Testing cancer drug side effects with organoids

The Problem

A major problem with cancer therapies is the side effects. And unfortunately, this is a problem that the newest therapies suffer from too, not just good ol’ fashion chemo. The new wave of drugs designed to help the patient’s immune system fight cancer (immunotherapies) are no exception.

The fundamental problem is cancer cells look a lot like your other, normal cells. How do you kill only a tumor when it looks the same as the healthy cell next to it?

People are willing to tolerate pretty severe side effects to get rid of cancer, but there’s a limit to what’s acceptable. As you might guess, full-body inflammation (called cytokine release syndrome) and organ damage can limit the usefullness and FDA approval of fancy new immunotherapies.

Maybe a somewhat surprising aspect of drug development is the simple fact that it’s hard to predict side effects. We do exhaustive cell culture work and pre-clinical animal studies to try and predict them, but it doesn’t always work. For the obvious reasons, people are working to develop new ways to test side effects more effectively without using animals.

The Solution

This week’s paper is an example of this effort. In the last decade or so, scientists have started taking tissue samples from patient’s and culturing them into what’s called an “organoid” (with the patient’s permission).

Organoids, as the name implies, are like mini-organs that copy the function and characteristics of organs while preserving the genes of the patient they came from. Put simply, it’s a ball of representative cells.

The authors used patient-derived tumor organoids to test out a new promising cancer immunotherapy called a T-cell engaging bispecific antibody (TCBs). Fancy name, but you can think of TCBs as a targetting system for immune cells. They bind to an immune cell on one end and the cancer on the other, letting the immune cell find and kill the cancerous cell.

A major problem with this approach is that the “cancer target” is often present in healthy cells too. But this is super variable between patients.

In some patients samples, the TCB therapy killed more of the healthy organoids than others, indicating that those patients would likely have more severe side effects. Importantly, this level of patient-specific findings isn’t something that traditional cell-culture or animal models would catch.

The image below drives this point home. On the left side, the different numbers correspond to different patient samples (3 = Patient 3, etc). The top categories are different types of TCB therapies. The red blots in each square show the dead cells in each organoid from the treatments.

As you can see, this varies wildly between patients. Some are barely affected (Patient 32) while some show largescale tumor death (Patient 30). Some respond well to some treatments but not others (Patient 30). It’s super variable. And that’s why using patient-derived models is a huge step forward; Patient 30 could be put on the EpCAM treatment while Patient 32 would save months by skipping it and moving on to something else.

In the near future, experiments like this might become clinical testing tools used by doctors to determine the best therapies to pursue. They’re powerful tools on the edge of the clinical world and hopefully make there way in soon.

See you next week for more science,

Neil

If you liked this post and want to keep getting cool science delivered to you, sign up for free: