Autism: The Shift Towards Personalised Medicine

23rd Apr 2020

Our understanding of the autistic spectrum and its subtypes has changed dramatically since Kanner’s (1943) seminal paper, including its nature as a neurodevelopmental disorder. In contrast to the socially-constrained savant portrayed by Dustin Hoffman’s character in ‘Rainman’, individuals with autism (also termed autistic spectrum disorders [ASD] or conditions [ASC]) display enormous variability in cognitive, emotional, social and functional dimensions. Autism encapsulates individuals with low IQ, minimal cognitive function who display seizure-like activity on electroencephalographic (EEG) tests through to those who are cognitively healthy, within or above a normal IQ (100+), but may present with social deficits, a limited ability to internalise emotions (alexithymia) and difficulty understanding the emotional states of others as distinct from oneself (theory of mind).

Conversely, other phenotypes include alternative/additional symptoms such as comorbid attention deficit hyperactivity disorders (ADHD) or insensitivity to sarcasm, such as Asperger’s Syndrome.

Indeed, individuals with no clinical ASD features fall on the autistic spectrum given the spectrum encompasses ‘normal’ functioning. Neurophysiological functioning across social, cognitive and motivational domains potentially differs in different ASD strata, explaining why reward processing, executive function and other prominent cognitive features differ so much between individuals. A prominent difficulty in ASD research is that many large scale studies utilise ASD samples centred around one region, community or social group, and this sample is often constituted by multiple ASD subtypes, all of which may present differently on behavioural, neuropsychological or functional magnetic resonance imaging (fMRI) tests. Attempting to average these responses together to examine mean group differences against control ‘healthy functioning’/ 'typically developing’ individuals often dilutes strata variability. Thus, many subsequent studies and treatments target a hypothetical mean which realistically does not represent a given subgroup's neurofunctional/neurobehavioural profile. This adds to the difficulty in providing suitable pharmacological or cognitive treatments to individuals who desire to improve deficits (relative to ‘normal’ functioning).

Therefore, some autism initiatives are seeking to address deficits at the individual level by comparing an individual's performance in one domain against that of their performance across other domains. By using more engaging task stimuli than many conventional cognitive paradigms and layering behavioural, psychometric and fMRI data together, an individualised biobehavioural profile can be produced - specific to the individual - which may hold clinical value in guiding treatment. This approach is being pioneered by the departments of Neuroimaging and Forensic and Neurodevelopmental Sciences (FANS) at King’s College London in collaborations with a number of EU partners.