Cancer Sends a Signal Before It Sends a Cell

Leptomeningeal metastasis is when cancer spreads into the membranes surrounding the brain and spinal cord. It's a devastating complication of solid tumors. The choroid plexus, the tissue that makes cerebrospinal fluid, is supposed to act as a barrier against this. A new paper in Nature Cancer asked how cancer gets past it, and whether it changes that barrier before it even arrives.

What they found

Before any tumor cells physically reach the choroid plexus, the chemical environment there has already changed. Using spatial metabolomics on brain tissue, the researchers found this shift at a pre-metastatic stage, before any cancer cells were present.

The molecule responsible is 5-HIAA, a breakdown product of serotonin. It travels there packaged inside extracellular vesicles released by the cancer cells, which target the blood vessels of the choroid plexus specifically. Once there, 5-HIAA activates a receptor called AHR in the vessel lining. This distorts the shape of the vessels and makes them leaky, before invasion even starts.

They proved this wasn't just correlation. Inject 5-HIAA into a healthy mouse and you get the same leaky, distorted vessels. Block the body's ability to make it and the effect reverses. That's how you prove a molecule is actually causing something, rather than just being present while something else does the work.

Is this intentional? No.

This makes complete sense to me. The second cancer appears in the body, molecules, or in this case vesicles, start flooding the system and making changes. But I don't think it's intentional. Cancer doesn't have a brain. This is a survival mechanism to keep proliferation going, most likely a byproduct. Because the signal comes from a cancer cell, wherever it ends up, it changes that place to make it more hospitable for a cancer cell.

It also points to something bigger. Metastases are likely set in motion from the get go. Stopping cells from leaving the primary tumor is very hard, if priming happens from the moment the very first cancer cell appears in the body.

Can we actually use this? Yes and no.

If you see these molecules in the bloodstream at higher levels, that could be an indicator that priming is happening. But that part is unlikely to be preventable. The real question is whether priming reliably leads to metastasis. If it does, knowing this in advance could support wider surgical margins. But the window between priming and metastasis is likely small. You have to deal with the primary tumor regardless. You cannot leave a patient with a tumor that is removable. And if a patient has low 5-HIAA, should that change the treatment at all?

If 5-HIAA does cause the leaky vessels, its role in brain metastasis is well founded, and the experimental logic here is sound. But I always push back on the idea that a single molecule is the answer. It's one piece in a much more complex system. Targeting it alone to stop the vessels leaking is unlikely to be the full story. That said, we have so few targets for metastasis that any improvement, however small, is worth pursuing.

Worth flagging: the paper doesn't test or propose an actual drug. The tools used to prove the mechanism (a serotonin-synthesis blocker, an exosome-release blocker, genetic knockouts) are research tools, not treatments. Identifying a target and having a drug for it are two different things.

The delivery mechanism, vesicles carrying 5-HIAA, isn't new either. Compared to genomic testing, clinicians could likely put it to use.

Why is this in Nature Cancer?

I don't fully understand why this paper is here. It's not groundbreaking. We already know priming happens, and there's little human translation or new methodology. It's one more molecule that may or may not turn out to be useful.

The science itself is careful and the causal logic holds up. But this is a solid piece of an old, important puzzle, the tumor microenvironment shapes distant sites before metastasis takes hold, not a new one. And that puzzle matters. The more the field brings the TME fully into the picture, the closer we get to targets worth pursuing. But I do worry about what going on in research and academia that results in this paper being in Nature.

Paper: Huang, Hou, Yang et al. "Leptomeningeal metastatic cancer cells induce a permissive choroid plexus vasculature through extracellular-vesicle-derived 5-HIAA signaling." Nature Cancer 7, 754-772 (2026). Read it here.