Making liquid biopsy and artificial intelligence technologies work together could help doctors make faster decisions to treat cancer on time. Philip talked to the CEO of Cambridge Cancer Genomics to find out what this approach can mean for the treatment of cancer. 

Although much progress has been made in recent years, cancer remains one of the most dreaded diagnosis. For many types of cancer, current treatment options only alleviate symptoms or stop working after a while, only delaying the inevitable for a few months.

One of the biggest challenges to treating cancer is its ability to rapidly mutate. “Cancer evolves. Your tumor today is not going to be your tumor in three months,” says John Cassidy, CEO of Cambridge Cancer Genomics. 

ADVERTISEMENT

The very act of treating cancer makes it evolve faster. “When you treat a tumor with a targeted therapy, that acts as a selective pressure that forces evolution,” Cassidy adds.

He draws the example of tumors with mutations in the EGFR gene. Patients with these mutations are treated with a cancer drug called erlotinib, but after 8 months 80% of patients become resistant to the treatment. “The tumor completely changes in 8 months,” says Cassidy. “And that is happening daily.”

Because of this, Cassidy believes that a single diagnosis is not enough. His company addresses precisely this problem by combining two emerging technologies: liquid biopsies and artificial intelligence.

Liquid biopsies make it possible to analyze the DNA of a tumor from a blood sample, removing the need to make an invasive biopsy procedure. This allows Cambridge Cancer Genomics to take a sample every couple of months to monitor how the tumor is changing.

Then, using AI, it can predict whether the current therapy is the most adequate for the patient or whether they should switch to another. With this method, the company claims that it can identify relapse an average of 7 months earlier than the standard practice — potentially saving the patient months of expensive and ineffective treatments.

“Especially when you start talking about treatments that cost half a million dollars a year,” says Cassidy, “you kind of have to know that it’s going to work.”

For example, for patients with the EGFR mutation, the company could determine when their cancer starts developing a mutation that makes it resistant to erlotinib, and recommend to change to a second-line therapy that is more likely to work from then on.

In short, Cassidy’s vision is to “scale the ability of a clinician to monitor a patient during treatment and rapidly change them between treatments.” To help it reach that point, Cambridge Cancer Genomics took part in the renowned accelerator program of Y Combinator last year and is now growing the team and working with physicians to figure out what is the information they need to make the diagnosis of patients the most accurate and up-to-date possible.

Cambridge Cancer Genomics is one of many in a new wave of companies — such as Grail, Guardant Health or Freenome — that are using liquid biopsies to augment the ability of physicians to better predict the outcome of their patients. Cassidy envisions a new breakthrough in the biomarker field comparable to the discovery of the Her2 marker for breast cancer in the 1990’s.

“Biomarkers like Her2 were really transformative for stratified medicine because they meant breast cancer patients were entirely different,” says Cassidy. “They were either Her2-positive or Her2-negative: different drug, different outcome, different relapse profile.”

With liquid biopsy and AI, we could now be closer to making that stratification dynamic, changing as the patient’s tumor changes. By helping doctors keep up with the rapid evolution of a tumor we could, hopefully, eventually outrun cancer and treat it more effectively.

Hear it all directly from our interview with John Cassidy:


Image via Shutterstock

Previous post

Biomarker Test for Crohn's Disease Gets Massive Support From Wellcome Trust

Next post

Branch Out With These Biotech Jobs

Let's Continue The Conversation

Feel free to send us comments about this article to comments@labiotech.eu and/or comment on that article on social media.