Science/Environment

Cancer algorithm uses game theory to double survival time

Using algorithms to monitor cancer evolution and apply game theory to their treatment has doubled the survival time of men with advanced prostate cancer

Cancer cells

Two prostate cancer cells in the final stage of cell division

Steve Gschmeissner/Science Photo Library

APPROACHING cancer treatment as a game has doubled the survival time of men with advanced prostate cancer. This achievement could mark the start of using game theory to target a range of cancers more cleverly.

“This approach is elegant and exciting, and shows real promise to delay treatment failure,” says Charles Swanton at the Francis Crick Institute in London.

People with cancer aren’t usually killed by their initial tumour, but by the rapidly evolving secondary tumours that occur once the disease has spread. To work out how each case of cancer is evolving, Robert Gatenby and his colleagues at the Moffitt Cancer Center & Research Institute in Florida created an algorithm. Built using clinical data, it also suggests the best treatment regime to maximise a person’s survival. This enables the team to use game theory to keep the upper hand over cancer.

 “This approach could revolutionise cancer therapy. I want to try it on every cancer we can”

In this “game”, the oncologists are predators, and the cancer cells are prey. The oncologists’ objective is to kill the prey, or to at least keep it in check. But conventional cancer treatment shifts this balance. By giving a patient repeated strong doses of a cancer drug, the cells are pushed to evolve resistance.

When this occurs, the oncologists stop leading the game and instead have to keep up with an evolving, stronger cancer. By using the algorithm to deploy drugs more subtly, and closely monitoring what the cancer does in response, Gatenby says oncologists can stay ahead.

To test this approach, his team turned to people whose prostate cancer has spread to other parts of the body. They kept track of whether the cancer was growing or shrinking by measuring how much of a chemical called prostate specific antigen was shed by tumours into the blood every month. As this chemical rises and falls, the algorithm calculates how much of the drug abiraterone to administer in each treatment cycle, tailoring dosages and treatment to each individual.

The approach works because it keeps the prostate cancer hooked on testosterone. Many cells in prostate tumours require this hormone, so men with this cancer are given a chemical to stop them making it. However, the cancer often evolves cells that make their own testosterone.

Abiraterone is then used to target these cells, but large doses of the drug wipes them out, prompting cells that don’t need testosterone to take over and ultimately kill the patient. By using less of the drug, and taking breaks between treatment cycles, the team can stop these more lethal cells from becoming dominant, retaining control over the cancer.

Outsmarting evolution

It typically takes prostate cancers 15 months to evolve resistance to standard doses of abiraterone, at which point the tumours are able to grow bigger than their initial size. But in an ongoing trial of the algorithm in the treatment of 17 men, this timescale has more than doubled to an average of 33 months – and could keep rising. Cancer has been able to progress in only three men in the trial, and some of the participants have now lived for four years without this happening, Gatenby told New Scientist. These results are so impressive that the team is beginning a larger trial, and plans to extend the treatment strategy to skin and thyroid cancer.

“This is an approach that could revolutionise cancer therapy, and it doesn’t even require the discovery and approval of a new drug,” says Carlo Maley of Arizona State University. “I want to try this on every cancer we can.”

The team will keep running their small trial until cancers resume progressing in all 17 men, but it could be a long wait. “These patients are still chugging along three to four years after the trial started,” says Gatenby, whose team published a preliminary report of the first 11 patients last year (Nature Communications, doi.org/gcm99n).

The approach is one of several to look at cancer evolution. Rather than go for broke trying to eradicate an advanced cancer, these strategies instead aim to maximise how long a person can continue living a relatively normal life with the disease.

Smart therapies

Game theory could be one way to keep rapidly evolving cancer cells under control, but it isn’t the only game in town.

The CAR-T approach genetically engineers a person’s own T immune cells so that they recognise and kill cancer cells. The strategy has shown promise in treating otherwise incurable blood cancers, including chronic leukaemias. It works because it can eliminate all cancer cells, including those that have evolved drug resistance.

Another tactic is to use immunotherapy to reawaken a person’s immune system. Cancers often evolve to evade a person’s immune system by producing molecules that make them look like normal cells. Drugs like nivolumab unmask these cells, enabling the immune system to see them and weed them out – a strategy that has sent cases of melanoma and lung cancer into long-term remission.

A newer technique is to test dozens of drugs on clones of a person’s cancer in the lab. This should enable oncologists to pick the best drug first time, giving the cancer less time to grow and evolve.

First published in New Scientist on March 7 2018. 

https://www.newscientist.com/article/mg23731682-600-cancer-algorithm-uses-game-theory-to-double-survival-time/

Leave a Comment

Your email address will not be published. Required fields are marked *

*