Measuring neural activity in language-sensitive brain regions of toddlers with autism spectrum disorder (ASD) may help identify those who go on to have poor language outcomes, researchers have discovered in a groundbreaking study.
By combining early brain imaging with performance on behavioral tests, the team could accurately predict language outcomes, thus helping to explain some of the heterogeneity and different developmental trajectories seen in children with ASD.
"These findings open insight into the first steps that lead to different clinical and treatment outcomes, and in the future, one can imagine clinical evaluation and treatment planning incorporating multiple accurate behavioral and medical prognostic assessments," coauthor Karen Pierce, PhD, codirector of the University of California, San Diego, Autism Center, said in a release.
"That would be a huge practical benefit for families."
The study was published online April 9 in Neuron.
High Level of Accuracy
The researchers recruited 60 children aged 18 to 24 months with ASD, matching them by sex and age to 19 language/developmentally delayed individuals and 24 typically developing individuals.
The participants underwent natural sleep functional magnetic resonance imaging (fMRI) to record neural activity while they listened to excerpts from children's stories that assessed complex forward sleep, simple forward speech, and backward speech.
The examinations were repeated approximately every 6 months, and each child underwent clinical assessment an average of 1 year following the original fMRI session.
The researchers identified 24 children with ASD who had poor language outcomes in childhood, defined as receptive and expressive language T scores on the Mullen Scales of Early Learning that were less than one standard deviation (SD) of the norm.
In contrast, 36 children with ASD had better language outcomes, defined as language scores greater than one SD below the norm.
Children with good language outcomes had normal patterns of neural activity in regions such as the superior temporal cortex on earlier fMRI assessment, whereas those with poor language outcomes had little activity in language-sensitive brain regions.
Crucially, the team was able to demonstrate that combining neural activity differences with performance on behavioral tests enabled them to predict language outcomes in early childhood with an accuracy of 80%, compared with 68% for each measure alone.
Distinct Subgroups
Explaining the significance of the research, first author Michael Lombardo, PhD, of the University of Cyprus and the University of Cambridge, commented that the findings underline the fact that there "is substantial heterogeneity within an omnibus diagnostic label like autism."
"It's probably better for us to move away from that thinking of autism as a singular thing, and more about how we can parse apart clinically meaningful heterogeneity within those individuals who all have the same diagnostic label of 'autism,' " he added.
"Coming back then to our study, what we've done here is to take something like early language development and then show that there are distinct subgroups within the omnibus 'autism' label," Dr Lombardo told Medscape Medical News.
"These subgroups also then happen to be neurally differentiated too, and that key point stands in contrast to the idea of treating autism as a singular thing...the challenge is to now parse apart the meaningful distinctions amongst that heterogeneity."
Dr Lombardo explained that if we are to understand the pathogenesis, it is perhaps better to think of brain differences in autism as a phenotype, but at the neural level.
"The causal mechanisms behind all this are probably things down within the genome, environmental facts, a combination of the two, and/or other kinds of molecular pathophysiological changes," he said.
The hope is that, from this greater understanding, novel interventions for autism may one day be developed. "One reason why most treatments for autism as a whole don't work well on everyone with autism is because they are probably only effective for some subset of individuals," Dr Lombardo explained.
"So by coming up with clinically relevant stratifications, hopefully we are improving the situation for future research that might want to hone in on specific causal biological mechanisms that could be amenable to treatment."
"We don't know what those are just yet, but the hope is that by having such stratifications, we will be much more sensitive at picking up on such mechanisms."
"Groundbreaking" Methodology
Commenting on the findings for Medscape Medical News, Anthony L Rostain, MD, professor of psychiatry and pediatrics at the Perelman School of Medicine at the University of Pennsylvania, in Philadelphia, described the study as a "very interesting and convincing analysis of brain differences in children with autism."
"What this study does is help identify which subjects seem to be underfunctioning, and what they did very nicely is show that, in many ways, the neuroimaging signature, if you will, can be detected early on and be a predictor of later course," he said.
However, for Dr Rostain, one of the most important aspects of the study lay in its methodology.
"It's groundbreaking stuff, in the sense that we now have large enough samples that are being followed across time," he said.
"With this study in particular, they were able to show that the development of specific language centers can be correlated with the outcomes we see clinically."
"It also helps us understand why there is such heterogeneity in autism. That's always been one of those questions that both researchers and clinicians ― and families and teachers and whoever ― are always scratching their heads about and trying to figure out," Dr Rostain concluded.
This work was supported by National Institute of Mental Health Autism Center of Excellence and the National Foundation for Autism Research grants and fellowships from Jesus College, Cambridge, and the British Academy, United Kingdom. The authors have disclosed no relevant financial relationships.
Neuron. Published online April 9, 2015. Abstract
Medscape Medical News © 2015 WebMD, LLC
Send comments and news tips to news@medscape.net.
Cite this: Early Neural Changes Predict Autism Language Outcomes - Medscape - Apr 17, 2015.
Comments