According to a study by Weill Cornell Medicine researchers, people with autism spectrum disorders can be classified into four distinct subtypes based on their brain activity and behavior.
The study, published in Natural neuroscience, leveraged machine learning to analyze newly available neuroimaging data from 299 people with autism and 907 neurotypical people. They found patterns of brain connections linked to behavioral traits in people with autism, such as verbal ability, social affect and repetitive or stereotyped behaviors. They confirmed that the four autism subgroups could also be replicated in a separate dataset and showed that regional gene expression differences and protein-protein interactions explain brain and behavioral differences.
“Like many neuropsychiatric diagnoses, people with autism spectrum disorders experience many different types of difficulties with social interaction, communication, and repetitive behaviors. Scientists believe there are likely many different types of autism spectrum disorders who might require different treatments, but there is no consensus on how to define them,” said co-lead author Dr. Conor Liston, associate professor of psychiatry and neuroscience at the Institute for Cancer Research. brain and mind of the Feil family at Weill Cornell Medicine.”Our work highlights a novel approach to uncovering subtypes of autism that may one day lead to novel approaches to diagnosis and treatment.”
A previous study published by Dr. Liston and colleagues in Nature Medicine in 2017 used similar machine learning methods to identify four biologically distinct subtypes of depression, and subsequent work has shown that these subgroups respond differently to various therapies for depression.
“If you put people with depression in the right group, you can assign them the best therapy,” said lead author Dr. Amanda Buch, postdoctoral neuroscience associate in psychiatry at Weill Cornell Medicine.
Building on this success, the team set out to determine whether there are similar subgroups among autistic people and whether different genetic pathways underlie them. She explained that autism is a highly inherited disease associated with hundreds of genes that have a varied presentation and limited treatment options. To study this, Dr. Buch pioneered new analyzes to integrate neuroimaging data with gene expression data and proteomics, bringing them to the lab and allowing testing and developing hypotheses about how risk variants interact in autism subgroups.
“One of the barriers to developing therapies for autism is that the diagnostic criteria are broad and therefore apply to a large and phenotypically diverse group of people with different underlying biological mechanisms,” Dr. Buch said. . “To personalize therapies for people with autism, it will be important to understand and target this biological diversity. It is difficult to identify the optimal therapy when everyone is treated as the same, when everyone is unique.
Until recently, there were not enough collections of functional magnetic resonance imaging data from people with autism to conduct large-scale machine learning studies, Dr. Buch noted. But a large dataset created and shared by Dr. Adriana Di Martino, research director of the Child Mind Institute’s Autism Center, along with other colleagues across the country, provided the large dataset needed to the study.
“New machine learning methods capable of handling thousands of genes, differences in brain activity, and multiple variations in behavior made the study possible,” said the co-lead author. Dr. Logan Grosenickassistant professor of neuroscience in psychiatry at Weill Cornell Medicine, who pioneered machine learning techniques used for biological subtyping in studies of autism and depression.
These advances allowed the team to identify four clinically distinct groups of people with autism. Two of the groups had above average verbal intelligence. One group also had severe social communication deficits but less repetitive behaviors, while the other had more repetitive behaviors and less social impairment. Connections between parts of the brain that process visual information and help the brain identify the most salient incoming information were overactive in the subgroup with more social impairment. These same connections were weak in the group with more repetitive behaviors.
“It was interesting at the brain circuit level that there were similar brain networks involved in these two subtypes, but the connections in these same networks were atypical in opposite directions,” said Dr. Buch, who completed his doctorate at the Weill Cornell Graduate School. of Medical Sciences in Dr. Liston’s lab and now works in Dr. Grosenick’s lab.
The other two groups had severe social impairments and repetitive behaviors, but had verbal abilities at opposite ends of the spectrum. Despite some behavioral similarities, the researchers found completely distinct brain connection patterns in these two subgroups.
The team analyzed gene expression that explained the atypical brain connections present in each subgroup to better understand what caused the differences and found that many genes were previously linked to autism. They also analyzed network interactions between proteins associated with atypical brain connections and searched for proteins that could serve as a hub. Oxytocin, a protein previously linked to positive social interactions, was a central protein in the subset of individuals with more social impairment but relatively limited repetitive behaviors. Studies have looked at the use of intranasal oxytocin as a therapy for people with autism with mixed results, Dr. Buch said. She said it would be interesting to test whether oxytocin therapy is more effective in this subgroup.
“You might have a treatment that works in a subgroup of people with autism, but that benefit disappears in the larger trial because you’re not paying attention to the subgroups,” Dr. Grosenick said.
The team confirmed their results on a second set of human data, finding the same four subgroups. As a final check on the team’s findings, Dr. Buch conducted an unbiased text-mining analysis she developed on the biomedical literature that showed other studies had independently connected genes linked to the autism with the same behavioral traits associated with the subgroups.
The team will then study these subgroups and potential treatments targeted at the subgroups in mice. Collaborations with several other research teams with large human datasets are also underway. The team is also working to refine its machine learning techniques.
“We’re trying to make our machine learning more cluster-aware,” Dr. Grosenick said.
In the meantime, Dr. Buch said he has received encouraging feedback from people with autism about their work. An autistic neuroscientist spoke to Dr. Buch after a presentation and said his diagnosis was confusing because his autism was so different from others, but his data helped explain his experience.
“Being diagnosed with an autism subtype might have been helpful to him,” Dr. Buch said.