The Biomarkers of Neurological Disease in Utero Study

BACKGROUND There are approximately 50 million people with epilepsy worldwide, with approximately 3-5 births per thousand to women with epilepsy; in the United Kingdom (UK) there are 2500 births per year to women with epilepsy. The teratogenic effects of anti-epileptic drugs (AEDs) on the fetus is a major public health concern. In 2016 UK national prescribing data showed around 18,000 women of child-bearing age were taking Sodium Valproate. for which current evidence suggests has the most potent teratogenic effects of all AEDs. The Medicines and Healthcare Products Regulatory Agency (MHRA) has formulated a specific prevention programme aimed at reducing the rates of valproate-associated teratogenicity. Longitudinal studies of fetal AED exposure show increased rates of congenital malformations, abnormal motor development, lower intelligence quotient, and poorer educational attainment. However, to date there is a major lack of research on the underlying neuroanatomical basis for such alterations in cognitive function, whether pre or post-natal.

Early exposure to neuromodulatory drugs likely induces abnormalities in structural brain development, neuromotor behaviour, and functional networks. Fetal MRI presents a novel multimodal approach for interrogating the effects of pathological exposures, identify biomarkers, define onset of abnormalities, and the dose response. This may aiding risk stratification, identify candidates for earlier intervention, and improve neurodevelopmental outcomes.

There are limited strategies available to the clinician to assess the risk to the fetus' brain health. Current guidance recommends Ultrasound Scan (USS) to screen for physical defects. However, USS is limited in its ability to comprehensively understand brain development. Fetal MRI presents an advance that is of clinical utility to neurologists, psychiatrists and obstetricians who routinely prescribe AEDs for a range of neurobiological conditions. Advances in fetal MRI have identified biomarkers for disease that may guide clinical management, and offers the potential of a multi-dimensional characterisation of the developing nervous system by analysing brain structure, neuronal networks, expression of metabolites and behavioural analysis, which may also provide the basis for long-term adverse outcomes. A small unpublished case series from the Chief Investigator's doctoral study elucidated several promising features of fetal MRI : (i) MRI confirms USS findings and can improve the characterisation of brain structure; (ii) longitudinal in utero imaging is feasible and sometimes necessary; (iii) brain anomalies are present and may be attributable to exposure to AEDs; (iv) clinical management of pregnancies and the postnatal period is directly impacted by fetal MRI.

This study will undertake a pilot study using fetal MRI on 20 pregnant women with the aim of developing a tool to further characterise in utero brain development. This is a pilot study which will include 2 cohorts: women with epilepsy taking AEDs (WWE) (N=10), and age-matched healthy pregnant controls (HC) with no known neurological disorder (N=10).

RESEARCH QUESTIONS The aim of this study is to investigate the adverse neurological effects on a fetus exposed to Anti-Epileptic Drugs.

STUDY DESIGN This case control observational study aims to characterise fetal brain development in the context of exposure to Anti-Epileptic Drugs. Participants will attend a single session at the SPMIC, University of Nottingham, in addition to their clinical antenatal care. Pregnancy outcome will be collated from clinical records. At their SPMIC session, they will complete a simple demographics data collection questionnaire on their personal medical history, followed by a 3 Tesla MRI Fetal MRI scan. There is no requirement for randomization or blinding. Each participant will be assigned a study identity code number, allocated at randomisation for use on study documents and the electronic database. The documents and database will also use their initials.

The Chief Investigator has overall responsibility for the study and shall oversee all study management. The data custodian will be the Chief Investigator. The Chief Investigator and co-investigators will meet regularly for the duration of the study to discuss the project implementation and overall progress.

The study is planned to be implemented over over 5 years from 2019. This will allow for recruitment and scanning of participants, and then a period of allowing participants to deliver the fetus, and post-partum outcome information to be collected. The end of the study will be marked by the final outcome from the last participant to deliver.

The imaging will all take place at the Sir Peter Mansfield Imaging Centre (SPMIC; University Park, University of Nottingham). Before commencing the scan the pregnant woman will be asked to fill in the SPMIC safety form to screen for any contraindications of MRI. The MRI will take up to one hour to perform and women will be in the centre for about 1-2 hours. Demographic and outcome data from the pregnancy will be collected by the research staff during the study. Women will be asked if they wish to be informed of the results of the study and asked to leave contact information for this purpose (Email or address).

MRI data will acquired using sequences including, but not limited to, T2-weighted structural, functional MRI (fMRI), Diffusion Tensor Imaging (DTI), and Cine (for motor analysis). Structural data will be acquired in 3 orthogonal planes, registered and reconstructed for further automated tissue segmentation. Segmented tissues will be used for volumetric analysis and the development of shape analysis methods. The fMRI data will analysed for resting state network differences between cohorts, and differences in Blood-Oxygen Level Dependent response during motor activity. DTI data will be analysed according to an analysis pipeline described recently, and will allow comparison of differences in tract formation between cohorts. Motor behaviour will be analysed according to principles of General Movement analysis between cohorts.

MRI is a safe, non-invasive and non-ionizing clinical imaging modality. Concerns naturally arise when using any imaging modality in the fetus, both in terms of immediate teratogenicity and long term development effects. In this regard, UK guidelines do not recommend MRI prior to 18 weeks gestation, unless there are exceptional clinical concerns. The magnitude of the primary magnetic field does not occur in natural phenomena, however there are no known adverse biological effects due to such interactions. However repeated radiofrequency pulses result in energy deposition in tissues and can lead to heating effects. This is measured by the specific absorption rate and the distribution of thermal energy deposition has been investigated in pregnancy. Animal models have demonstrated that thermal energy generated by metabolic processes in the fetus dissipate through two routes: thermal energy transfer from fetal to maternal blood in the placenta and directly from the fetal tissues into the surrounding amniotic fluid. Clinical studies into the long term effects of MRI exposure to fetuses have also shown no adverse outcomes in terms of growth deficits, or changes in physiological parameters.

Descriptive statistics will be produced of the variables collected. Graphical displays will be generated as appropriate. Data summaries for continuous variables will be expressed as median and interquartile ranges and for categorical variables will be expressed as n (%). Further statistical testing will depend on the distribution of the data. Parametric (e.g. t tests) and non-parametric statistics (e.g. Mann-Whitney U-test) will be used to compare differences in the continuous and non-continuous variables respectively, between groups. Chi-square test or Fisher's exact test will be used to examine associations between categorical variables. Limits of agreement and Bland-Altman plots will be used to examine intra- and inter-observer reliability of measurements.

STUDY SETTING Demographic and medical history will be collected from two sources: hospital records, and directly from a patient administered questionnaire on the day of the SPMIC visit. MRI data will be acquired using the 3 Tesla wide-bore MRI scanner at SPMIC, University of Nottingham. The choice of scanner permits greater participant comfort particularly later in gestation where changes in body habitus would benefit from a larger circumference scanner. There is also a substantial amount of experience for scanning pregnant women on this scanner.

Patients will be recruited from a single site at the Joint Obstetrics-Epilepsy Service at Nottingham University Hospitals, which has specifically designed to manage women with epilepsy that are both considering or are currently pregnant. The clinical team consists of Consultant Obstetricians, Consultant Neurologists/Epileptologists, Obstetrician Specialist Trainees, Physician Specialist Trainees, and Specialist Nurses.