A blood test for colon cancer, and the rush to catch disease sooner

Biotech & Longevity 4 min 16 sources

The FDA cleared Freenome's blood test to screen for colorectal cancer, part of a wider push to spot disease years earlier. Plus a melanoma drug backed over its own reviewers' doubts, a Duchenne therapy rejected, and a weight-loss drug that cut heavy drinking.

Key takeaways

The FDA approved a blood test to screen for colon cancer, part of a wider push to catch disease years earlier - even before symptoms.
An advisory panel backed a melanoma drug despite the FDA's own staff doubting the data, and rejected a Duchenne heart therapy the same week.
A next-generation weight-loss drug sharply cut heavy drinking in a mid-stage trial, one more sign these drugs quiet more than appetite.

Catching cancer from a vial of blood

The US Food and Drug Administration, the country’s drug and device regulator, approved a blood test from the company Freenome to screen for colorectal cancer [1]. It is one of a small number of blood-based screens cleared for a cancer that kills tens of thousands of Americans a year.

The pitch is simple. A colonoscopy - the standard screen, where a doctor threads a camera through the bowel - is thorough but invasive, and many people who should get one never do. A blood draw is easy. The bet is that a test people will actually take, even if it misses more, catches more cancer across a whole population than a better test people skip.

That is also the catch. Blood tests generally find fewer early cancers, and far fewer of the precancerous growths a colonoscopy can spot and remove on the same visit. Ease buys participation; it does not buy the sharpest eyes. An approval clears a test for use - it does not settle whether the test, in the real world, means fewer people die of this cancer.

The wider bet: find it years before it shows

Freenome’s approval sits inside a bigger shift - the drive to detect disease earlier and earlier, sometimes before any symptom.

The US health secretary, Robert F Kennedy Jr, called this week for wider Alzheimer’s screening [2]. But the federal panel that vets which screens actually help, the US Preventive Services Task Force, has been kept from meeting for more than a year, and outside experts told the Guardian the evidence for routine cognitive screening is “not great” [2]. Guidelines for cervical cancer screening, meanwhile, have shifted as new tests arrive, including HPV self-sampling [3].

The lab science is running further ahead. A five-protein signature in the blood flagged fatty-liver disease up to 16 years before it appeared, across more than 50,000 people in a study in Nature Aging [4]. Another blood study, in Nature Medicine, predicted which carriers of ALS-linked genes would go on to develop the disease, and roughly when [5]. A separate model read routine blood counts to screen for leukemia across several hospital groups [6]. None of these is a product yet. All point the same way: pull the moment of detection earlier.

A melanoma drug cleared over its reviewers’ doubts

At the same regulator, an advisory committee voted in favor of Replimune’s treatment for advanced melanoma, the deadliest skin cancer [7]. That came days after the FDA’s own staff had raised concerns that the trial data did not clearly show the drug worked [8]. The panel’s vote is advice, not a final decision, and the split shows how much judgment sits inside “the evidence.”

The same panel cycle delivered a rejection. FDA advisers voted against Capricor’s cell therapy for the heart damage of Duchenne muscular dystrophy, a fatal muscle-wasting disease that mostly strikes boys [9]. Agency staff had questioned whether the treatment worked [10]. For families, an advisory “no” can close the nearest door.

A weight-loss drug that curbed drinking

One trial result landed from an unexpected direction. Altimmune said its next-generation GLP-1 drug - the class behind the big weight-loss medicines, which mimics a gut hormone that blunts appetite - sharply cut heavy drinking in a mid-stage trial in people with alcohol use disorder [11]. The company called the effect “unequivocal” [11].

Treat that as a signal, not a verdict. This was a phase 2 study - the middle test, sized to hunt for a hint of benefit, not the large final trial that decides approval. Company-reported results also flatter easily. Still, the idea that these drugs quiet more than hunger keeps gathering evidence.

Money moving behind the science

The business end stayed busy. Argenx agreed to buy Forte Biosciences for about $2.2 billion, chiefly for an early-stage drug for vitiligo, the condition that removes skin pigment in patches [12]. GSK laid out plans to cut costs while pushing more drugs into late-stage trials [13], and signed an artificial-intelligence drug-discovery deal worth up to $110 million with Relation Therapeutics [14]. Johnson & Johnson lowered its 2026 profit forecast, blaming the cost of its own deals [15].

And a quieter story of a disease long written off: alpha-1 antitrypsin deficiency, an inherited disorder that can wreck the lungs and liver, has gone “from lost cause to gold rush,” with biotechs now racing to treat it [16]. Neglected diseases get their turn when someone finds a way in.

02 · Lesson · why it matters

How catching a disease earlier can improve the statistics without improving a life

Survival is counted from the day of diagnosis. Move that day earlier and the count rises - even if the patient dies at the very same moment.

The vial of blood and the promise it carries

This week the FDA cleared a blood test that screens for colon cancer. Behind it sits a much bigger idea, running through Alzheimer’s screening, cervical checks, and lab work that spots disease years before a symptom appears. The idea is intuitive and almost sacred: find it early, and you save a life.

Sometimes that is exactly what happens. Catch a tumor before it spreads, cut it out, and the person lives who would have died. But not always. And the trap is that even when earlier detection changes nothing at all, the numbers it produces still look like a triumph. Learning to see that gap is one of the most useful things you can carry out of a health headline.

Where the clock starts is a choice

Here is the sentence to hold onto. Survival is almost always measured from the day of diagnosis. “Five-year survival” means: of the people diagnosed, how many are alive five years later.

Now picture two people with the identical cancer, growing at the identical speed, who will both die on the exact same day. The first is diagnosed when a lump appears - say, three years before death. Her survival is three years. The second is diagnosed by a blood test that caught the same tumor five years earlier. His survival is eight years.

Nothing about the disease differed. Nobody lived a day longer. But the second person’s “survival from diagnosis” more than doubled. All that changed was where the clock started.

The name for the trick

This is lead-time bias. Screening does not add time to a life; it adds time to the diagnosis. The clock starts sooner, so the interval between diagnosis and death gets longer on paper, whether or not the person dies later.

It is why “people diagnosed early live longer after diagnosis” is a sentence that can be completely true and completely empty. Of course they live longer after diagnosis - you diagnosed them sooner. The one thing that phrase never proves is that anyone actually died later.

There is a quieter cousin, too. Screening tends to catch slow, mild cases more often than fast, aggressive ones, simply because slow cancers sit around longer waiting to be found. So the screened group is stacked with gentler disease from the start. That also makes screening look better than it is.

The only question that survives the trick

So how do you tell a real win from a bookkeeping one? You stop asking whether people live longer after diagnosis, and ask whether fewer people die.

Not survival from the day of the test - deaths in the whole population. Take everyone offered the screen and everyone not offered it, wait, and count the coffins on each side. If the screened group buries fewer people, the test earned it. If the two groups die at the same rate while the screened group merely carries its diagnosis for more years, all you bought was a longer stretch of knowing.

This is why a regulator approving a test is not the same as the test saving lives, and why good screening trials are enormous and slow. They have to be. Only a body count large enough, watched long enough, can see past a clock that anyone can move.

The thing that binds us to the numbers

It would be easy to read this as a reason to distrust screening. It is not. Some early detection genuinely saves lives, and the tests arriving now may do real good. The point is narrower and stranger. The very same result can be a breakthrough or an illusion, depending on where you start the clock. And the illusion feels exactly as reassuring as the real thing.

That reassurance is the part that reaches you. When a headline says a group lived longer after diagnosis, or a company’s blood test “caught it early,” the warmth you feel is doing the persuading, not the evidence. We are inside these numbers, not above them - patients, worriers, people who will one day sit in a waiting room hoping the early news was the good kind. Knowing that where the clock starts is a choice does not make the hope smaller. It makes it harder to sell you a longer wait dressed up as a longer life.

03 · Lab · your turn

Start the Clock

Rehearse lead-time bias: move detection earlier and watch survival-from-diagnosis rise while the day of death holds fixed.

04 · Hope · carry this

The tools to read a disease years before it speaks are genuinely arriving, and that is real progress. And the same science that can be fooled by an early clock is the one teaching us to ask the harder question - not whether the number looks better, but whether people actually live longer.