The FDA delays a rare-epilepsy drug three months — and analysts blame last year's staff cuts

Biotech & Longevity 3 min 80 sources

A finished drug application sits in a longer queue, as the agency that reviews it runs shorter on people. Plus a gene therapy chief steps down, and a busy week of new results piling up behind them.

Key takeaways

The FDA pushed back its decision on a rare childhood-epilepsy drug by three months, and analysts blame last year's cuts to agency staff — a reminder that approval is a job that needs people, not just science.
In the same week, the FDA's acting gene-therapy review chief stepped down, thinning the pool of specialists exactly as complex new treatments pile up behind them.
New trial wins and deals from Otsuka, AstraZeneca and others kept the pipeline full, while a separate FDA program raced to speed up drug manufacturing — capacity added in one queue, quietly draining from another.

The US drug regulator has delayed its decision on a rare childhood-epilepsy drug by three months — and Wall Street analysts think they know why. The science is done. The reviewers to read it are thinner on the ground.

A three-month wait, and who’s counting

On June 29, the FDA — the US Food and Drug Administration, the agency that decides which drugs can be sold — pushed back its verdict on relutrigine, a Praxis Precision Medicines drug for two severe forms of infant epilepsy [22]. The drug had been accepted for priority review back in March, the fast lane the agency reserves for serious conditions [22].

The delay is three months. TD Cowen analysts said they “strongly believe FDA bandwidth constraints from last year’s agency downsizing likely contributed to the delay” [22]. That is an analyst’s read, not the FDA’s stated reason — but it names something rarely said out loud: an approval is not just a scientific judgment. It is a task that has to sit on someone’s desk, and someone has to have the hours to do it.

Relutrigine treats SCN2A and SCN8A developmental and epileptic encephalopathies — rare, severe seizure disorders that begin in infancy [22]. For the families waiting, three months is not an abstraction.

The people who read the files

The same week, the FDA’s acting head of the office that reviews cell and gene therapies stepped down. Vijay Kumar told colleagues by email that he and agency leadership “mutually decided not to renew my detail” [56].

His exit lands inside what STAT called a “rocky year of turnover” at the Center for Biologics Evaluation and Research, the FDA division that houses gene-therapy review [56]. Kumar had been promoted to the role only about a year earlier, after his predecessor Nicole Verdun was pushed out [56]. Cell and gene therapies — one-time treatments that fix a disease at the level of a patient’s own cells or genes — are among the most complex things the agency reviews. They need specialists who understand them. Those specialists are the exact people leaving.

None of this means an approval has been refused. It means the machinery that grants approvals is running with fewer hands, at a moment when the workload is climbing.

The queue behind them

Because the pipeline is not slowing down. In the same week’s news:

Otsuka reported a positive phase 3b trial for its ADHD drug centanafadine, weeks before an FDA approval decision it is now waiting on [18]. A phase 3 trial is the large, final human test before approval — hundreds of patients, measured against a dummy pill or the current standard.

AstraZeneca unveiled phase 3 data for a rare bone-disease drug it forecasts could earn $5 billion a year, a successor to its existing treatment Strensiq [28]. Ionis licensed ex-US rights to a near-approval rare-disease drug to Recordati for $30 million upfront [4].

Each of these is a file that eventually lands in the same building, in front of the same shrinking pool of reviewers. When capacity falls and submissions rise, the gap does not vanish. It turns into waiting.

The other side of the ledger

The agency is not standing still. On June 30 it named seven companies for its new PreCheck pilot, a program to speed up building drug-manufacturing plants inside the US [43]. Over 80 firms had applied [43]. The stated goal is a more predictable path and a sturdier domestic supply chain [43].

So the FDA is trying to move faster on manufacturing while, by the analysts’ account, moving slower on review. Both can be true at once. Speed in one queue does not add hours to another.

The science kept arriving anyway

Underneath the regulatory churn, the lab work continued. Epicrispr said it produced the first clinical sign of increased lean-muscle volume in a form of muscular dystrophy, using gene silencing — a technique that quiets a harmful gene without cutting the DNA [53]. It is early, company-reported data from a small patient group, not an approval — the kind of result that, if it holds, will one day need a reviewer to read it too [53].

02 · Lesson · why it matters

The step nobody counts: someone still has to read it

Between a drug that works and a drug you can have sits a person with a stack of files and only so many hours — and that person is easy to forget until the stack stops moving.

Two clocks, not one

There are two clocks running on any new medicine.

The first is the science clock. It runs for years — the lab work, the animal tests, the three rounds of human trials, each bigger than the last. When it stops, the headline says the drug “works.” That is the clock everyone watches.

The second clock starts where the first one ends. The finished application lands at the FDA, and now the question is not does it work but has anyone read it yet. This clock is quiet. It has no press release. And this week, for a rare-epilepsy drug from a company called Praxis, it added three months.

We treat the first clock as the whole story. The drug is done, so surely the rest is a formality. But the rest is its own clock, and it does not run on the science. It runs on people.

The input that hides in plain sight

Think about what an approval actually is. Not a switch someone flips. A human being — a specialist who understands how a one-time gene therapy behaves in the body — sitting down with a document the size of a phone book. Reading it, carefully, against everything they know.

That reading is the product. And like any product, it has a capacity. There are only so many qualified readers, and each has only so many hours.

Here is the trap. When you tally up what a drug needs, you count the money, the trial, the science. You forget to count the reading, because the reader is invisible — a name you never learn, in a building you never see. So when that capacity shrinks, nothing looks broken. The lab still works. The trials still finish. The drug still “works.” It just doesn’t arrive.

This week you could watch the input drain in real time. The agency’s acting head of gene-therapy review stepped down, in what one outlet called a rocky year of turnover. The specialists who read the hardest files are the exact people leaving. The capacity to say yes is falling — silently, because nobody was counting it.

A full queue meets a shrinking desk

Now hold two facts next to each other.

On one side, the pipeline is fuller than ever. In the same few days: a positive final-stage ADHD trial waiting on a decision, a rare bone-disease drug forecast to earn billions, a rare-disease licensing deal, an early muscle-growth result. Every one of them is a file that will land on the same shrinking desk.

On the other side, that desk has fewer people at it than a year ago.

When the amount of work climbs and the number of hands falls, the gap does not disappear. It has to go somewhere. It goes into the one place that absorbs mismatch without anyone deciding it should: the wait. Three months here. Longer there. Not a refusal — a queue.

And notice: the same agency, the same week, launched a program to speed up building drug factories, with over eighty companies lined up to apply. It can pour effort into making one queue faster while another quietly lengthens. Speed is not a mood the whole system is in. It is hours, spent in one place and therefore not another.

Where you’re standing in this

It is tempting to read all this as someone else’s logistics — regulators, analysts, companies you’ll never deal with. But the queue ends at a person, and eventually that person is you, or someone you love, waiting for a treatment that exists but hasn’t arrived.

The families waiting on the epilepsy drug are not waiting on the science. It’s finished. They’re waiting on the second clock — the one made of somebody’s working hours. You are downstream of that clock every time you need something a system has to process: a claim, a permit, a scan result, a case. The thing that decides how long you wait is often not the difficulty of your case. It’s how many people are on the other side, and how many others are in line ahead of you.

The choice that poses as a fact

The last move is to see that the capacity itself is a choice, not a law of nature. How many reviewers the agency has, how fast files move — those are set by decisions about budgets and staffing made far upstream, long before your file arrives. By the time you’re in the queue, the length of the queue already feels fixed, like weather. It isn’t. Someone chose it, in a room you weren’t in, about a number you never saw.

That choice can serve more than one thing at once. Fewer reviewers can mean a leaner agency and a longer wait — both real, at the same time. Naming the trade isn’t accusing anyone; it’s seeing the whole shape.

So the humility is this. When something you need is slow, the reflex is to ask what’s wrong with your case. Often the truer question is quieter. Who’s reading, how many of them are there, and who decided that — none of which you can see from the line. The step nobody counts is usually the one you’re waiting on.

03 · Lab · your turn

The Desk

Rehearse how a drug's wait is set not by its difficulty but by arrivals against reviewer capacity — and feel the gap turn into the line.

04 · Hope · carry this

A queue is one of the few problems that a decision can fix overnight — add hands to the desk and the line moves. The science this week didn't slow down, and neither did the people working to close the gap between a drug that works and a child who gets it.