Biotech & Longevity · Wednesday, 26 August 2026

01 · Briefing · what happened

Answering the new virus with the old antibodies, in a week about what people bring to a vaccine

Biotech & Longevity 3 min 16 sources

Unvaccinated adults in Nigeria met Omicron with antibodies raised against the 2020 coronavirus - and when those older antibodies were stripped out, the response to Omicron went with them.

93%

carried an earlier, pre-Omicron infection

in the first Nigerian group; 58% also showed Omicron exposure [1]

4,089

people whose antibodies were read before vaccination

across 8,687 samples and 185 immune targets [2]

4 in 5

rectal cancer patients who kept their rectum

at 12 months, against 3 in 5 on radiotherapy alone [6]

$2.7m

list price of a newly approved gene therapy

a one-time treatment for glycogen storage disease type Ia [8]

At a glance

Unvaccinated adults in Nigeria whose most recent infection was Omicron still made antibodies that recognised the original 2020 strain better than the variant that had just infected them. [1]
In the first group, 93% carried signs of an earlier, pre-Omicron infection and 58% showed signs of Omicron; a second, separate group repeated the pattern. [1]
Stripping out the antibodies that stick to the original strain wiped out the ability to block Omicron in most people, so the new response was built from old parts. [1]
A separate US study read 185 antibody targets across 8,687 blood samples from 4,089 people and picked out who would respond weakly to a Covid vaccine before they had it. [2]
About 5% of healthy people mounted a blunted response, while some people on immune-suppressing treatment responded strongly - the health label alone predicted little. [2][3]
Immune ageing runs on two clocks. Men lose the normal balance between helper and killer T cells earlier and more sharply than women, which the authors say should shape vaccination. [4]
More than 1,000 genetic switches behave differently in male and female immune cells, part of why lupus can affect as many as nine women for every man. [5]
In a European trial of more than 400 patients, four in five people with rectal cancer avoided major surgery for a year. They had chemotherapy with radiotherapy; three in five managed it on radiotherapy alone. [6]
Patients whose cancer was not cleared still had an operation to remove what remained, so the gentler route is a first attempt with a fallback, not a gamble. [6]
The FDA approved the first treatment for glycogen storage disease type Ia, a one-time gene therapy listed at $2.7 million per patient. [8]
It also cleared the first drug for warm autoimmune haemolytic anaemia, where the immune system destroys its own red blood cells. On the higher dose 24% held a lasting rise in haemoglobin, against 8% on placebo. [9]
Massachusetts's industry group warned that the earliest, riskiest startups are being skipped, while licensing deals for China-discovered drugs reached $79 billion last year against $1 billion in 2019. [14]

Forces in play

Old immune memory High→

the response to the 2020 strain still dominates years and several variants later [1]

Reading people first Building↑

antibody fingerprints taken before vaccination separated strong responders from weak ones [2]

One dose for everyone Easing↓

age, sex and infection history each change what the same shot does [4][5]

Cost of the new fixes High↑

$2.7m for one gene therapy, about $1.4m a year for a rare bone drug [8][10]

In play Institute of Human Virology, Nigeria, and Cambridge — ran the unvaccinated groups that isolated the effect Arizona State University's Biodesign Institute — read antibody fingerprints to predict vaccine response The STAR-TREC investigators in Leeds, London and Birmingham — tested keeping the rectum instead of removing it The FDA — cleared two first-in-disease treatments and delayed a third

How it unfolded

Early 2023 blood is taken from two separate groups of unvaccinated Nigerian adults [1]
Wed the FDA clears the first gene therapy for glycogen storage disease type Ia [8]
Thu a US study reports antibody patterns that predict vaccine response before the shot [2]
Mon the rectal cancer trial reports four in five patients avoided major surgery [6]
Tue the imprinting result is written up, and a first drug for warm autoimmune haemolytic anaemia is approved [1][9]

Where this points

Watch whether updated Covid vaccines start being tested against people's real infection histories rather than an average volunteer - that is what the Nigerian result puts in question. [1][2]

Full briefing

Why the old answer wins

Immune memory is built for speed. Cells that already recognise a shape can answer in hours; a cell trained from scratch needs days the body may not have. Because the original coronavirus and Omicron still share most of their outer surface, the old cells do recognise the new virus - imperfectly - and they get there first. [1]

That stays a hunch until you can take the old antibodies away. The Nigerian and Cambridge team did exactly that, removing everything that stuck to the 2020 version of the spike, the knob the virus uses to get into cells. What was left could barely block Omicron in most people. The new answer had been built almost entirely out of old parts. [1]

The setting is what makes the result clean. In Europe and North America nearly everyone carries layers of vaccination and reinfection, so no one can separate what a shot did from what an infection did. Nigeria had low vaccine coverage and wide spread of the virus, which leaves infection alone as the teacher. [1]

What it does not settle

The team measured antibodies, not T cells, which do a different job. And they read each person’s infection history from blood tests, not by sequencing the viruses themselves. So a “pre-Omicron” infection could have been Delta or Beta, not the original strain. Repeated exposure to Omicron softened the bias without removing it. [1]

Doing less, deliberately

The week’s other large result changed no drug at all. In the rectal cancer trial, the chemotherapy and radiotherapy were not made stronger; the order changed. Surgery, which for decades often meant a permanent colostomy bag, was held back as a rescue rather than given by default. [6] That is safe only because a cancer left behind can still be found and operated on later. Prostate cancer is being pulled the same way on thinner evidence. Focal therapy destroys only the lesion visible on a scan, not the whole gland. A UK study found that ten years on, 0.1% of men treated that way had died of prostate cancer. The American Urological Association still calls it experimental. A paper last week found half the men receiving it sat outside the group the guidelines cover. [7]

The bill, and the caveats

The new fixes arrive expensive. Regeneron’s newly cleared antibody drug for a rare disease that turns muscle and tendon into bone lists at roughly $1.4 million a year. [10] The FDA also pushed a Duchenne cell therapy decision to 22 November after the company narrowed what it was asking permission for. [11] Around that, the week ran on caveats. A promising mRNA booster for cancer vaccines has so far only cleared tumours in mice. [12] County-level reporting missed measles transmission crossing the epidemic line inside US schools. [13] One critic called a protein-restriction longevity review a beautiful mouse story sold as human advice. [15] And a map of where new neurons stall in depression shows association, not cause. [16]

02 · Lesson · why it matters

Why the first answer is the hardest one to replace

A system that solved something once keeps answering with that solution - and the better the first answer was, the harder a new one becomes.

How it works

A first infection trains cells to one exact shape
The virus drifts; the new shape still overlaps the old
The trained cells recognise it roughly, and answer first
Answering rewards those cells, so they multiply again
The from-scratch response is never needed, so it never grows

The twist

The body is not failing to learn. It is succeeding at remembering - and the reward for a fast old answer is that a better new one never gets built.

Where you've seen this

Languages

the sounds of a first language decide which sounds you can still hear decades later

Software

a file format that worked in version one gets carried forward long after it stops fitting

Institutions

a body built for the last emergency meets the next one holding the last one's tools

The catch

Imprinting is not simply a loss - the old antibodies do block the new virus, just less well, and repeated exposure wears the bias down.

Full lesson

The blood said the wrong year

Take an adult in Nigeria who was never vaccinated, and whose most recent brush with the coronavirus was Omicron. Draw their blood. Test it against the virus that infected them, and against the version that circulated in 2020. The 2020 version comes off better.

That is odd only if you think memory is a filing cabinet. It is not. It is a set of trained cells, and the training happened once.

Speed decides who answers

When something enters the body, two responses are possible. One is instant: cells that already recognise a similar shape start copying themselves within hours. The other is slow: the body builds a new answer from scratch, and that takes days.

The original coronavirus and Omicron are not the same, but they are not strangers either. Most of their outer surface still matches. So the old cells recognise the new virus - badly, but first.

And because they answer first, they win. They multiply, they get rewarded, and their numbers grow again. The from-scratch response never becomes necessary, so it never gets built. Each infection makes the old template a little more dominant, which makes the next infection more likely to be handled the same way.

This is not the immune system failing. It is the immune system doing exactly what it was built to do, at full strength, on a problem that has quietly changed shape underneath it.

The reference shape was an accident

Now step back from the body to the arrangement around it.

There is nothing special about the version of the coronavirus that reached the world first. It was not the most dangerous version, or the most representative one. It was simply the one that arrived. And because it arrived first, it became the shape that vaccines were designed to, that assays were calibrated to, and that billions of immune systems were trained on.

An accident of timing became the reference for everything after it. That is not a conspiracy and nobody chose it; it is what happens when a system has to lock onto something in order to act at all. But it is worth noticing, because from inside it looks like a fact of nature rather than a decision the calendar made.

The same shape shows up on the other side of this week’s news. For decades, the reliable way to cure a cancer in the rectum was to remove the rectum, often leaving a person with a permanent bag. That was a real answer to a real problem, and it worked. It also became the default, and stayed the default long after gentler routes were worth testing - which is what the trial reported this week finally did.

What gets known depends on where you look

There is a second arrangement here, quieter than the first.

Almost everything the world learned about immune memory during the pandemic came from countries where nearly everyone was both vaccinated and repeatedly infected. In those populations you cannot pull the two apart. The question was not hard to ask; it was impossible to answer with the data anyone was collecting.

Nigeria could answer it because the mix there was different - lots of virus, few vaccines. The finding was not waiting on a cleverer experiment. It was waiting on someone funding a long-running study in a place the research maps mostly left blank.

So the picture of how immunity works was shaped, for years, by where the cohorts happened to be. What counts as known is not just a matter of what is true. It is also a matter of who got studied.

You are already carrying yours

None of this is a story about other people’s bodies.

Whatever the first flu you met was, it set a template you still carry. So did the first cold, the first vaccine, the first infection you were too young to remember. Those encounters are quietly shaping how you will answer a virus that does not exist yet. You have no way to inspect the settings, or to know which of them will help.

The same is true one level up. Every institution you rely on is running on an answer that worked once, formed at a moment nobody now remembers clearly, and is answering today’s problem with it. From the inside, that always feels like competence, because it usually is - right up until the shape changes.

Seeing that does not tell anyone what to do next. It mostly makes the confident version of any answer, including this one, a little harder to hold tightly.

03 · Lab · your turn

Recall or Rebuild

Rehearse why leaning on a memory that still half-works quietly stops a better one from ever being built.

04 · Hope · carry this

The clearest look yet at how immune memory really works came from two groups of adults in Nigeria, in a country the pandemic's research maps largely passed over. Widening who gets studied keeps turning up answers that were never going to be found anywhere else.