Female breast cancer is one of the most common and deadly cancer types worldwide, along with lung and colorectal cancer. It’s estimated that about 1 in 8 U.S. women will develop invasive breast cancer over the course of their lifetime, and according to the World Health Organization, there were 18.1 million new cases and 9.6 million deaths worldwide last year alone.

Early detection significantly improves outcomes, and fortunately, efforts are underway at Google, MIT, and NYU to improve the accuracy of breast cancer screenings with artificial intelligence. They’re not the only ones — in a paper and accompanying blog post today, scientists at IBM’s Zurich office detailed a partnership with the University of Zurich to develop a system that can identify and classify tumor and immune cells as well as their relationships.

Their work is featured in the journal Cell.

“While researchers have been working hard to develop novel therapeutic approaches to fight against breast cancer, the main reasons for cancer-associated deaths are still therapy resistance, relapse, and metastasis,” IBM computational systems biology researcher Marianna Rapsomaniki wrote in the blog post. “The goal is for this work to lay the foundation for future precision medicine approaches that could potentially help patients win the fight against breast cancer.”

Toward that end, Rapsomaniki and team hypothesized that breast cancer is a heterogeneous disease — i.e., that it comprises tumor cells with characteristics determined by genetic makeup, and environmental influences that communicate and interact with surrounding non-cancer cell types such as immune cells, stromal cells, and vascular cells. Furthermore, they theorized that patterns within these ecosystems might be linked to disease progression and therapy response.

IBM Breast Cancer Study Image

To prove out their idea, the team took non-tumor samples from 144 patients and used mass cytometry — a variation of flow cytometry — to measure more than 70 proteins in over 26 million cancer and immune cells. Next, they used an AI-driven technique to identify various populations of tumor and immune cells and create a detailed atlas of breast cancer ecosystems, which they then used to define the heterogeneity of individual tumors and quantify their abnormality in comparison to matched non-tumor tissue.

Lastly, the researchers analyzed tumor-associated macrophage and T cell populations (which they note can exhibit both tumor-suppressing and tumor-supporting functions), and they associated their findings with clinical information including disease grade or tumor aggressiveness.

In the end, the team found that highly aggressive tumors are often dominated by a single tumor cell phenotype and that each tumor is unique in its cellular composition, with the more aggressive tumors differing most from the rest. Additionally, they found similarities in the tumor-associated immune system among more aggressive tumors.

They believe the work lays the foundation for the design of precision medicine treatments and suggests that immunotherapy might be a viable approach for certain groups of breast cancer patients.

“This could be a reason why a one-size-fits-all approach to cancer treatment is not always effective,” said Rapsomaniki. “Based on our findings, we believe that a specific group of breast cancer patients could benefit from immunotherapy as well. Moving forward, we will investigate the possibilities of immunotherapy in additional studies, potentially leading to a clinical study.”