Biotech & Longevity · Thursday, 27 August 2026

01 Briefing what happened

The first drug to lengthen life in pancreatic cancer is approved, at $39,800 a month

Biotech & Longevity 24 sources

The FDA cleared Rasonque six and a half months early for a cancer that had almost nothing to offer. Two thousand people were already taking it free.

13.2 months

median survival on the drug

against 6.7 months on chemotherapy, in 500 patients [1]

$39,800

list price per month

about $478,000 a year, mostly carried by insurers [2]

2,000+

already taking it, free

since May, three months before there was anyone to bill [2]

6.5 months

early against the FDA's deadline

cleared under four separate speed-up programmes [1]

The lead story — what happened

The FDA approved Rasonque, also called daraxonrasib, on Wednesday for pancreatic cancer that has already spread. [1][3]
It is the first treatment to attack a genetic cause of this aggressive, highly lethal cancer. [3]
In a 500-patient trial, people on the drug lived a median 13.2 months against 6.7 months on chemotherapy. [1]
RAS is a protein that tells cells to grow, and a faulty copy drives most tumours of this kind. [1]
The drug is built to fit several faulty versions of RAS at once, not one exact version. [1]
The most common side effects listed are rash, mouth inflammation, diarrhoea, nausea, fatigue, swelling and bleeding. [1]
The label covers adults who already had one course of drug treatment, or cannot take the standard combination. [1]
Revolution Medicines set the price at $39,800 a month, about $478,000 for a year, mostly billed to insurers. [2]
More than 2,000 people have been taking it free since May under an early-access programme, and now move onto insurance. [2]
The agency cleared it 6.5 months before its own deadline, and said so in its headline and two official quotes. [1]
About 67,000 Americans are diagnosed with pancreatic cancer a year; it is 3.2% of cancer cases and far more of the deaths. [1]
It is not a cure, though some people who got it in trials have now lived for years. [2]
Median is the midpoint: half the trial did better than 13.2 months, and half did worse.
Immune-based cancer drugs have struggled in this disease [4], and that field has slowed while chasing copies of one old success. [5]

Who is involved

Revolution Medicines

made the drug, gave it away first, then set the price
The FDA

approved it 6.5 months early and led with its own speed
Pancreatic Cancer Action Network

its chief scientific officer said no benefit like this had been seen
Insurers

will receive claims of about $478,000 per patient per year

How it unfolded

Until now pancreatic cancer had late detection, fast decline and few options [1]
May the FDA lets the company hand the unapproved drug out free [1][2]
Since May more than 2,000 people start taking it outside any trial [2]
Wednesday approval arrives, 6.5 months before the deadline [1]
Next the free patients move onto insurance at $39,800 a month [2]

Where this points

Watch what insurers do with a $478,000-a-year claim, and whether the 2,000 free patients transfer across without a coverage fight. [2]

What is pushing on the whole day

The bar and the word are our reading of how hard each one is pushing today. The arrow is where it is heading. The evidence is in the stories below.

Unmet need High→

pancreatic cancer is 3.2% of US cancer diagnoses and a far larger share of the deaths [1]

Price pressure Building↑

$39,800 a month, and almost all of it lands on insurers rather than the maker [2]

Regulator speed Building↑

cleared 6.5 months early, with the agency naming its own urgency in the headline [1]

Depth of evidence Steady→

one 500-person trial where everyone knew who got what; the 2,000 on free access produce no comparison [1][2]

The rest of the day

9 more stories on this beat.

Each with its own sources. None of these is a link to the story above.

02
Measles kills two, in the worst US year since 1991

Two unvaccinated Lancaster County residents died, Pennsylvania officials said Tuesday. [7][8] They are the first US measles deaths of 2026, and the state's first in 35 years. [9][10] The CDC counted 2,777 confirmed cases as of 20 August. [6]

Why it matters — Five Americans have died of measles in two years, more than in the three decades before. [7] In November, officials decide whether the country still counts as having eliminated it. [7]
03
Child vaccination slips as exemptions hit a record

New CDC data show non-medical exemptions at a record 4.2%, with about 280,000 kindergarteners lacking a measles shot. [12] Rates fell in more than half of states. [12]

Why it matters — The health secretary spent the week of the deaths arguing with a governor rather than promoting the vaccine. [11]
04
An English hospital finds its own gaps

Worcestershire Acute Hospitals reviewed 8,500 staff after a three-month measles outbreak, and found over a quarter in patient-facing roles unvaccinated or without natural immunity. [13]

Why it matters — Three vulnerable patients, among them a pregnant woman and two transplant recipients, needed protective antibodies. [13]
05
Hepatitis B drug cleared in Japan

Japanese regulators approved Hibsago, from GSK and Ionis, the first clearance anywhere for the drug. [14] Its makers call it a functional cure: the infection controlled rather than wiped out. [14]

Why it matters — GSK lists it among 15 drugs it expects to become top sellers, with US and other decisions still to come. [14]
06
Bristol Myers walks, a factory maker cuts staff

Bristol Myers ended a deal worth up to $380m with Cellares, saying its automated cell-therapy factories could not make Breyanzi at commercial scale. [15] The layoffs land by 20 October. [16]

Why it matters — Living-cell treatments are still built close to one patient at a time, and that is where they keep failing. [15]
07
One dose for sleeping sickness clears review

European regulators backed acoziborole, a single-dose treatment for a parasite disease that is almost always fatal untreated. [17] Trials reported success rates up to 96% at 18 months. [17]

Why it matters — The previous best option took ten days, and Sanofi has pledged to donate this one in affected countries. [17]
08
Dozens of ageing studies used the wrong antibody

An independent sleuth reports more than 50 papers on cell ageing appear to list an antibody that binds a bacterial protein, not the mammalian one intended. [18]

Why it matters — Nobody has checked whether the findings still hold, and validating a reagent is slow, costly and rewarded by no one. [18]
09
$795m for an epilepsy drug with no results yet

SK Biopharma took worldwide rights to Biohaven's epilepsy drug opakalim, paying up front rather than waiting for the main trial results due later this year. [19]

Why it matters — The money is a bet on the mechanism, placed before the evidence that would price it exists. [19]
10
Three ageing results, none of them in people yet

Nature Aging reported a gut bacterium that cut inflammation and extended healthspan in aged mice. [20] A Yale group blocked an immune protein called C3 and reduced age-related inflammation, also in mice. [21] A third paper built ageing clocks from 172 measures in over 100,000 Chinese adults. [22]

Why it matters — Two of the three are mouse work, and the third is a snapshot in time, not a test of anything. [22]

02 Lesson why it matters

Why the sharpest tool reaches the fewest people

Shape a remedy to fit one case exactly and it fits only that case - and widening it is never free.

The twist

Precision and reach pull against each other - the tighter a remedy is shaped to one case, the fewer cases it fits, and widening it is paid for elsewhere.

How it works

A drug works by fitting a shape on a protein
RAS comes in many faulty versions, and patients carry different ones
A drug cut for one version fits it exactly and misses everyone else
Cut the shape wider and it reaches most patients with the disease
But a wider shape also matches the healthy copies every cell carries
So reach is bought with side effects, never won for free

Where you've seen this

Building rules

a code written for one soil type protects those houses and leaves every other house uncovered

Sanctions

aimed at one firm they are easy to walk around; aimed at a whole sector they catch bystanders

Benefits and grants

a payment defined by an exact circumstance misses near-identical people just outside the wording

Spam filters

a rule tuned to one scam lets the next one through, and a loose rule eats real mail

The catch

Wider is not automatically better. A broad remedy that harms too much is worse than a narrow one that helps a few people completely.

And the whole of it

Everyone who writes a rule, a form or a policy is choosing a width, usually without noticing. From inside one case it always looks obvious that the tool should have been cut to fit exactly here.

03 Truth what's really going on

What is really going on

A drug that adds a median six months to a lethal cancer arrives priced at about $478,000 a year, with 2,000 people already taking it free.

Why it works on us — The comparison offered is 13.2 months against 6.7, which reads as a doubling. Doubling a very small number leaves a small number, and the word does the work.

Who gains

Revolution Medicines — A first approval six and a half months early, a $39,800-a-month price, and 2,000 patients already started on the drug. [1][2]
The FDA's leadership — An approval it can present as proof of speed, in an announcement it wrote and headlined itself. [1]
SK Biopharma — Worldwide rights to an epilepsy drug bought before the trial reads out, when the price of uncertainty is lowest. [19]

Who pays

Insurers and the people who fund them — About $478,000 a year for each patient treated, with out-of-pocket costs varying by plan. [2]
Cellares staff — Bristol Myers ended a deal worth up to $380m and the layoffs land by 20 October. [15][16]
Unvaccinated households in Lancaster County — Two deaths, 70 hospitalisations and a third of infections in children, in an outbreak still spreading. [10]

What nobody knows yet

Open questions from across today’s stories — ours included.

01

How many people the new pancreatic drug will actually reach.

The label requires a prior course of drug treatment, and nobody has published how many patients get that far. [1]
02

How the Pennsylvania outbreak is being counted.

Reuters reports 393 cases across 28 counties statewide; the New York Times reports more than 390 in Lancaster County alone. [9][10]
03

What happened to the 2,000 people on free early access.

They took the drug outside a trial, so there is no comparison group and no result to publish. [2]
04

Whether more than 50 ageing studies are wrong.

The sleuth who found the antibody mix-up says it may compromise some of them; nobody has rechecked the work. [18]

04 Hope carry this

A disease that had almost nothing to offer now has a treatment that lengthens life, and a single-dose cure for sleeping sickness cleared review in the same week.