Epilepsy summit report

Between 10% and 40% of autistic people also have epilepsy and this rate increases steadily with age – in contrast to a one per cent prevalence rate in the general population.

We hosted a two-day Global Summit including world leading researchers, autistic adults, relatives and professionals to develop world-class research ideas to investigate the co-occurrence of autism and epilepsy.

Why research autism and epilepsy?

Our 2016 report Personal Tragedies, Public Crisis highlighted the appalling rates of early death in autistic people relative to the general population, with epilepsy being a leading cause of death. The combination of disabilities significantly impacts on quality of life as well as physical and mental health, but remains under-researched.

It is more urgent than ever that we invest now to understand the interaction between autism and epilepsy and improve outcomes for autistic people who have epilepsy.

At Autistica we have committed to addressing the global need for increased and improved research into autism and epilepsy to tackle the unacceptable disparity in autistic life expectancy. Research could help us to understand the underlying mechanisms, risk factors, environmental triggers and developmental trajectory of the co-occurring disorders and how to better detect and treat seizures.

Research ideas from the Epilepsy Summit

Jump to: 1. Distribution and determinants of autism and co-occurring epilepsy 2. Understanding epilepsy onset in autism 3. Medication and onset of epilepsy 4. Epilepsy medication side effects 5. Autism characteristics versus epilepsy seizures 6. Stress / anxiety 7. Sleep 8. Biomarkers 9. Genetics – which genes? 10. Genetics – the biological consequences 11. Genetics – understanding seizures 12. Risk factors for premature death. 13. Premature death education

Why is this helpful?

Improve understanding of the pattern of co-occurrence.
Potentially highlight unknown risk factors and mechanisms of the co-occurrence (e.g. specific seizure types, age of onset etc.).
Enable future research that can save care and management costs

How would the research be designed?

Form an international expert group to conduct a scoping review of academic literature and identify relevant international datasets.
Perform a meta-analysis - a statistical approach to combine the results from multiple studies in an effort to improve estimates.

Who would be involved?

Partnerships would be formed with autistic people and an international multidisciplinary research team including epidemiologists and biostatisticians.

Things to consider

Definitions of autism and epilepsy and mode of ascertainment.
Whether international differences in terminology would need to be adjusted in datasets.
Whether datasets are derived from surveillance or survey.
Whether other conditions could be included, such as catatonia and chronic fatigue.
How far this study has been achieved, and whether existing work could be built upon.

Why is this helpful?

Determine who is at risk to enable early prevention measures and/or treatment.
Identify the frequency, duration, severity, type and treatment response of seizures to inform treatment plans.

How would the research be designed?

Conduct a prospective study following multiple cohorts for five years across the age range 0-25 years to understand more about the onset and course of epilepsy in autism.
Potential risk and protective factors (e.g. family history, prenatal and perinatal environment, and hormones) would be measured by repeated questionnaires and genetic samples. The use of wearable sensors and/ or brain activity measures i.e. electroencephalogram (EEG) could also be explored.

Who would be involved?

Partnerships would be formed with autistic people, third sector organisations, and a multi-disciplinary research team including epidemiologists, biostatisticians, geneticists and endocrinologists.
The project could be multi-site, capitalising on existing cohorts.
Participants would have an established childhood diagnosis of autism. Both genders would be well represented and people with learning disability would be included.

Things to consider

Recruitment strategy, including consideration of late autism diagnoses.
Retention strategies and a protocol for participant drop-outs, considering how this will affect the age distribution of the study.
Whether data / genetic samples could be collected from parents and/or siblings.
Whether the age range could be extended beyond 25 years.

Why is this helpful?

Build knowledge of associations between medication usage and epilepsy onset in autism to inform safe medication practices.

How would the research be designed?

Utilise NHS datasets to look at prescribing patterns and epileptogenesis in autistic people with late-onset epilepsy and compare with a control group of autistic people without epilepsy.
Establish associations between medications and onset of epilepsy.
Consider the dose and duration of medication use and time lag to symptom onset.

Who would be involved?

Partnerships would be formed with autistic people, relatives, Department of Health, NHS England and the devolved nations, NHS Digital, Public Health England and researchers.
Participants would be autistic people whose epilepsy began between 12 and 30 years of age.

Things to consider

How to determine whether an association is meaningful.

Why is this helpful?

Build knowledge of epilepsy medication side effect profiles.
Inform treatment plan amendments to improve quality of life and life expectancy.

How would the research be designed?

Telephone or virtual interviews to establish a detailed understanding of medication history, adherence, and side effect profiles.
Compare to pharmaceutical companies’ published side effect lists.
Participants would be recruited via existing cohorts (ideally with contemporaneous clinical data). There would be an initial focus on people in the typical IQ range, with people with a learning disability included at a later
stage.

Who would be involved?

Partnerships would be formed with autistic people, relatives, third sector organisations, and researchers.
Participants would be 100-300 autistic people, families and carers.

Things to consider

How to account for recall bias when retrospectively discussing side effects.
How to distinguish between medication side effects and autism characteristics.
The effect of medication adherence on side effect profiles.
Gender differences in side effect profiles.
Age range of the sample - parents could be asked about side effects for young children.
Whether reviewing medical records and prescription forms might be beneficial to the study design.

Why is this helpful?

Optimise treatment (e.g. withdrawal of anti-epileptic drugs).
Improve participant identification for research studies.

How would the research be designed?

Conduct a scoping review of the application of wearable technology to detect seizure in epilepsy and engage research groups to add an autism comparison group to ongoing epilepsy studies.

Who would be involved?

Partnerships would be formed with autistic people, companies specialising in wearable technology and a multi-disciplinary research team including experts in epilepsy, signal detection and behaviour observation.

Things to consider

Definitions of seizures and ‘other behaviours’.
Stratification of the sample (e.g. by age and autism characteristics severity).
Whether ‘other behaviours’ could be related to later onset of seizures.
Whether certain behavioural profiles are associated with different seizure treatments.

Why is this helpful?

Inform the design of stress / anxiety management tools, which can be tested for efficacy of stress / anxiety and seizure reduction.
Improved quality and duration of life.

How would the research be designed?

Use a mixed method approach, including focus groups and objective measures (e.g. wearable devices; galvanic skin conductance), to explore the factors leading to anxiety in autistic people with epilepsy and develop a severity scale.

Who would be involved?

Partnerships would be formed with autistic people with epilepsy and lived experience of stress and/or anxiety and a multi-disciplinary research team including sociologists, psychologists, and clinicians with experience of autism and epilepsy.
The sample would be stratified by age, gender, ability level – and later by seizure type and stress / anxiety level. Medication use would also be taken into consideration.

Things to consider

Definitions of stress / anxiety and epilepsy.
How far existing stress / anxiety scales apply to autistic people with epilepsy.
Whether measures of depression can be integrated.
Impact of participant lifestyle (e.g. exercise).

Why is this useful?

Better understanding of how sleep can be managed for autistic people with and without epilepsy, improving daily living for them and their families.
Inform understanding of how sleep could relate to seizure onset.

How would the research be designed?

Utilise wearable technology to investigate group differences in a) behavioural impact of sleep (e.g. anxiety) and b) physiological impact of sleep (e.g. drug absorption).

Who would be involved?

Partnerships would be formed with autistic people and relatives, companies specialising in wearable technology, pharmacists, and a multidisciplinary research team including experts in sleep.

Things to consider

How sleep quality will be defined and measured, taking into account that ‘sleep’ may not be characterised in the same way as in the general population.
Whether sleep quality measured during the study represents general sleep quality.
The expense of using wearable technology.
Whether a measure of brain activity i.e. electroencephalogram (EEG) can be integrated.
Whether pharmacogenomics - the study of how genes affect a person’s response to drugs – can be integrated.
Whether an epilepsy only comparison group might be beneficial to the study design.

Why is this useful?

Improved knowledge of risk of developing epilepsy in autism.
Potential to inform prevention strategies.
Potential to inform treatment options.

How would the research be designed?

Longitudinal study with multimodal data collection, including classic brain activity measures i.e. electroencephalogram (EEG), neuroimaging, blood-based biomarkers, wearable devices / activity trackers, and medication use. High density EEG data collection and investigation of participants’ microbiome would also be considered.
Machine learning / big data analysis approaches (e.g. connectivity, cluster, features selection and dimensionality reduction) would be used to investigate potential biomarkers.
A qualitative study (e.g. focus groups or interviews) evaluating the experiences of autistic people and families.

Who would be involved?

Partnerships would be formed with autistic people and relatives, companies specialising in wearable technology, and a multi-disciplinary research team including experts in electrophysiology, neuroimaging, metabolomics and machine learning data analysis.
Cohorts with high risk of autism and epilepsy (e.g. genetic disorder groups) may be included to enhance the sample.

Things to consider

Incentive strategy to prevent participants from dropping out.
How to avoid the research procedures, which some people may find intrusive, preventing people from taking part or inducing anxiety and potentially triggering seizures.
How co-occurring conditions might affect biomarker correlations.
How medication may affect data.
Whether the inclusion of a control group, such as people with epilepsy who are not autistic or autistic people without epilepsy, might be beneficial to the study design.

Why is this useful?

Inform screening and early diagnosis.
Enable genetic counselling - the process by which people or relatives at risk of an inherited disorder are advised of the consequences and nature of the disorder, the probability of developing or transmitting it, and the options open to them in management and family planning.
Inform the development of prevention and treatment options.

How would the research be designed?

The project would focus on assembling a large cohort of autistic people, people with epilepsy, and autistic people with epilepsy with existing genetic data to enable the investigation of the genes that drive the conditions.

Who would be involved?

Partnerships would be formed with autistic people, relatives, Health Maintenance Organizations (e.g. Kaiser Permanente and Geisinger), an outreach team and a multidisciplinary research team including geneticists, biologists, psychiatrists and neurologists.
Data could be collated from existing cohorts in the US and UK and mined for relevant data (e.g. brain activity measures i.e. electroencephalogram (EEG), epileptiforms in autism cohorts and neuropsychology measures in epilepsy cohorts).

Things to consider

The types and quality of genetic data available.
How autism and epilepsy are defined with regard to inclusion and exclusion criteria.
Whether other common co-occurring conditions should be included.
Timeframe to establish cohort, gain access to data and investigate data.
Social consequences of identifying genetic drivers e.g. stigma.

Why is this useful?

To reveal potential drug targets.
Develop potential therapies based on biological mechanisms.

How would the research be designed?

Regular (less than one per year) pathway analyses of genes would be conducted.
Causative genetic changes would be modelled through a) animal models of single gene changes and b) induced pluripotent stem cells of copy number variants and complex genetics.
Biological consequences in the models would be examined and interrogated through a) synapses b) neuronal morphology c) neuronal differentiation d) neuronal activity e) behaviour f) identified biomarker (see project 8).

Who would be involved?

Partnerships would be formed with autistic people, relatives, an outreach team and a multidisciplinary research team including geneticists, biologists and experts in electrophysiology and bioinformatics.

Things to consider

How to account for social and environmental factors.
How this project could link to establishing an autism and epilepsy genetics cohort (project 9).

Why is this useful?

Greater understanding of seizure characteristics.
Better preparation for seizure onset.
Improved and targeted seizure management.

How would the research be designed?

Population study of autism and seizures.
Exclusion criteria of a) prematurity, b) evidence of prenatal or perinatal injury, c) single gene disorder or cortical malfunctions with no variation in phenotype outcome (e.g. treatment resistance).
Genotyping and epigenetics may be done on existing ASD and epilepsy cohorts.

Who would be involved?

Partnerships would be formed with autistic people, relatives and a multidisciplinary research team including geneticists and experts in bioinformatics.

Things to consider

Specificity and justification of exclusion criteria.
How to acquire treatment histories for participants.
Whether ‘treatment resistance’ includes surgery and/or brain stimulation as well as medication.
How this project could link to establishing an autism and epilepsy genetics cohort (project 9).

Why is this useful?

Generate accurate information on the risk of premature death in autism and epilepsy.
Enable premature death prevention strategies to be identified.

How would the research be designed?

Literature review for causes (and age) of death among autistic people with epilepsy.
Interrogate death records for cause of death using postmortem reports and health records.
Look for cognitive ability and health conditions (e.g. cardiac, respiratory, mental health).
Use statistical analyses to consider group differences across cohorts of autism, epilepsy, and autistic participants with epilepsy.

Who would be involved?

Partnerships would be formed with autistic people, families, caregivers and researchers, including epidemiologists.

Things to consider

Whether family interviews will be required to establish diagnoses and cause of death due to poor quality of death records.
Influence of other co-occurring conditions on premature death.
Comparison to premature death in intellectual disability.

Why is this useful?

Provision of accurate information on the risk of premature death in autism and epilepsy to enable stakeholders to make informed decisions.
Enable premature death prevention strategies to be identified.

How would the research be designed?

Qualitative study interviewing autistic people, families and caregivers to understand existing knowledge of premature death (including SUDEP), preferred method of information dissemination (e.g. face-to-face, media, written), attitudes towards medication adherence, and knowledge of brain donation possibilities.

Who would be involved?

Partnerships would be formed with autistic people, families, caregivers, third sector organisations such as SUDEP Action and qualitative researchers.

Things to consider

If it is possible to build predictors of SUDEP and sensitively communicate these to autistic people (in an autism-friendly way) and families.

What happens next?

This workshop is only the beginning of the process. We are committed to taking the research concepts forward and evolving them into fundable research proposals that will lead to longer, happier and healthier lives for autistic people. We will work with partner organisations and funders over the coming months to get the remaining questions addressed.

How can you get involved?

Follow us on social media for the latest research news and developments on this project or Join our Autistica Network for updates.

If you think you can help in making this research happen, get in touch with Lorcan Kenny at lorcan.kenny@autistica.org.uk.