 Post-mortem evaluation If the infant dies (UCH cases only), parents will be given options of conventional autopsy or minimally invasive autopsy (MIA) at Great Ormond Street Hospital (GOSH) using the protocol established for the MaRIAS (Magnetic Resonance Imaging Autopsy Study)\[20\]. 3.6 Qualitative interviews About 6 months after death or discharge a parental sub group from 20 NE cases recruited by stratified purposeful sampling from Royal Victoria Infirmary (Newcastle) will be invited to participate in an interview by a psychology Research Fellow. The parental understanding of, and attitudes towards use of, MR techniques for predicting long term outcomes and making decisions about withdrawal of life support (during hospital stay), will be collected thematically and analysed (aim 2.2.3). Tape recorded transcripts of parental interviews will be analysed and emerging themes grouped. 4\. ENDPOINTS AND DATA ANALYSIS 1H MRS thalamic \[NAA\] (mean (SD); mmol/kg wet weight brain tissue) in NE infants will be compared with outcome. Sensitivity and specificity with 95% CI and area under the curve (Standard Error) will be reported for \[NAA\] and Lac/NAA peak-area ratio and the clinical significances of unit changes in these measures in terms of unit changes in motor and cognitive outcome scores examined. The prognostic accuracy of early acquisition 1H MRS will be compared with late acquisition using area under curve and diagnostic odds ratios. The accuracy of \[NAA\] for prediction of outcome will be compared with conventional MRI. 1H MRS thalamic \[NAA\] (mean (SD)) in healthy term infants will be reported.