Blood tests could predict how long it takes until bowel cancer returns based on the same principle used in forecasting the weather, a new study has found.

The blood tests, or ‘liquid biopsies,’ could also pick out people unlikely to respond to the drug in the first place.

Researchers developed a computer model that can track the evolution of cancer from tumor DNA in blood samples when they are taken frequently and at set intervals.

Knowing when drugs are likely to stop working could give clinicians more time to prepare for the next step in a person’s treatment, and to consider offering them alternative therapies or enroll them in clinical trials.

Researchers at The Institute of Cancer Research, London and The Royal Marsden NHS Foundation Trust analyzed tumor and blood samples from people with advanced bowel cancer who took part in an academic clinical trial.

The study is published in the prestigious journal Cancer Discovery today, and was supported by funders, including Cancer Research UK, the NIHR Biomedical Research Centre at The Institute of Cancer Research (ICR), and The Royal Marsden, the Wellcome Trust and the European Union.

Tracking tumor evolution over time

The team combined tumor DNA analysis from frequent liquid biopsies—taken at least every 4 weeks—with mathematical modeling, to track how the genetic make-up of the tumor evolved over time.

Tissue biopsies and blood samples are key tools to help detect changes in tumor DNA, which can signal changes in the way tumors are growing and responding to treatment.

When cancer cells contain certain genetic changes, they can be—or evolve to become—resistant to treatment.

The new study found that a computer model could predict the estimated waiting time until the tumor would stop responding to cetuximab, and the cancer would return.

In 75% of people who initially responded to cetuximab, liquid biopsies picked up changes in the RAS gene before a CT scan was able to show that their bowel cancer had returned.

The researchers also looked at the way changes in a key cancer-related gene, called RAS, affect the response and resistance to cetuximab, a type of drug known as an EGFR inhibitor.

Liquid biopsies picked out 54% of people with bowel cancer whose tumors already had changes in the RAS gene before they were treated with cetuximab, meaning the drug would have been unlikely to work for them.

'The genetic make-up of cancer is highly complex'

Study co-leader Dr. Andrea Sottoriva, Chris Rokos Fellow in Evolution and Cancer, and Team Leader in Evolutionary Genomics and Modelling at the ICR, said:

“In our study, we analyzed tumor DNA from frequent blood samples, which allowed us to closely track the genetic evolution of bowel cancer within patients.”

“Our computer model used information from liquid biopsies to predict how a tumor’s genetic make-up would evolve, and estimate how long it would take for the cancer to return—in much the same way that computer models can forecast the weather.”

“The genetic make-up of cancer is highly complex and ever-changing, and understanding how tumors evolve in response to treatment is key in combating drug resistance.”

Forecasting tumor evolution

Study co-leader Dr. Nicola Valeri, Team Leader in Gastrointestinal Cancer Biology and Genomics at the ICR, and a Consultant Medical Oncologist at The Royal Marsden NHS Foundation Trust, said:

“Our study showed that liquid biopsies are better than traditional tissue biopsies at picking out people with bowel cancer whose tumors are unlikely to respond to a drug called cetuximab.”

“We also found that analyzing tumor DNA from frequent blood samples, which are already taken throughout a person’s treatment, can help predict cancer’s next move.”

“Forecasting how tumors will evolve in individual people with bowel cancer could open up the very exciting possibility of using liquid biopsies for personalized, adaptive treatment.”

Predicting response to treatments

Professor David Cunningham, Director of the NIHR Biomedical Research Centre at The Royal Marsden and the ICR, who was Chief Investigator of the clinical trial associated with the study, said:

“Despite tailored selection, we know cetuximab only works for 50% of patients so it is crucial to identify those who will not respond to this treatment as early as possible. This avoids unnecessary treatment with a drug we know will not work, and means we can consider alternatives at an earlier stage.”

“Tumors shed tiny fragments of DNA into the blood, and these fragments—known as circulating tumor DNA—can be picked up by a simple blood test, or ‘liquid biopsy.’ Liquid biopsies are quicker, cheaper, and less invasive than taking a sample of body tissue. By analyzing the DNA for genetic alterations, also known as somatic mutations, we can follow the progression of a patient’s tumor over time and predict whether they will respond to a particular treatment.”

“At The Royal Marsden, targeted, precision medicine is extremely important, so this study is a significant step forward, which could benefit some of our patients with bowel cancer.”