The study, now published in Cancer Cell, shows for the first time that this variation is not just due to the tumor’s environment, but is encoded in the DNA of the cancer cells themselves. This means that even if most of a tumor responds to immunotherapy, some subclones may remain undetected and cause the disease to return.

Tumors as Genetic Family Trees

“We know that even within the tumor of a single patient, there are major differences in genetic makeup depending on where you look,” explains Oncode Researcher Krijn Dijkstra, first author of the study. “In fact, a tumor is made up of a kind of family tree of cancer cells, which are related to each other but can also be very different.”

Until now, it was unclear whether these subclones also differed in how the immune system sees them. “That’s important to know,” Dijkstra adds, “because we see that current immunotherapies rarely clear an entire tumor—suggesting that some subpopulations are less sensitive to immune pressure than others.”

Breakthrough with Patient-Derived Organoids

To investigate, the team used advanced organoid models—miniature 3D tumor cultures grown directly from lung cancer tissue. From three patients, they isolated and grew around 25 organoids per tumor, resulting in 92 distinct subclones. Each represented a genetically unique branch of the tumor.

“This was technically very challenging,” Dijkstra says. “The cells are taken straight from the tumor and need to grow into separate subclones in the lab and survive long enough to be tested with immune cells. That was not possible for a long time.”

By combining organoid technology with immune cells from the same patients, the team could test how well the immune system recognized each subclone. Some triggered strong immune responses—others were barely noticed.

Evidence of Built-In Immune Resistance

When the researchers analyzed the DNA of the subclones, they found that immune-sensitive and immune-resistant cells belonged to separate genetic branches of the tumor’s “family tree.”

“This is the first time we have been able to directly show, at the level of individual clones, that cancers contain subpopulations that are inherently better at evading the immune system,” says Dijkstra. “And that the difference between subpopulations is inside the cells themselves, not just caused by their surroundings.”

Why This Matters for Patients

While immunotherapy has significantly improved outcomes for many cancer patients, full remission remains rare. This study offers a clear reason why.

“The fact that immune resistance is already built into certain subpopulations of cancer cells suggests we will eventually need immunotherapies that can also eliminate those specific subclones,” Dijkstra explains. “Even a few surviving cells can lead to tumor regrowth.”

Importantly, the immune-resistant subclones weren’t rare exceptions. “We were able to isolate them fairly easily,” he notes. “That suggests these aren’t obscure outliers, but common and clinically relevant players in cancer relapse.”

What’s Next

The team now aims to scale up the method and study more patients to see how broadly these findings apply. Another key question: why do tumors maintain this kind of internal diversity? The researchers suspect immune-resistant subclones may play a role in long-term tumor survival or progression. Their organoid models will be critical tools for exploring these questions.

Collaboration and Support

The study was a close collaboration between the labs of Emile Voest (Netherlands Cancer Institute), Charles Swanton (Francis Crick Institute), and Sergio Quezada (University College London). The research began during Dijkstra’s time in Swanton’s lab, in partnership with Quezada’s immunology team, and was completed at the Netherlands Cancer Institute, where organoid expertise was key.

The project was supported by Cancer Research UK and the Oncode Institute.