Cancer survives by hiding — and this week's news is about making it visible again

Biotech & Longevity 5 min 40 sources

At the big annual cancer meeting, the theme was the same trick from many angles: tumours stay alive by hiding from the immune system, so the new treatments strip the disguise or train the body to see through it. One study found the hiding itself can be a weakness. Plus a longevity startup raises $435 million before its first human trial, a masterclass in reading past a trial headline, and a lawsuit about why consent exists.

Key takeaways

At oncology's biggest meeting, the unifying theme was making cancer visible to the immune system again — an Oxford "invisibility cloak" drug shrank tumours by 30%, an injected drug made tumours vanish in 15 last-line patients, and a Baylor study found cancer's favourite hiding trick can itself become a weakness (all early-stage).
A longevity startup raised $435 million to rejuvenate old cells before running a single human trial — a reminder that in this field the money routinely sprints far ahead of the evidence.
The beat's core skill is reading past a headline: a drug with great efficacy can be sunk by a safety signal (Abivax), and a failed trial can be re-pitched on a cherry-picked subgroup (AstraZeneca) — the word after "but" is where the real story hides.

Cancer’s oldest survival trick isn’t growing fast. It’s hiding. A tumour that the immune system can’t see is a tumour the body won’t attack. This week, at oncology’s biggest annual meeting, most of the headline results were variations on one idea: make the cancer visible again.

Teaching the body to see the tumour

Start with the disguise. Healthy immune cells patrol for anything foreign, but cancer cells learn to switch off the signals that mark them as targets — an “invisibility cloak.” Researchers in Oxford reported an experimental tablet, GRWD5769, designed to strip that cloak off. In an early trial across the UK, France, Spain and Australia, 83 patients with six of the most common cancers took it alongside an existing immunotherapy; tumours shrank by at least 30% [7]. Early-stage, small, and paired with another drug — but a clean demonstration of the principle.

A second approach trains the attack directly. A “triple-action” injected drug called amivantamab was given, in an 11-country trial, to patients whose cancer had spread or returned and stopped responding to both chemotherapy and immunotherapy [18]. More than a third saw tumours shrink; in 15 patients, the tumours disappeared entirely [18]. “Unprecedentedly strong responses,” said Kevin Harrington of the Institute of Cancer Research — in a group who had nearly run out of options [18]. The caveat rides along: these are last-line patients in a trial, not a routine cure.

A third result is quieter but striking. In a bowel-cancer trial called NEOPRISM-CRC, patients with a particular genetic profile — about 10–15% of cases — were given the immunotherapy pembrolizumab before surgery [13]. Many had their tumours clear so thoroughly that chemotherapy became unnecessary. Chris Burston, 73, had three doses over nine weeks; three years on, he remains cancer-free [13]. It only applies to that genetic subgroup — but for them it reorders the whole treatment.

When the hiding place is the weakness

The most interesting finding flips the whole picture. Scientists at Baylor College of Medicine and the University of Michigan found that when cancer cells shut down a key “show-yourself” molecule called MHC-I — the very trick they use to dodge the immune system’s “killer” T-cells — they become more vulnerable to a different squad, the “helper” T-cells [28]. The disguise that hides the tumour from one hunter exposes it to another. The team says it overturns a belief that guided immunology for decades [28]. It’s lab-stage work, not a treatment yet. But it hints that cancer’s favourite escape may come with a built-in trapdoor.

Longevity: $435 million, zero human results

Away from cancer, the longevity industry had a big-money week. NewLimit, a South San Francisco startup trying to “rejuvenate” old cells — coaxing aged cells to behave young again — raised $435 million, and it hasn’t yet run its first human trial [5][39]. That’s an enormous bet placed entirely on laboratory and animal promise.

Worth setting against a sober note from the journal Nature, reviewing a new book pointedly titled Morbid: Debunking Modern Longevity Science [1]. The honest scientific position on how long humans can live is: we don’t know [1]. There’s even real argument that the past century’s gains in life expectancy came mostly from stopping people dying young — better nutrition, medicine, sanitation — rather than from anyone pushing the natural ceiling higher [1]. The money is sprinting; the evidence is still tying its shoes. Both things are true at once, and the gap between them is exactly where hype lives.

How to read a trial result

This is the beat’s most useful skill, and two stories this week teach it. A headline number is never the whole story.

The French firm Abivax reported strong efficacy from a late-stage trial of its drug for ulcerative colitis, a painful inflammatory bowel disease. Good news — except reports of cancer cases among trial participants overshadowed it, and the company’s stock fell hard [3]. Efficacy tells you whether a drug works; it says nothing about what else it does. A drug is the whole package — benefit and harm — and the harm can sink it no matter how well it treats the target.

AstraZeneca offered the other lesson. One of its drugs missed the main goal of its phase 3 trial — the big, final test a drug must pass to win approval [4]. Rather than shelve it, the company highlighted data from a subgroup of patients who seemed to do better, using that to pitch a narrower approval for a rare disease [4]. Sometimes a subgroup signal is real. But slicing a failed trial until some slice looks good is a known way to find a “win” that’s really just chance — which is why regulators treat after-the-fact subgroups with suspicion. When a trial “fails but…”, the word after “but” deserves a hard look.

Why the caveats exist

End on the reason all of this care is non-negotiable. The families of two Black infants have sued the US government, saying the babies were enrolled in a 1960s vaccine trial — for the respiratory virus RSV — without their families’ knowledge or consent, at a Washington, D.C., clinic between 1965 and 1966 [23]. Both died about a year later, at 14 and 16 months old [23]. The families say they only learned of the trial recently [23].

The trial stages, the consent forms, the safety monitoring, the insistence on caveats — none of it is bureaucratic throat-clearing. It exists because medicine has a real history of treating people, disproportionately the poor and the Black, as material rather than patients. Every “we don’t yet know” and “only in this subgroup” in the stories above is a small descendant of lessons learned the hardest way. The excitement is earned. So is the caution.

02 · Lesson · why it matters

The cost of the cloak

Cancer's best trick is to go invisible — switch off the tag that lets the immune system's killers find it. This week, scientists found something unsettling and useful: that very disguise makes it visible to a different set of hunters. The hiding place was also a trap. That pattern is everywhere.

The trick that backfires

Your immune system’s “killer” cells hunt by looking for a tag on the surface of cells — a kind of “show-yourself” badge. A cancer cell that drops the badge becomes invisible to those killers, and that’s exactly what aggressive tumours learn to do. For decades, that was understood as cancer winning: no badge, no target, no attack.

Then researchers at Baylor and Michigan looked closer. A second squad of immune cells — “helper” cells — doesn’t hunt by that badge at all. It hunts differently. And to that squad, a cell that has dropped its badge stands out. The disguise that hid the cancer from one hunter exposed it to another. The escape route had a trapdoor built into it.

Every disguise is built against one watcher

Here’s the general truth hiding in that biology. A disguise is never universal. It’s always built against a specific watcher, with a specific way of looking. It works against that watcher — and only that one.

Camouflage that makes a moth vanish against bark makes it obvious against a white wall. A code that hides what you’re saying still reveals that you’re saying something, and to whom, and how often. The very act of hiding is itself a kind of signal to anyone watching for hiding. There is no cloak that works against every set of eyes, because different eyes look for different things. Defend against one, and you’ve usually turned your back on another.

What people hide, and what it costs them

Take it out of biology, because people do this constantly.

The person who hides their struggle to look strong is camouflaging against one watcher — judgment. And it often works: no one judges what they can’t see. But the same disguise hides them from the people who would have helped. They’ve defended against being thought weak and exposed themselves to being alone with it. The cloak worked exactly as designed; it just had a cost they weren’t counting.

It scales up. A company that buries a weakness so rivals won’t see it often buries it from itself, until the problem it hid grows in the dark. A system locked down so tight no attacker can get in can become so tight no user can either. A life arranged to avoid one kind of risk — never failing, never being rejected, never looking foolish — quietly takes on another: never trying, never connecting, never growing. The defense was real. So was the bill, paid to a watcher you weren’t looking at.

What to carry out of today

None of this means never hide. Sometimes a disguise is exactly right — the cancer researchers, after all, want to use the trapdoor, to catch the tumour through the very move it made to escape. Hiding is a tool, and tools have their place.

The skill is just to know the cost of the cloak, because there always is one. When you reach for a defense — a wall, a secret, a way of disappearing from something you fear — ask the two questions the immune system answers by accident: which watcher is this built against, and which different watcher might it be exposing me to? “Safe from this” is not the same as “safe.” The disguise that saves you from one set of eyes is, almost always, waving at another. Knowing that doesn’t stop you hiding. It stops you being surprised by the bill.

03 · Lab · your turn

The Two-Hunter Problem

Pick a tumour cell's disguise and watch hiding from one immune squad expose it to another — feeling that no disguise is universal, and every defense built against one watcher reveals you to a different one.