An 82% reduction in a disease-linked molecule sounds like a medical breakthrough. But for children with Hunter syndrome, the more consequential question is whether that number represents something they can actually experience: another word learned, another memory retained, another year of independence.
On February 9, the U.S. Food and Drug Administration declined to approve Regenxbio’s experimental gene therapy RGX-121, exposing a difficult question in rare-disease medicine: how much evidence is enough when waiting for certainty may itself cause irreversible harm?
The Gap Between a Biomarker and a Clinical Benefit
Hunter syndrome, or mucopolysaccharidosis type II (MPS II), is an inherited disorder affecting predominantly boys. A deficiency of the enzyme iduronate-2-sulfatase causes complex sugars to accumulate throughout the body, damaging organs and, in severe cases, progressively impairing neurological development.
Existing enzyme-replacement treatments can alleviate some physical symptoms but cannot adequately reach the brain. Regenxbio developed RGX-121, also known as clemidsogene lanparvovec, to address this gap by delivering a functioning gene into the central nervous system.
The results appeared encouraging.
In data presented in September 2025, Regenxbio reported that 13 participants in the pivotal phase of its CAMPSIITE trial experienced a median 82% reduction in cerebrospinal-fluid heparan sulfate D2S6, a biomarker associated with neurological disease. Participants also demonstrated developmental stability or skill acquisition over one year.
Yet the FDA questioned whether these findings established a sufficiently reliable connection between biochemical change and clinical benefit.
According to Regenxbio’s February 9 announcement, the agency's February 7 complete response letter raised three principal concerns: whether enrollment criteria distinguished severe neurological disease from milder forms, whether historical untreated patients provided an appropriate comparison group, and whether the biomarker reasonably predicted meaningful improvements.
The distinction matters because the FDA's accelerated approval pathway permits treatments to reach patients before long-term clinical benefit is conclusively demonstrated, provided their surrogate endpoints are reasonably likely to predict that benefit.
Regenxbio argued that its evidence satisfied this standard. The FDA disagreed.
The decision followed a separate January safety setback. Regulators had placed RGX-121 on clinical hold after a brain tumor was identified in a child who received another Regenxbio gene therapy, RGX-111. The tumor's cause remained unconfirmed, and no comparable tumors had been reported among 32 RGX-121 recipients at that point.
For patient advocates, however, delaying access carried its own consequences.
Terri Klein, president and CEO of the National MPS Society, declared in Regenxbio's announcement that families had waited two decades for new treatments. Joseph Muenzer, a pediatric geneticist at the University of North Carolina, emphasized that neurological deterioration makes additional waiting particularly consequential.
Regenxbio planned to seek another FDA meeting and potentially submit longer-term evidence. But the regulator's suggested alternatives—including additional participants, new studies, or untreated controls—posed formidable challenges for an exceptionally small patient population.
The dispute remained unresolved: stronger evidence could protect patients from an ineffective, difficult-to-reverse intervention, while obtaining that evidence could consume the very developmental years treatment was intended to preserve.
A Childhood Cannot Be Reduced to a Biomarker
Underlying the regulatory dispute is another question: who defines meaningful improvement when the children most vulnerable to neurological decline may be least able to articulate their own priorities?
In a video interview featured on Takeda's Hunter syndrome awareness website, Kyle, filmed at 16, discussed ambitions that extended beyond medical treatment: pursuing a career, traveling, and visiting Ireland.
Kyle has the attenuated form of Hunter syndrome, making his experience clinically distinct from that of children with severe neurological involvement. That distinction is essential. It also illustrates why selecting comparable patients, one of the FDA's concerns, is so consequential.
Another Takeda-published interview recounts how Zayd, a young adult with Hunter syndrome pursuing doctoral studies, found belonging through music during adolescence.
“I got good at the trombone, and people were starting to know me for that instead of my condition,” he recalled.
Neither patient was commenting on RGX-121. Their experiences nevertheless reveal dimensions of treatment benefit that laboratory measurements struggle to capture: participation, relationships, identity, and the freedom to imagine one's future.
For a generation approaching adulthood alongside increasingly predictive medical technologies, Regenxbio's setback exposes a troubling asymmetry. Scientific predictions can guide decisions made long before young patients are capable of consenting to their consequences. Yet the uncertainty surrounding those predictions does not disappear when regulators authorize treatment—or when they refuse it.
Approving too early transfers uncertainty into children's bodies. Demanding more evidence transfers part of that uncertainty into their remaining time.
Neither outcome is ethically neutral.
The FDA's dilemma therefore extends beyond deciding whether an 82% biomarker reduction constitutes sufficient evidence. It concerns whose risks become acceptable, whose observations count as knowledge, and who gets to decide what a better future means for a child.
In rare pediatric diseases, time is simultaneously the evidence medicine needs and the resource patients cannot recover.
Explore more exclusive insights at nextfin.ai.
