President Donald Trump has yet to name a new leader for the U.S. Food and Drug Administration (FDA) — the agency that vets and approves prescription drugs. But it’s clear that he aims to speed up the drug approval process by eliminating some of the current regulations.

In a January meeting with pharmaceutical executives, Trump said he would cut FDA regulations by 75 to 80 percent. He reiterated his plan in an address to Congress last week, saying a move to “slash the restraints” would lead to more medical breakthroughs.

Those remarks have sparked concern among some autism researchers: Many parents of children with autism are known to turn to unproven — often dangerous — ‘remedies,’ in part because of the dearth of drugs approved for the condition’s core features.

“We have a long history [in autism] of flash-in-the-pan treatments that turned out not to do anything useful,” says Jeremy Veenstra-VanderWeele, associate professor of psychiatry at Columbia University.

Other researchers note that changes to how the FDA evaluates clinical trials for a drug might ease drugs’ path to approval.

At least two of the reported front-runners for FDA chief are likely to further Trump’s vision for the agency. Joseph Gulfo, a medical entrepreneur based in New Jersey, has strong ties to the pharmaceutical industry. Having someone closely aligned with the industry leading the FDA would be a problem, says Randi Hagerman, professor of pediatrics at the University of California, Davis. “Drug companies are about making money, and that’s like putting a wolf in the chicken coop.”

Another candidate, Jim O’Neill, managing director of the investment firm Mithril Capital Management, has said the FDA should approve drugs as soon as they are deemed safe and try to determine whether the drugs work after they are on the market.

But such a policy would flood the market with treatments that have no benefit, caution experts. “Resources, efforts, hopes, emotions [will] be diverted and not channeled in the right way,” says Antonio Hardan, professor of psychiatry and behavioral sciences at Stanford University in California.

[Trump nominee Scott Gottlieb, a former doctor, has stated that the FDA should tolerate more uncertainty in deciding whether to approve drugs. Gottlieb also has drug-company connections and has worked in various positions at the FDA, including as deputy commissioner for medical and scientific affairs.]

Room for change:

Relaxing some FDA guidelines might help speed the search for autism treatments. For example, the FDA requires that researchers choose an ‘outcome measure’ — a marker of improvement that a drug must meet in a clinical trial — before the trial begins. It also requires that the drug show statistically significant improvements across the entire set of trial participants.

These requirements are restrictive, and sometimes prevent good drugs from reaching the public, experts say.

“If a new medication helps 30 percent of individuals with autism, that is great. But it doesn’t meet the FDA standards,” says Hagerman. “I am in favor or loosening up some of the rules for efficacy.”

For instance, mouse studies suggested that a drug called arbaclofen would ease the features of fragile X syndrome, a condition related to autism. In 2012, a trial of 150 people showed that the drug helps some people with fragile X, based on a clinician’s assessment of their features. But the study’s primary outcome measure was social behavior, which the treatment did not alter. The FDA deemed the trial a failure, and Seaside Therapeutics, the drug’s manufacturer, subsequently went out of business.

If the FDA had allowed for a different trial design, the first-round attempts might have shown a benefit, says Elizabeth Berry-Kravis, who co-led the trial at Rush University Medical Center in Chicago. “A better understanding of what we’re all trying to do in the trial might shorten things,” she says.

Allowing the use of varied outcome measures might help autism drugs get to the market sooner.

“If the new head of the FDA had a different vision for how to potentially measure outcomes, then it would certainly have an impact,” says Alexander Kolevzon, clinical director of the Seaver Autism Center for Research and Treatment in New York City.

Checks and balances:

There are no reliable biological markers of autism, so trials of autism drugs typically rely on subjective reports of improvement. As a result, these trials are particularly susceptible to large placebo effects, and parents may start using a drug before it has been proven to be effective.

For example, the drug secretin rose to prominence in 1998. In a study of three children with autism, researchers reported that the children’s behavior seemed to improve after they were given the drug for a gastrointestinal procedure¹. The FDA had already approved secretin for a test of pancreatic function, and so it was deemed safe, but it had not yet been tested for autism. Still, based on the case study and anecdotal reports in the media, parents began pushing pediatricians to prescribe the drug or purchased it on the black market².

It took 16 studies showing that secretin had no beneficial effect on people with autism to convince the autism community to move on³. “If we had stopped at safety, we’d probably still have tens of thousands of kids in this country, maybe more, taking something that the [later] studies carefully studying efficacy showed did nothing,” Veenstra-VanderWeele says.

The money and effort spent on these studies could have supported more fruitful research. And even drugs that meet safety standards can have side effects. So it is important to know that the potential benefits of a treatment outweigh the risks, he says.

“You really need to show there’s a benefit in order to take that sort of risk,” he says.