UCSD scientists report new therapy approach that makes colon cancer cure ‘truly possible’

Cancer stem cells are often difficult to find, can return after treatment or resist therapy altogether. But a new targeted approach from scientists at UC San Diego in La Jolla, with help from artificial intelligence, shows promise in reprogramming colon cancer cells and triggering some to “self-destruct,” according to a study published Oct. 20 in the scientific journal Cell Reports Medicine.

“A cure is actually truly possible. And that’s not us dreaming about it,” said Dr. Pradipta Ghosh, a professor of cellular and molecular medicine at the UC San Diego School of Medicine and the study’s senior author.

Ghosh described cancer stem cells as “shapeshifters” in a statement released by UCSD Health Sciences. Catching them, she told the La Jolla Light, is akin to finding a needle in a haystack.

“Almost none of us would really care about cancer the way we think about it if not for just one thing — that when it relapses and if it spreads, that’s a death sentence,” Ghosh said. “If those … things didn’t happen, we would just treat it as diabetes [or] hypertension, call the surgeon, get it taken out and we’re done.”

The new approach delves into differentiation therapy, a treatment that makes acute promyelocytic leukemia highly treatable. The process aims to revert cancer cells to their normal or near-normal states without damaging others around them.

To get a grip on these cells and identify treatment targets for specific tumors, Ghosh and her team built a machine learning tool named CANDiT, for Cancer Associated Nodes for Differentiation Targeting.

The tool rapidly uses an extensive database to “identify new treatment targets for a specific tumor based on its unique genetics,” the UCSD statement said.

Starting with CDX2, a significant gene found in colon cancer, the tool identified a treatment target called PRKAB1.

By activating that protein, which scientists say assists cells in responding to stress, they were able to restore a pathway that helps with cell differentiation.

Reprogramming the stem cells led to their self-destruction, said Saptarshi Sinha, interim director of the Center for Precision Computational Systems Network of UC San Diego’s Institute for Network Medicine and the study’s first author. This, he said, implies they couldn’t live without their cancerous identity.

“It’s very, very hard to find drugs that kill cancers but not hurt you,” Ghosh told the Light. “It’s almost like ‘We can kill cancer cells’ and the footnote is ‘We also killed the normal ones.’ That doesn’t do us any good.”

“For decades,” Sinha said in the statement, “the Holy Grail of cancer has been its stem cells — resilient, elusive and beyond our ability to identify or track them. They are able to outsmart every form of treatment, even the most advanced immunotherapies. To be able to track and selectively kill them brings us closer to rewriting the rules of cancer treatment.”

The study, Sinha said, spans three models — an organoid, mice and cell lines. The organoid, he said, is special because it replicates the human body in a unique way.

Organoids are small, three-dimensional masses that are derived from stem cells and mimic organs. Scientists used UCSD’s Humanoid Center for the models, accounting for differences such as genetic mutation types, tumor stages and more.

“Organoids are incomplete. … It doesn’t have an immune system and doesn’t have other things,” Ghosh said. “However, what the organoids allowed us to do here is actually conduct a clinical trial in a dish. We could line up multiple patients’ tumors — organoids, that is — each with an annotation of what are the other genetics. Because everyone’s tumor is different.

“It is very clear to us that some of these findings either would have been completely clouded by 2D studies, if not entirely missed or wrongly concluded. So organoids gave us the confidence, the accuracy, the real-world testing [and] the impact of mutation.”

Researchers also developed a gene signature to predict how people would respond to the new form of therapy. Testing it across 10 patient groups and 2,100 people, they concluded the CDX2-restoring drug could cut the risk of colon cancer recurrence and death by up to 50%.

Further research is underway to make a more potent version of the compound, push it to critical trials and deploy the CANDiT technology in studies of different tumor types.