The New Fight Against Cancer Is Moving Beyond Genetic Mutations

Imagine a cancer treatment that successfully destroys the cells carrying the mutation doctors have spent weeks identifying, only for the tumor to return months later. The treatment worked, at least for a while. But somewhere inside the tumor, other cancer cells survived, changed their behavior and began growing again.

That problem lies at the center of a different way of thinking about cancer. Instead of treating a tumor primarily as a collection of genetic mutations, some researchers are studying it as a system of distinct cellular states that can coexist, change and adapt under pressure.

Andrea Califano, a Columbia University scientist who studies cancer at the level of individual cells and their regulatory networks, argues that this framework could eventually change how precision oncology approaches drug resistance.

His premise is that the mutations found in a tumor may be less important therapeutically than the cellular programs those mutations ultimately activate.

“The balls in that game can follow an incredible number of different trajectories, just as cancer cells may have an incredible number of different mutational patterns. But, in the end, they can only go in a limited number of slots,” Califano said, using the Pachinko game as a metaphor.

The idea is deceptively simple. Cancer cells can carry an extraordinary variety of genetic alterations, but they may ultimately settle into a much smaller number of functional states.

Those states can determine whether a cell proliferates, evades the immune system or avoids being killed by a drug.

Califano, the Clyde and Helen Wu Professor of Chemical and Systems Biology at Columbia University and head of Biohub, New York, and his colleagues have studied this phenomenon across several tumor types, including pancreatic ductal adenocarcinoma, diffuse midline glioma, cholangiocarcinoma and several types of sarcoma.

In pancreatic ductal adenocarcinoma, or PDAC, researchers found six distinct cell states within each tumor. In diffuse midline glioma, malignant cells from different patients could also be grouped according to patterns of regulatory activity.

The findings suggest that cancer’s apparent genetic complexity may conceal a more limited biological architecture.

The machinery behind a cancer cell

To understand those states, Califano’s group reconstructs networks of regulatory interactions inside cancer cells. These networks map how regulatory proteins influence the genes that determine what a cell does.

“Analysis of these networks tells us, very accurately, which master regulators run the show in each cancer cell,” said Califano.

The researchers call the proteins that control these programs “master regulators.” They act as a kind of command system, integrating signals from mutations and the cell’s surrounding environment and translating them into a particular cellular behavior.

“Master regulators are responsible for integrating the effects of the incredible number of different mutation patterns and extracellular signals in a specific tumor type to implement and maintain the state of individual cells,” explained Califano.

The significance, he said, is that these regulatory systems appear to recur. “Even more importantly, these states and their master regulator proteins are virtually identical in different patients with the same type of tumor,” he said.

If that finding holds across cancers and survives clinical testing, it could offer a different route to precision treatment.

Traditional precision oncology often begins by sequencing a tumor and identifying mutations that can be targeted with drugs. The approach has produced important advances for some cancers, particularly when a tumor depends heavily on a particular molecular driver.

But a tumor is rarely a uniform population of cells.

Different cells can carry different mutations, respond differently to their environment and occupy different functional states. A drug designed to eliminate one population can therefore leave others behind.

That surviving population can become the starting point for relapse.

The problem with chasing mutations

Cancer evolves partly because its cells are not static. As they divide, they can acquire additional genetic alterations and change their behavior.

Califano argues that this makes a strategy focused exclusively on individual mutations inherently vulnerable to resistance.

“Targeting individual mutations will likely never cure cancer.” The problem can begin even before treatment starts.

“Many cells in a newly detected tumor already have mutations that make them resistant to drugs that target the driver mutation,” said Califano.

That means a tumor can contain cells that are effectively prepared for a treatment before the patient receives it.

There is another complication. Two cells can have similar or even identical genetic alterations while behaving differently because they occupy different cellular states.

A mutation may therefore identify an important feature of a tumor without fully explaining how every cancer cell will respond to treatment.

“Drug resistance always wins in the end when you only target a specific mutation.”

The alternative proposed by Califano is to identify the cellular states within a tumor and then find drugs capable of eliminating each one.

Rather than asking which mutation should be targeted, researchers could ask which regulatory programs are keeping the cancer cells alive, which proteins control those programs and which combination of drugs can eliminate the entire collection of states.

A different kind of precision medicine

This does not necessarily mean abandoning personalized medicine. Instead, it could shift what is being personalized.

A future treatment could be based on the particular combination of cancer cell states found in a patient’s tumor.

If researchers can identify a finite set of recurring states, each controlled by particular master regulators, they could design drug combinations that cover the full set.

“The master regulators represent the Achilles’ heels of each cell state,” Califano explained. “The fact they are conserved in all patients suggests that a combination of drugs that targets all of them could be curative for all patients, removing the need for personalized cancer therapies.”

That is a significant claim, and the distinction between a promising research strategy and an established cancer treatment remains important.

The work described so far includes laboratory and computational research. Whether targeting multiple cell states can produce durable cures in people will depend on clinical testing, including whether the relevant malignant states can be identified reliably and whether the drugs needed to eliminate them can be combined without unacceptable toxicity.

Califano’s group has nevertheless begun testing the concept against specific cancers.

In diffuse midline glioma, the researchers identified three existing drugs capable of targeting all seven possible cellular states they had identified. In laboratory experiments, combining two drugs aimed at complementary states significantly extended survival compared with treatment using a single drug.

The result offers evidence for the underlying strategy, but it does not yet establish that the same approach will cure patients.

Cancer does not stay still

There is another reason Califano’s framework could matter: cancer cells can change states during treatment.

A tumor that begins with one distribution of cellular states may have a different composition after exposure to a drug.

“Tumor cells are remarkably plastic,” noted Califano. “They can reprogram themselves, take refuge in drug-resistant states, and proliferate there until they find a recipe to evade the effect of the drug altogether.”

That plasticity complicates treatment because resistance does not always require a new mutation. A surviving cell can sometimes change its regulatory program and enter a state that tolerates the treatment.

“Knowing the states and identifying the drugs that deplete them specifically gives you a combination recipe that blocks the tumor’s ability to seek refuge in these drug-resistant states,” explained Califano.

This suggests that cancer treatment could eventually become more dynamic.

Instead of profiling a tumor once, at diagnosis, doctors might repeatedly examine its cellular composition during treatment. The goal would be to identify changes that could warrant a different therapeutic combination.

“Treatment can create new cell states or select for some states that are resistant,” said Califano.

One example is prostate cancer. Some tumors can develop an aggressive neuroendocrine state after androgen-deprivation therapy. Monitoring cellular states could, in principle, allow researchers to identify such transitions and investigate treatments aimed at the newly dominant population.

The approach could therefore complement, rather than simply replace, genetic analysis. Mutations remain important. But a mutation alone may not reveal what a cancer cell is doing at a particular moment.

Bringing the immune system into the picture

Califano’s work also extends beyond the cancer cell itself.

Tumors do not grow in isolation. They interact continuously with immune cells and the surrounding tissue, creating an environment that can help malignant cells evade immune attack.

At Biohub, Califano is studying how different cancer cell states influence this immune environment and whether immune cells themselves can be reprogrammed to mount a stronger response.

That adds another layer to the state-based view of cancer. A treatment might have to do more than eliminate malignant cells. It may also need to prevent those cells from creating conditions that allow them to hide from the immune system.

“Defeating cancer will require targeting the multiple coexisting cell states and defeating its ability to hide from the immune system.”

The broader proposition is that cancer may be less like a single enemy with a single weakness and more like a system that can occupy several different configurations.

That is where Califano believes artificial intelligence could eventually become useful.

If researchers can map the regulatory networks governing individual cancer cells, identify recurring states and determine which drugs affect each state, computational systems could help search through combinations that would be difficult to test manually.

The concept that cancer is quantum “will allow the use of the sophisticated frontier AI approaches we are developing to create better and better and, more importantly, curative therapies for cancer and, ultimately, for other diseases,” he noted.

For now, the “quantum” description is a conceptual framework for understanding cancer’s organization and adaptability, not a replacement for clinical evidence. But it points to a fundamental shift in the question researchers ask.

Instead of focusing only on what mutations a cancer carries, they can also ask what kind of cell each cancer cell has become, what keeps it in that state and whether it can switch to another one when treatment begins.

The answer could determine how effectively doctors can prevent the cells left behind after an apparently successful treatment from becoming the cells that eventually bring the disease back.

Stay ahead of the Stories shaping our world. Subscribe to Impact Newswire and join our 
WhatsApp Channel for updates on global tech, business, and innovation—all in one place.

Dive deeper into the future with the Cause Effect 4.0 Podcast, where we explore the ideas, trends, and technologies driving the global AI conversation.

Got a story to share? Contact Us to reach a global audience with Impact Newswire.