A gene therapy halted over spots no child can feel, in a week medicine spent looking harder

Five spines, and no way to read them

Five boys have small masses on their spines. They feel nothing. Their radiologists think the masses are probably harmless. Their investigators wrote down non-serious. And the FDA has stopped the trial anyway.

Nobody is being unreasonable. The scans are perfectly clear. What is missing is not resolution but a second picture: the same scan, in children with the same rare disease, who never got the therapy. That picture does not exist. Children with Hunter syndrome are not routinely given spine MRIs, because until now there was no reason to point a scanner there.

So the finding sits there, sharp and unreadable. It could be the therapy. It could be the disease, quietly doing this all along in every child who has it, seen for the first time because someone finally looked.

A number is half of a sentence

We talk about measurement as if a reading were a fact. It is not. A reading is one half of a comparison, and the half that gives it meaning is the other one.

Five out of a group means nothing until you know what five out of an untouched group looks like. Your blood pressure reading means one thing if your usual is 110, another if your usual is 140. If nobody took it while you were well, both stories fit.

This is why the second half is so easy to lose. It is not exciting. It is a measurement of nothing happening, taken from people nothing is happening to, filed away against a day that may never come.

The moment passes, and it does not come back

Here is the part that catches everyone. You cannot go and get it later.

Once every child in the trial has been dosed, there is no untreated group left to scan. The comparison had to be gathered while nobody yet wanted it - while the question that needs it had not been asked. By the time the question arrives, the window is shut.

The trail through this year makes the point almost too neatly. In February the FDA told Regenxbio to enrol more patients and add an untreated comparison group. In June it dropped both demands and agreed to review the studies the company already had. That group was requested to prove the therapy works, not to explain a scan. But it is close kin to the very thing whose absence now leaves everyone guessing.

Everyone is walking into this at once

The same shape is spreading fast, and not only in gene therapy. A blood test for Alzheimer’s-related brain changes was cleared this week for people with mild memory decline. Newborn programmes are sequencing for more than seven hundred conditions where the old heel-prick test looked for sixty-six. Hospitals are reading whole genomes in three days.

Every one of these finds real things in people who feel entirely well. And each new finding lands in the same gap: what does this look like in everyone who was never tested? The instruments are running ahead of the comparisons, and the comparisons cannot be manufactured after the fact.

You are inside this too. Most of what a doctor knows about your body was measured on a day you already felt bad. The record of your ordinary is thin, because collecting it costs money and time and helps nobody today.

What the shape of it costs

Baselines get cut for a reason that is not stupidity. Most are never used. They are dull to collect, expensive to keep, and their value shows up only in a future nobody can name. The one that turns out to matter is invisible in advance - which is exactly why it is first out of the budget.

The families waiting on this therapy are paying for a decision made months ago by people acting sensibly. Nobody chose to leave them guessing. The saving was real and the cost arrived later, on someone else’s ledger, in a form nobody could have priced.

That is worth holding lightly next time a number arrives and seems to say something. Ask what it is standing next to. Often the honest answer is nothing, and the honest thing to do is say so - which is, in the end, all Regenxbio actually did.