The Drug That Keeps Surprising

No, really. Exenatide — one of the first GLP-1 receptor agonists — was developed from exendin-4, a hormone found in the venom of a Gila monster lizard that happens to mimic a naturally-occurring human gut peptide. The lizard doesn't eat very often. Researchers noticed the hormone helped regulate blood sugar. Eventually, after enough molecular tinkering, that observation became a class of drugs that now includes semaglutide — Ozempic, Wegovy — which became the most talked-about medication of the 2020s.

The story of GLP-1 agonists is the story of a drug that keeps rewriting its own job description.

It started as a diabetes medication. Then it caused dramatic weight loss. Then it appeared to reduce cardiovascular events. Then it seemed to help with addiction. Now, in data presented at the 2026 American Society of Clinical Oncology annual meeting, it appears to be meaningfully reducing cancer progression and incidence across multiple cancer types.

A single molecule, touching everything.

i · the weight loss drug that wasn't only a weight loss drug

GLP-1 stands for glucagon-like peptide-1, a hormone your gut produces when you eat. It tells your pancreas to release insulin, tells your brain you're full, and slows gastric emptying. GLP-1 agonists mimic this signal, which is why they work for both blood sugar and weight management.

That's what they were designed to do.

What nobody fully anticipated was the extent to which "fixing your metabolism" would turn out to be touching a much larger system. Chronic inflammation — the persistent, low-grade immune activation that obesity and metabolic dysfunction produce — is a driver of a remarkable number of diseases. Heart disease, obviously. Type 2 diabetes, yes. But also cancer.

Tumors don't grow in isolation. They grow in an environment — surrounded by immune cells, inflammatory signals, blood vessels they co-opt, metabolic substrates they exploit. When you change the metabolic environment dramatically, you change the field in which cancer either flourishes or struggles.

GLP-1 receptors are expressed throughout the body — in the brain, the gut, the heart, the kidneys, and in tumor tissue. The signal this molecule carries doesn't stay local. When you activate GLP-1 receptors, you're potentially speaking to multiple systems simultaneously. The drug was designed for the pancreas. It turns out it's also fluent in the language of inflammation, immune function, and cellular proliferation.

This probably explains why the cancer data is now impossible to ignore.

ii · what asco 2026 showed

The breast cancer finding got the headlines: a study of 111,646 women with BMI ≥ 25, ages 45 to 80, found that those taking GLP-1 medications were about 30% less likely to develop breast cancer compared to non-users. The effect held in matched cohort analysis — a 30.5% reduction controlling for confounders. University of Pennsylvania research, presented at ASCO's 2026 annual meeting.

But the broader cancer progression data is, if anything, more striking.

Real-world analysis showed that GLP-1 users were 38 to 50 percent less likely to progress to stage IV disease across four obesity-related cancer types: non-small cell lung cancer (50% reduction in metastatic progression), breast cancer (43%), liver cancer (38%), and colorectal cancer (31%). The raw stage IV development rates were stark: in lung cancer, only 10% of GLP-1 users progressed to stage IV, compared to 22% on comparison medications. In breast cancer: 10% vs. 20%.

Then there's a molecular signal harder to explain through metabolic improvement alone. Tumor tissue with high GLP-1 receptor expression showed a 33% lower risk of death overall. In breast cancer specifically, high receptor expression was associated with 45% lower mortality. This suggests the drug may have direct effects on tumor biology — not just indirect effects via weight loss or inflammation reduction.

Which means either we don't fully understand the drug, or we don't fully understand cancer biology, or both. Probably both.

The proposed mechanisms are multiplying: reducing hyperinsulinemia, which drives tumor growth signals; suppressing chronic inflammation, which promotes tumor survival; reshaping the tumor microenvironment to allow immune cell access; and potentially direct receptor-mediated effects in tumor cells themselves.

This is nested coherence. A molecule that evolved to manage post-meal blood sugar turns out to be entangled with immune function, inflammation, and tumor biology — not because anyone designed it that way, but because these systems share infrastructure. The boundary between "metabolic drug" and "cancer drug" was always partly a classification artifact. GLP-1 agonists are bumping into that artifact hard enough to break it.

iii · what "keeps surprising" actually means

There's a version of this story about pharmaceutical triumph. That version is being written in a lot of places. It's probably right — as far as it goes.

There's another version worth sitting with: all of these findings are observational. The ASCO studies are real-world data, not randomized controlled trials. They show correlation; they don't prove causation. Confounding is possible. Clinical trials for GLP-1 drugs in cancer prevention are being planned but haven't been conducted. The data will mature — or not.

This caveat matters, but it doesn't explain away the pattern. The same signal keeps appearing across different studies, different populations, different cancer types. At some point, the consistency of the surprise becomes its own information — an indication that the original framework ("this drug does X, full stop") was a simplification carrying more weight than anyone realized.

Drugs that touch fundamental regulatory systems will always have effects that exceed their stated purpose. Aspirin did it. Metformin did it. Statins did it. GLP-1 agonists are joining that company — drugs whose mechanisms turn out to be entangled with processes so central to physiology that benefits spread in directions nobody predicted.

This is what happens when you find a signal the body already uses everywhere. You push on one node of a distributed network and feel it all the way across. The drug isn't surprising because it's exceptional. It's surprising because we thought the body was more modular than it is.

What the GLP-1 story keeps revealing is how connected the system is. Inflammation in fat tissue is not a separate problem from cancer proliferation, is not a separate problem from cardiovascular risk, is not a separate problem from metabolic dysfunction — they are expressions of the same underlying state. Touch one coherently enough and you touch them all.

The tens of millions of people now taking semaglutide-based drugs may be getting additional benefits nobody intended. Or they may not be — the trials will tell. But the research keeps pointing in the same direction, and the direction is interesting: a drug designed for one thing, running on the operating system of everything.

The Gila monster didn't set out to contribute to oncology. The lizard just had an interesting hormone.