Clear Search

Unraveling Breast Cancer’s Mysteries Through Math

By BCRF | March 19, 2019

Novel mathematical methods funded by BCRF may lead to more effective breast cancer treatment and care.

Throughout human history, mathematical approaches have consistently driven the discoveries that advanced civilization. In an effort to better understand breast cancer from all angles, BCRF launched the Mathematical Oncology Initiative in 2017 which brings together mathematicians and computer scientists with biologists and oncologists.

“Mathematics has been fundamental to medicine and oncology for decades,” said BCRF researcher Dr. Allen Tannenbaum whose project belongs to the initiative. He cites the Norton-Simon Hypothesis as a modern example which impacted many lives through a model supporting different chemotherapy schedules.

“New mathematical models, however, are more comprehensive and can be used to study the underlying biology of cancer,” added Dr. Joseph Deasy.

Drs. Deasy and Tannenbaum joined forces five years ago with the goal of applying new mathematical tools to better understand the evolution of cancer – the changes that occur during its progression – as well as what makes one cancer different from another.

“While other methods [such as DNA sequencing and protein analysis] are helpful in pinpointing molecular changes in a cancer from one point to another, mathematics can create a continuum of points and keep track of them through time,” Dr. Deasy explained.

They call their method “high dimension systems analysis” (HDSA). This is where they compare the entirety of a complex system – for instance two tumors – and identify what is different about them. This is often the level of genes or proteins, which may drive or suppress tumor growth

Their expectation is that HDSA can identify patients more or less likely to respond to specific drugs. More importantly, it allows the scientists to identify changes that occur when new things are introduced into the system, such as chemo, radiation therapy or targeted therapies. This in turn can inform the causes of drug resistance and provide insights into more effective treatment approaches.

Computational methods like HDSA accelerate discoveries that can then be tested in biological systems and in clinical trials. Drs. Deasy and Tannebaum have confirmed that their methods work using existing cancer datasets and will soon publish results describing a new subtype of breast cancer identified using their HDSA methodology.

In their next phase of research, the team will work on integrating results from HDSA, i.e. changes that occur during the evolution of the tumor, to information obtained from tumor images through machine learning technologies.

“With BCRF support, we hope our method will evolve, merge with other technologies and provide more precise and predictive information in a clinical setting,” Dr. Deasy said.