Diffusion‐weighted imaging and magnetic resonance spectroscopy findings in a case of neonatal hypoglycaemia

Abstract

SIR–We would like to present our findings of a case of neonatal hypoglycaemic encephalopathy due to transient hyperinsulinaemia investigated with magnetic resonance (MR), diffusion-weighted imaging (DWI), and proton MR spectroscopy (H-MRS). Our case highlights some interesting points and to our knowledge there is only one other case report examining the H-MRS appearances. The parents of the child described in this case report have consented for our findings to be published. A female born at term, birthweight 2.6kg, presented on day 2 of life with lethargy, poor feeding, and decreased responsiveness. The pregnancy and post-delivery checks were normal. Investigations revealed plasma glucose of 0.8mmol/L and bicarbonate of 12mmol/L. She received an intravenous 10% dextrose bolus of 2mL/kg followed by a glucose infusion. Over the next 24 hours, the infant had recurrent episodes of hypoglycaemia despite a high glucose infusion (10mg/kg/min) and she was commenced on diazoxide and chlorthiazide. Further tests revealed a non-ketotic hypoglycaemia. Her serum insulin levels were 9.4pmol/L with simultaneous plasma glucose level of 1.8mmol/L with an I/G ratio of 47.0 (local normal range 0.0–18.6) consistent with neonatal hyperinsulinaemia. On day 2, the child had several short unresponsive episodes that on electroencephalogram showed rhythmical discharges localizing in the left fronto-temporal region, with a relative excess of prominent delta slow waves. These seizures were controlled with phenobarbitone. Intracranial MR was performed on day 6 of life at 1.5T (Infinion; Philips Medical Systems, Best, Holland). Standard, whole-brain, turbo-spin-echo T2-weighted, T1-weighted, and gradient-echo T2*-weighted images were acquired. Whole-brain, single-shot, spin-echo-planar DWI datasets were acquired from which trace and apparent diffusion coefficient maps (ADC) were calculated. Single-voxel H-MRS was used to acquire spectra from the basal ganglia (15×15×15mm3) and occipital grey/white matter (10×10×10mm3) using point-resolved spectroscopy. Imaging revealed loss of grey–white matter differentiation in the occipital and parietal lobes bilaterally with a high T2-weighted signal in these regions (Fig. S1a; supporting information published online). Restricted diffusion was apparent with a high DWI signal and corresponding low signal on the ADC map (Fig. S1b, c; supporting information published online). The H-MRS showed a normal N-acetylaspartate to creatine ratio and normal lactate (Fig. 1). Normal magnetic resonance spectrum from occipital lobe. On day 15 her blood glucose stabilized and she was discharged home. Phenobarbitone was weaned off successfully after 3 months. Clinical follow-up at 15 months showed that she is seizure free, independently standing (since 12 mo), cruising along furniture, spoon feeding herself, and speaking 10 single words. Follow-up imaging has not been performed but is planned at 2 years. Our case highlights some interesting points. The mechanism of injury in neonatal hypoglycaemia is complex and not well understood. Established evidence shows that the injury has a predilection for the occipital and parietal lobes in neonates. Hypotheses for this include relatively higher glucose demands during growth periods and vulnerability to reduced glucose availability. DWI is useful as it not only shows changes more readily than T2-weighted imaging, but also allows differentiation between cytotoxic cell damage and vasogenic oedema. In our case, the occipital and parietal regions showed true restricted diffusion confirmed by hypointensity on the ADC map. This pattern indicates the mechanism of brain injury produces cytotoxic oedema as opposed to vasogenic oedema when the ADC is normal or high, as seen in posterior reversible encephalopathy syndrome. Interestingly in our case, the H-MRS was normal, thus indicating no substantial cell death. This is surprising in view of the DWI changes being consistent with cytotoxic oedema and may suggest there is a spectrum of severity and in our case the changes were not marked enough to cause abnormal spectroscopy. Our final point is that in our case the child’s outcome has been good. At 15 months following insult the child has normal neurodevelopment and no seizures. This is in contrast to the only other case report with spectroscopy findings in neonatal hypoglycaemia.1 In these two cases H-MRS revealed reduced N-acetylaspartate and increased lactate consistent with axonal loss and both infants have developmental delay and one has intractable seizures. These cases did have more extensive damage involving deep white matter than seen in our case, but we have seen several cases of neonatal hypoglycaemia with delayed development and seizures with damage localized to the occipital lobes on MR imaging. From this, we postulate that normal H-MRS findings in neonatal hypoglycaemia may indicate a more favourable outcome. Abnormal H-MRS has been shown to predict worse outcome in hypoxic–ischaemic encephalopathy and traumatic brain injury. We hypothesize it is also likely to be useful in neonatal hypoglycaemia as a predictor of long term outcome in cases where changes are confined to the occipital lobes as DWI may not be demonstrating cytotoxic oedema, but rather vasogenic oedema secondary to a vascular phenomenon as also seen in posterior reversible encephalopathy syndrome and post subarachnoid haemorrhage reversible vasospasm. Figure S1a: T2-weighted magnetic resonance imaging showing loss of grey–white matter differentiation in the parietal and occipital lobes in neonatal hypoglycaemia. Figure S1b: Diffusion-weighted imaging showing abnormal high signal in parietal and occipital lobes in neonatal hypoglycaemia. Figure S1c: Apparent diffusion coefficient map showing restricted diffusion in parietal and occipital lobes in neonatal hypoglycaemia. Please note: Wiley-Blackwell are not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.