There's a frustrating pattern in cancer research that drives everyone crazy. A new drug looks absolutely brilliant when you test it on cancer cells in a dish. It kills tumors left and right. Everyone gets excited. The drug moves forward. Then it reaches patients and... nothing. It flops. The tumors don't respond the way they did in the lab.
What's going on? According to a study in Cell Stem Cell, part of the problem might be something obvious that we've been ignoring: real tumors have nerves growing through them. Lab dishes don't. And it turns out those nerves can change drug sensitivity by up to 100-fold.
That's not a minor adjustment. That's potentially explaining why drug development keeps face-planting.
The Uncomfortable Truth About Our Cancer Models
Cancer researchers spend a lot of time growing cancer cells in plastic dishes. It's convenient, it's fast, and you can screen lots of drugs quickly. The problem is that cancer cells floating around in pink nutrient broth don't behave the same way as cancer cells that are part of an actual tumor, surrounded by blood vessels, immune cells, connective tissue, and yes, nerves.
The researchers in this study ran CRISPR screens looking for metabolic vulnerabilities in gastric cancer. They tested their approaches across different lab models: traditional cell lines (the 2D stuff in dishes), spheroids (3D clumps of cancer cells), and organoids (mini-organs that more closely mimic real tumor architecture).
Here's the uncomfortable finding: they got different answers depending on which model they used. And when they compared their results to what actually happened when they tested drugs in mice with real tumors, the organoid results matched reality. The cell line results often didn't.
This means drugs that look promising in traditional cell cultures might be failing in patients for reasons that would have been obvious if we'd used better models from the start.
The Plot Twist: Adding Nerves Changes Everything
Now for the really interesting part. Real tumors in the gut are woven through with enteric neurons, the nerve cells that normally control digestion. These neurons don't just sit there as innocent bystanders while cancer does its thing. They interact with tumor cells in ways that matter.
The researchers decided to find out what happens when you add enteric neurons to tumor organoids. Would the nerves just hang out? Would they die off? Would anything change?
What happened was striking. The neurons didn't just survive; they actively modulated tumor metabolism. Specifically, they affected how the tumor cells handled lipids, which are fats that cells need for energy and building materials.
And here's the number that should make every drug developer pay attention: drug sensitivity changed by up to two orders of magnitude depending on whether neurons were present. A drug that worked beautifully in a neuron-free organoid might barely dent a tumor when nerves were in the picture. Or vice versa.
If you're screening drugs without neurons and your results are potentially off by 100-fold, you're basically guessing.
The Molecular Story Behind the Drama
The researchers didn't just observe that neurons change things; they figured out why. The mechanism involves an enzyme in the neurons called acetyl-CoA carboxylase. This enzyme affects whether a particular class of drugs (lanosterol synthase inhibitors) would work against the tumors.
In other words, the neurons are producing signals that change how tumor cells respond to therapy. It's not a random effect. It's a specific molecular mechanism that researchers can now target.
This opens up a potential combination therapy approach. Instead of just targeting the tumor, you might need to target both the tumor (with lanosterol synthase inhibitors) and account for the neuronal influence (by co-targeting acetyl-CoA carboxylase). Hit both pathways, and suddenly you're seeing real effects.
Finding Drugs That Actually Work
The study identified a combination of lanosterol synthase and acetyl-CoA carboxylase inhibitors that showed impressive potency, reaching nanomolar levels. Even better, this combination worked across diverse human gastric cancer organoids, including ones that were resistant to standard therapies.
That's the hopeful part. There may be effective treatments hiding in plain sight that we've been missing because our screening approaches didn't capture the real complexity of tumors.
The Bigger Lesson for Cancer Research
The methodological implications here are significant. If neurons in the tumor neighborhood can swing drug sensitivity by 100-fold, then any screening approach that ignores neurons is flying blind. And it's not just neurons. Tumors exist in complex environments with immune cells, blood vessels, and supporting tissue. All of these could potentially affect drug response in ways we're not capturing.
The path forward probably involves more complex co-culture systems that actually resemble the tumor environment patients have. Yes, these are harder to work with than simple cell lines. Yes, they're more expensive. But if simpler models keep giving us wrong answers, what exactly are we saving?
Sometimes the problem isn't finding new drugs. It's testing them in models that actually resemble reality. Nerves aren't just architectural details in a tumor. They're active participants in how cancer behaves and responds to treatment. Ignore them at your peril.
Reference: Chan BKC, et al. (2025). A combined enteric neuron-gastric tumor organoid reveals metabolic vulnerabilities in gastric cancer. Cell Stem Cell. doi: 10.1016/j.stem.2025.08.006 | PMID: 40902593
Disclaimer: The image accompanying this article is for illustrative purposes only and does not depict actual experimental results, data, or biological mechanisms.