570 VOLUME 22 | NUMBER 6 | JUNE 2016 NATURE MEDICINE has capital letters that you can then put into a search engine,” she says. One common measure is the Autism Diagnostic Observation Schedule (ADOS), a diagnostic tool developed by Lord that relies on clinician observations. The ADOS is reliable, but it’s also lengthy and requires substantial expertise. “It takes 45 minutes and a very skilled clinician to do it,” Lord says. That is doable if a study includes a dozen participants. However, for a large drug trial involving hundreds of children, the test becomes prohibitively expensive. And the tool doesn’t do a good job of measuring subtle changes over weeks or months. “It’s better suited to measuring presence or absence of the disorder, and a little less well suited to measuring individual differences in severity of certain symptoms,” says Stephen Sheinkopf, an autism researcher at Brown University in Providence, Rhode Island. Drug trials often rely on another tool called the Aberrant Behavior Checklist (ABC), says Connie Kasari, an autism researcher at the University of California, Los Angeles. In fact, the trials of the only two drugs on the market for children with autism relied on this measure to show improvement. The ABC is usually completed by a parent, and according to Scahill, it has many advantages over other outcome measures. “It’s reliable, it’s valid, it measures something that is reasonably clear and it’s sensitive to change,” he says. But there are also drawbacks: Scahill points out that it was not developed for autism. And Kasari notes that the results can be biased, especially if someone who is invested in the research completes the checklist. Spooren notes that tools that depend on parent reports are prone to strong placebo effects. “If you have a child with autism that is really sick and he or she is in a drug study, you just have this incredible hope that’s it’s going to work,” he says. In January, researchers reported that a Novartis drug called mavoglurant failed to show any benefit in two studies of patients with fragile X syndrome, a genetic disease that can cause ASD. Earlier trials of a different drug being developed by Roche and Seaside Therapeutics also failed to show any benefit. Although it is not clear why these drugs failed, the outcome measures could have played a part, says Mustafa Sahin, a neurologist at Boston Children’s Hospital in Massachusetts who studies childhood neurological diseases. “They looked very It has always been tricky to develop therapies for autism, in part because the disorder is so diverse. Some children with autism don’t speak at all; others hardly stop talking. Some have IQs so low that they are considered to have an intellectual disability; others score so high they would be classified as extraordinarily gifted. “Autism is probably 100 diseases,” says Will Spooren, head of Behavioural Pharmacology and Preclinical Imaging at Hoffmann–La Roche in Basel, Switzerland. Researchers have yet to understand fully the neural circuitry that underlies autism spectrum disorder (ASD). But there is another problem that hampers drug development: although there are hundreds of tools to assess treatment response in children with autism, most of them are deeply flawed. The majority of these measures were not designed for children with autism. Many weren’t even designed as outcome measures. Experts point out that even commonly used measures are prone to bias and cannot reliably detect improvements in symptoms. Consequently, even medicines that could help children with autism might fail to show any benefit in a clinical trial. That makes drug companies with an interest in the disorder reluctant to move forward, says Lawrence Scahill, an autism researcher at Emory University in Atlanta, Georgia. But better tools are on the way. Last year, two massive initiatives were launched by institutions in the US and Europe to identify new biomarkers for autism that could help researchers to stratify individuals with the disorder into subgroups for clinical trials, or that could even serve as surrogate endpoints. In addition, the Simons Foundation Autism Research Initiative (SFARI), based in New York, awarded $2 million in January to four research groups working to develop new measures. And in April, Catherine Lord, director of the Center for Autism and the Developing Brain in White Plains, New York, and colleagues published a new measure for social disability in young children with autism1. Because there are several promising medications for autism that have shown promise in clinical studies, “there’s a sense of urgency and attention to this whole matter,” Scahill says.
Download Full PDF Version (Non-Commercial Use)