A first treatment for Menkes disease, tested against children who received none

On January 12 the Food and Drug Administration approved Zycubo, a copper histidinate injection, as the first treatment for Menkes disease in pediatric patients [s1].

The approval is worth reading twice: once for what it does for a small group of children, and once for how the agency decided it worked. The second reading is the one with implications beyond this disease.

The disease

Menkes disease is a neurodegenerative disorder caused by a genetic defect that impairs a child's ability to absorb copper [s1]. It is characterised by seizures, failure to gain weight and grow, developmental delays, and intellectual disability, and it produces abnormalities of the vascular system, bladder, bowel, bones, muscles, and nervous system [s1].

Children with classical Menkes — about 90 percent of those affected — begin to develop symptoms in infancy and typically do not live past three years [s1]. The disease affects approximately one in every 100,000 to 250,000 live births worldwide, and is more common in boys [s1].

Those two facts together — a fatal course in early childhood, and a few hundred births a year worldwide at most — define the problem the FDA had to solve. There has never been an approved treatment, and the population is too small and too ill for a conventional randomised trial in which half the infants receive placebo.

How the drug works

Zycubo is a copper replacement therapy given by subcutaneous injection [s1]. It delivers copper in a form that bypasses the genetic defect in intestinal absorption, allowing the body to better use the mineral [s1].

That is a mechanistically clean idea: the defect prevents copper crossing from gut into circulation, so the drug delivers copper past the gut. Whether the idea works is an empirical question, and the answer depends on when it is given.

The evidence, and its structure

FDA evaluated Zycubo in two open-label, single-arm clinical trials in pediatric patients treated for up to three years [s1]. Overall survival was assessed by comparing treated patients with untreated patients from contemporaneous external control groups; the analysis included 66 treated patients and 17 untreated patients, most of them from the United States [s1].

The headline result: children who began treatment within four weeks of birth had a 78 percent reduction in the risk of death compared with untreated patients [s1]. Nearly half of early-treated patients survived beyond six years, and some survived more than 12 years [s1]. No patients in the untreated control group survived beyond six years [s1].

Children who started treatment later than four weeks after birth also experienced a substantial survival benefit, according to the agency [s1].

What "external control" means, and what it costs

This is the part that matters methodologically. In an external-control design, the comparison group is not randomised alongside the treated group. It is assembled from patients who were not treated — here, contemporaneous ones — and their outcomes are set against the treated group's.

The design is not a substitute for randomisation. It cannot rule out that treated and untreated children differed systematically in ways that affect survival: which families reached specialist care, how early the diagnosis was made, what supportive care was available, whether the underlying genetic variant was more or less severe. Some of these can be adjusted for statistically; none can be eliminated the way randomisation eliminates them.

What makes the design defensible here is the natural history. In a disease where children with the classical form typically do not live past three years [s1], and where no untreated control patient survived beyond six [s1], survival past six and in some cases past twelve is a large effect against a well-characterised baseline. Large effects against near-uniform natural history are exactly the circumstance in which external controls carry the most weight — and small effects against variable natural history are exactly where they carry the least.

FDA's own framing acknowledges the trade-off. "The company demonstrated a large improvement in overall survival compared with untreated patients, using an innovative trial design that addressed the challenges of studying an ultra-rare disease," said Tracy Beth Hoeg, Acting Director of the Center for Drug Evaluation and Research [s1].

Safety

The most common side effects reported with Zycubo included infections, respiratory problems, seizures, vomiting, fever, anaemia, and injection site reactions [s1].

Because copper can accumulate in the body, patients receiving the drug should be closely monitored for potential toxicity [s1]. That is a real constraint on a therapy whose mechanism is delivering more of a mineral that is toxic in excess.

Regulatory path

The application received Priority Review, Fast Track Designation, Breakthrough Therapy Designation, and Orphan Drug Designation [s1]. The FDA approved Zycubo for Sentynl Therapeutics [s1].

Christine Nguyen, Deputy Director of the Office of Rare Diseases, Pediatrics, Urologic and Reproductive Medicine in the Center for Drug Evaluation and Research, framed the result in survival terms: children with the disease "will have an FDA-approved treatment option and the potential to live longer" [s1].

What remains open

The trials followed patients for up to three years of treatment [s1]. Survival beyond that is reported from the treated cohort but the trial duration bounds what was studied prospectively.

The approval also does not establish what happens to neurological function. The reported endpoint is overall survival [s1]. Menkes disease produces developmental delay and intellectual disability [s1]; whether early copper replacement changes that trajectory, and by how much, is not answered by a survival analysis.

And the four-week window is doing a great deal of work. A treatment whose largest measured benefit depends on starting within a month of birth requires diagnosis within a month of birth — which, for a disease affecting perhaps one in 100,000 to 250,000 births [s1], is a newborn screening question as much as a pharmacological one.

This article is informational and is not medical advice.