Striving for objective autism measures

Their training is rigorous­—and they use great skill in diagnosing and treating their patients—but without objective, quantifiable measures to characterize autism, clinicians who work with children who show symptoms of the disorder are at a disadvantage. Despite the “shrewd clinical eye” they bring to their practice, as James C. McPartland, Ph.D., associate professor of child psychiatry and psychology in the Child Study Center (CSC), describes it, they still largely rely “on the same tools that were used when autism was first written about in 1943.” Says McPartland, “We think we could do better, that we can measure the specific processes in the brain that are being impacted in autism.”

McPartland is principal investigator, and the School of Medicine is the lead institution, in a five-site research study called the Autism Biomarkers Consortium for Clinical Trials (ABC-CT). The goal of the four-year effort is to meticulously monitor 200 children with autism and 75 typically developing children over a six-month period, and use the findings to establish a reliable set of biomarkers that can determine the most appropriate treatments and measure any resulting improvements from treatment.

Investigators will analyze readings from electroencephalograms (EEGs) and also from devices that precisely measure eye movements—including factors such as changes in pupil size—when children are shown social information. EEG data are recorded from caps containing 128 electrodes that children wear on their heads. According to McPartland, “We are examining the brain’s activity at rest and also the brain’s response to things we think are socially important, like human faces and point light displays of people moving.”

They are doing so with unprecedented precision. From site to site, eye tracking equipment is identical, “right down to the manufacturer of the splitter cables,” says Adam J. Naples, Ph.D., associate research scientist in the CSC. “If you go through the literature,” Naples explains, “you can read different papers and see different equipment. These things can change the data quality and they matter.”

Naples traveled to Boston Children’s Hospital; Duke University; the University of California, Los Angeles; and the University of Washington, setting up equipment and training the staff at each location. “We have totally identical equipment, totally identical protocols, down to what you can say to a child who is not looking at the screen,” Naples says. “We have custom-built light meters. We can track the ambient light throughout the experiment. The level of standardization is unbelievable.”

McPartland says multiple sites are necessary for two reasons. “One is that it will be really hard to get this many children in one place at one time. Two, part of the goal of the research is to create this infrastructure so that the government and industry can use us as a tool for future clinical trials. Those kinds of trials necessitate multiple sites, because if I want to try a drug for a certain kind of autism, there may not be enough patients to do it in New Haven, but if I have a network of five sites across the country, we can get the required sample reasonably.”

Pharmaceutical companies are active participants in the National Institutes of Health (NIH)-led consortium that is funding and overseeing the project, because drug development is at a dead end without good biomarkers. “When a pharmaceutical company wants to see if its drug is working,” McPartland explains, “it is not adequate to bring that child in and have me play with them and say whether I’m seeing something subjective. It’s necessary to know that a particular neural circuit is being engaged and be able to quantify how it is changing.”

Biomarkers may also be invaluable in cases where patient evaluation is especially difficult. “Sometimes we have kids who are challenging to clinically conceptualize,” says Julie M. Wolf, Ph.D., assistant clinical professor in the CSC, who is one of the clinicians evaluating Yale Medicine patients enrolled in the study. “Every now and then you get those kids where you’re really on the fence. They have some traits of autism and other features that are not as consistent with the diagnosis, and you’re not sure how best to help the child. So, something more quantitative could be really helpful in those instances.”

Beyond the scientific and clinical ground it intends to break, ABC-CT is also an example of how the research infrastructure at the Yale Center for Clinical Investigation (YCCI) has advanced in ways that let medical school investigators do work that otherwise might be impossible. When McPartland learned of NIH interest in forming such a multisite consortium, he sought out Tesheia H. Johnson, M.B.A., M.H.S., deputy director and chief operating officer of YCCI and associate director for clinical research for the School of Medicine, who walked him through the ways YCCI could support this research.

As Johnson recalls, “I said, ‘Well, we can cover biostats, informatics, the statistical coordinating center. Who’s going to do your monitoring?’” YCCI was just forming its Multicenter Research Support Unit, which opened during the summer of 2016 as a standing resource for investigators who otherwise would need to assemble operational and administrative teams from scratch.

The unit worked with McPartland in writing the grant proposal, assured NIH of the resources it possessed in support of a scientist planning to direct an exceedingly complex study, and contributed personnel to the integrated ABC-CT team that McPartland now leads. Says Johnson, “The whole concept [for investigators] is to take your science and just plug it into a mechanism that’s all ready to go with professionals who know how to do the work.”

McPartland adds, “I think that YCCI has really specific expertise in areas that many investigators don’t know how to do well. Tesheia and her team were critical, and continue to be critical.”

Even before all the data come in, McPartland says children his team is assessing are already receiving some benefit. All participating children receive a detailed evaluation of their individual cognitive abilities, which results in a report that goes to their parents and teachers. "So, by virtue of being involved in our research, families and teachers are getting a better understanding of these children and how to help them.”

Even better, concludes McPartland, will be a day he hopes is not far away, when biomarkers his team develops begin to help patients directly. “It’s probably a field changer,” he says. “People with autism are complex, and we need nuanced tools.”