FDA Should Practice the Science It Preaches

The FDA has plenty of advice for pharmaceutical companies. It publishes guidance documents, establishes scientific expectations, and expects drug developers to explain themselves when they take a different approach. Companies spend years and millions of dollars designing development programs around those expectations. But who holds FDA to the same standard?

The question takes on particular urgency when the patients are children. Pediatric drug development isn't a matter of adjusting an adult dose for body weight. Developing organs respond differently to medicines, and some adverse effects may not become apparent until years after treatment begins. A reassuring safety profile today doesn't necessarily tell us what happens five or ten years down the road.

FDA knows this. In 2021, it adopted the International Council for Harmonisation's S11 guidance on nonclinical safety testing for pediatric pharmaceuticals. The guidance addresses a problem that has bedeviled pediatric drug development for decades: how to identify developmental risks before exposing children to medicines that may interfere with biological processes we are still struggling to understand.

S11 calls for a weight-of-evidence assessment that considers the age of the youngest patients, duration of treatment, biological pathways involved, and adequacy of existing safety information. Additional nonclinical studies may be warranted when important questions remain. When existing evidence is sufficient, unnecessary testing should be avoided.

FDA expects industry to take this framework seriously. The agency should do likewise.

Consider medicines that interfere with pathways involved in normal brain development. Synaptic maturation and myelination aren't processes that can be adequately evaluated through routine blood tests or a few months of clinical observation. Some effects may be subtle, cumulative, or recognizable only after a child reaches a later developmental milestone.

FDA's S11 guidance anticipates this problem. It discusses circumstances in which targeted juvenile animal studies, including neurological endpoints, may be appropriate even when adult safety information is available. The question isn't whether FDA should demand another animal study. It's whether the evidence already available adequately addresses a scientifically plausible developmental risk.

A recent citizen petition involving infigratinib, an investigational treatment for achondroplasia, provides a useful example. The drug inhibits FGFR3, a receptor involved in the abnormal bone growth associated with the condition. But it also inhibits FGFR1 and FGFR2, which participate in normal neurological development. Evidence indicates that infigratinib can cross the blood-brain barrier.

The petition argues that publicly available studies haven't adequately addressed the possible neurological consequences of prolonged treatment in young children. It requests additional scientific scrutiny and, if necessary, targeted non-clinical testing. The petition was filed by Ascendis Pharma, a commercial competitor. That fact deserves disclosure, but it doesn't relieve FDA of its obligation to address legitimate scientific questions on their merits.

None of this establishes that infigratinib causes neurological harm. Nor does a citizen petition prove that a development program is deficient. FDA may have additional evidence that resolves these concerns. The question is whether the agency has applied its own scientific framework and can explain why the available evidence is sufficient.

This is not just a pediatric drug development problem. Modern medicines increasingly target molecular pathways that perform multiple biological functions. Blocking an abnormal signal in one tissue may also affect normal activity elsewhere. Understanding those consequences is basic pharmacology. It shouldn't require a regulatory controversy to get FDA's attention.

Better biomarkers, targeted nonclinical studies, and more sophisticated analytical methods can help resolve these uncertainties without automatically adding years to development programs. FDA has been talking enthusiastically about fit-for-purpose evidence and modernizing regulatory science. Fine. But calling something modern doesn't make it scientifically sound.

FDA guidance shapes investment decisions, clinical trial designs, and regulatory submissions. Sponsors are expected to explain departures from agency recommendations, often in painstaking detail. FDA should be prepared to do the same when its decisions appear inconsistent with its published scientific framework. Otherwise, guidance becomes less a predictable expression of regulatory expectations than a document whose importance depends on who is being asked to follow it.

Guidance documents aren't binding regulations. FDA must retain the flexibility to exercise scientific judgment, particularly when new evidence or unusual clinical circumstances warrant a different approach. Nobody benefits from an agency that mistakes regulatory consistency for bureaucratic rigidity.

But there is a substantial difference between exercising judgment and improvising a justification after the fact.

The agency should be more innovative, not less. It should embrace smarter trial designs, better biomarkers, sophisticated statistical methods, and nonclinical approaches that answer meaningful questions without imposing unnecessary burdens. Regulatory rigor shouldn't be measured by the number of studies FDA demands or how long it takes to reach a decision. The real measure is whether the agency has asked the right questions and has enough reliable evidence to answer them.

This matters enormously in pediatrics. The youngest patients may face the longest exposure, and developmental consequences can be difficult to recognize until years later. Waiting for a postmarketing safety signal may be reasonable for some uncertainties. For others, it may mean discovering a problem only after the opportunity to prevent it has passed. S11 was written to help regulators make those distinctions. FDA should be prepared to explain how it has done so.

Patients shouldn't have to choose between timely access to promising medicines and confidence that FDA has done its scientific homework. Nor should drug developers have to guess when the agency intends to follow its own guidance.

Scientific credibility isn't established by issuing guidance documents. It's earned by applying them consistently—and explaining, with evidence, when a different approach is warranted. FDA wrote the rules of the road. It shouldn't be the first to disregard them.

Peter J. Pitts is a former FDA Associate Commissioner and President of the Center for Medicine in the Public Interest.