Multimodal Non-Sedated MRI during an RCT of Simvastatin in Neurofibromatosis Type 1 (NF1)-Syndromic Autism.

Background: Understanding neurochemical and physiological dysregulation is thought to have potential as a basis to model developmental abnormalities in autism. The most common non-invasive technique used to acquire measures of both anatomical, neurochemical and physiological parameters in the developing brain is multi-modal Magnetic Resonance Imaging (MRI). Simvastatin is an HMG-CoA reductase inhibitor.

Objectives: To assess the feasibility and results of multi-modal MRI assessment of a range of neurochemical and neuro-physiological measures in non-sedated, awake children with Neurofibromatosis Type 1 (NF1) syndromic autism undergoing an early phase trial of Simvastatin intervention.

Methods: The patient cohort was acclimatised to awake MRI scanning using a social story and two weeks of exposure to mp3 recordings of specific MRI scan sequence noises at specified time points during the day when awake and during sleep preparation. Scanning was performed at baseline, then following twelve weeks of simvastatin or placebo exposure.

We measured the concentration of GABA and other neuro-metabolites in the left frontal white matter and deep grey nuclei using a single-voxel MEGA-PRESS sequence. We also acquired anatomical data using a whole brain gradient echo T1 volume acquisition with voxel size of 0.9mm³. Brain perfusion measures were assessed using arterial spin labelling (ASL), whilst six-directional diffusion tensor imaging (DTI) was used to determine regional apparent diffusion coefficient (ADC) and regional fractional anisotropy. MEGA-PRESS data were analysed with the jMRUI v.5 software; the AMARES routine, was used to calculate metabolite concentration with tissue percentages extracted from a participant's T1 images for partial volume correction.

Results: In total 31 children presented for imaging, with 26 completing both baseline and week twelve imaging assessments (mean age 7.9, range 4.6-10.4). Absolute measures of GABA in the frontal white matter were seen to significantly increase in the treatment group (p=0.02) compared to the placebo group. Analysis for measure of change also identified significance at the 5% level in ASL perfusion in the ventral diencephalon and ADC in the cingulate gyrus in the treatment group, again not seen in controls. Finally there was significance in the interval change in glutamine + glutamate in the deep grey nuclei (p<0.05) again in the treatment group and not in the control cohort.

Conclusions: We have demonstrated that in this age group, made more challenging by the diagnosis of autism, careful patient preparation can result in successful multi-modal MRI assessment of both neuro-chemistry as well as physiological imaging in the awake patient. This is important as the effects of anaesthesia or sedation on the brain will impact on multi-modal imaging findings in such studies.

We have shown significant GABA changes after simvastatin exposure in this age group of NF1 autism. We also present the specific neurochemistry, perfusion and diffusion changes seen in these children indicative of measurable brain physiology changes after twelve weeks of treatment. These changes may be used to illuminate the pathogenesis of autism in this cohort, as well as being used as biomarkers for future phamaco-intervention studies.