Can Autism Therapy Be Tailored?

Over the last decade, genetic sequencing studies have uncovered more than 100 high-confidence autism risk genes, with hundreds more expected to be found in the next few years. This wealth of information is bringing a long-cherished hope into focus: the possibility of designing personalized treatments for different genetic subtypes of autism.

Researchers have long known that autism is far too heterogeneous to be amenable to one-size-fits-all therapies. Until recently, tailored treatments for genetic subtypes of autism were out of reach, but now a few are starting to emerge. For instance, clinical trials in children and adults are underway for drugs that have been found to ‘rescue’ animal models of fragile X syndrome, an autism-related disorder characterized by intellectual disability, lack of eye contact, and difficulties with speech. Trials are also being planned for a drug that may improve speech in individuals with deletions in the genetic region 16p11.2, a genetic variant that underlies about one in every 200 cases of autism.

For another autism risk gene, SCN2A, researchers are using their understanding of the gene’s molecular mechanisms to try to figure out which existing drugs are most likely to reduce the frequency of the seizures that often accompany this form of autism. “Instead of having to spend six months going through a lot of medicines to see which one helps an individual, we hope to be able to say, ‘Don’t use this drug, it’s likely to make things worse — start with this other drug,’ ” says John Spiro, the deputy scientific director for the Simons Foundation Autism Research Initiative (SFARI).

SCN2A and other genes linked to ion channels in the brain are among the most promising targets for drug therapies. The pharmaceutical prospects are less immediate for other autism risk genes, such as those involved in remodeling chromosomal materials. But a nonpharmaceutical approach is rapidly moving out of the realm of science fiction: repairing or replacing the defective gene itself.

In recent years, researchers have devised numerous gene therapy methods, from viruses that insert functional copies of genes into cells to a technique called CRISPR that directly edits a cell’s genome. “There’s been an explosion of knowledge about vectors for getting genes into the central nervous system,” Spiro says.

“These gene therapy approaches are potentially transformative,” he says. “Instead of saying, ‘Let’s choose from 10 medicines,’ we’ll be able to say, ‘Let’s go in and fix the gene that’s broken.’ ”

The Food and Drug Administration has already approved a method for delivering a replacement gene to individuals with spinal muscular atrophy, a disorder unrelated to autism in which muscles gradually weaken. “We know that the response is durable — that the added genes stay there,” says Wendy Chung, the principal investigator for the Simons Foundation Powering Autism Research (SPARK) initiative, which is in the process of sequencing tens of thousands of children with autism and their families.

Within the next five to 10 years, Chung and Spiro say, researchers are likely to attempt gene replacement treatments for some autism subtypes. “There are legitimate reasons to be excited that some of these will be the real deal,” Spiro says.

Identifying personalized drugs and gene replacement therapies is only part of the work that lies ahead. Researchers will have to figure out how to determine whether a promising treatment truly has a beneficial effect. Autism’s core traits are slippery things to measure, in contrast with those of epilepsy, for example, where it’s possible simply to count seizures. And some biological treatments for autism may need to be combined with behavioral or educational therapies before their benefits become apparent. “We need to get these issues sorted out so that we don’t miss an effective treatment because of measurement issues,” says Paul Wang, the deputy director for Clinical Research Associates at SFARI.

For any given autism mutation, scientists will also have to determine the window of development in which therapies are most likely to be beneficial. For some mutations, treatment in adulthood or adolescence might be too late to have an effect. For others, individuals might be able to benefit at any age. For instance, researchers have shown that mice with nonfunctioning copies of the autism gene SHANK3 can recover from certain deficits even in adulthood if the gene gets activated. “There have been some stunning examples in the literature of relatively late reversals in mice,” Spiro says.

For therapies that only work early in development, newborn screening could be an effective way to identify candidates for treatment, thanks to the growing availability and affordability of genetic testing. Several states are already screening infants for the spinal muscular atrophy mutation. “Potentially, some would even dream about fetal screening and treatment for some types of autism associated with severe disabilities,” Chung says. “That’s not in the 10-year plan, but if you think farther in the future, there are technical ways of doing this.”

While much of this vision remains far off, families are already starting to benefit from the growing understanding of different autism subtypes. SPARK offers families the option of learning what mutation their child has (when it can be identified); they can then potentially join Simons Searchlight, an initiative that helps form supportive family communities for more than 150 different autism-linked mutations.

Understanding a child’s genetic profile can enable SPARK’s researchers to give parents crucial information about their child’s prognosis. “It helps parents enormously, having some sense of where they’re going,” Chung says.

Sometimes this prognosis is good news, reassuring parents that their child will eventually learn to speak, for instance. And even when the prognosis isn’t so great, that information can help parents prepare to meet their child’s needs. For instance, SPARK has been able to alert some parents that their child’s form of autism is commonly associated with epilepsy, and in some cases the parents realized that their child was already having seizures that flew under the radar. “For some children, it’s been quite helpful because the seizures were at the root of some of their behavioral and educational challenges,” Chung says.

Regardless of whether a specific prognosis or treatment is available, learning that a child has an autism-linked mutation can end parents’ painful speculations about whether they inadvertently caused their child’s autism through some misstep. “Understanding the genetics helps answer the fundamental question ‘Why does my kid have autism?’ ” Spiro says. “For some families, that ends a diagnostic odyssey.”