New eco-evolutionary mathematical model can predict prostate cancer aggressiveness

Reviewed by Emily Henderson, B.Sc.May 14 2020

Prostate cancer is the most common cancer in men. One in nine men will be diagnosed during their lifetime. While most men will not die from prostate cancer, there is a small subset of patients whose disease is so aggressive at the time of diagnosis that surgery and radiation are not able to control their cancer. Researchers in the Center of Excellence for Evolutionary Therapy at Moffitt Cancer Center want to better understand what is happening in the tumor microenvironment to drive prostate cancer to become aggressive and grow rapidly. In a new article published in Nature Ecology & Evolution, the research team provides a closer look at a multiscale mathematical model they developed to analyze integrated biologic and pathologic data to determine tumor aggressiveness.

The prostate is made up of epithelial tissue surrounded by supporting stroma. While prostate cancer begins in the epithelial tissue, the surrounding stroma contributes to cancer growth and progression. The Moffitt researchers wanted to learn how the stromal ecology, which can be inhibitory, highly reactive or nonreactive, shapes prostate cancer tumor evolution.

Our previous models show that reactive stroma has a role in prostate cancer progression, but this new model allowed us to look more closely at how the stromal ecology alters tumor evolution, growth and invasiveness."

Alexander R.A. Anderson, Ph.D., director of the Center of Excellence for Evolutionary Therapy and chair of the Integrated Mathematical Oncology Department at Moffitt

He led the Moffitt team with collaborators from the University of Texas Health Science Center and NorthShore University.

To do this, the researchers developed an eco-evolutionary mathematical model that mimics how prostate cancers grow and simulated tumors growing in different stromal ecologies. The model predicted larger tumors with highly reactive stroma and smaller tumors in nonreactive stroma. Surprisingly, it also predicted that these tumors evolved differently, with the smaller tumors driven by nonreactive stroma being far more aggressive.

This result was validated in animal models and quantified in patients. Anderson and his team analyzed pathology data from prostate cancer patients guided by the mathematical model. They used patients' Gleason scores, a standard grading for prostate cancer diagnosis, along with reactive stroma grading, a measurement based on the amount of reactive stroma in the prostate, to create a new integrated cancer biomarker capable of more accurately scoring a patient's prostate cancer. That integrated cancer biomarker was then cross validated in a large cohort of prostate cancer samples and was able to accurately stratify all Gleason scores.

"Our results demonstrate that traditional metrics like the Gleason score, which are often tumor-cell centric, could be improved with ecological metrics," said David Basanta, Ph.D., co-author and associate member of Moffitt's Integrated Mathematical Oncology Department.

The mathematical models and patient samples all agreed and showed that stromal ecology can explain prostate cancer growth dynamics. Importantly, they also showed that the stroma is a major driver of the aggressiveness of prostate cancer cells.

"This suggests that aggressive, environmentally independent prostate cancer may be a result of poor stromal ecology, supporting the concept that incorporating markers of stromal ecology into prostate cancer scoring could improve a patient's prognosis," said Anderson.

The researchers are further testing this theory in a clinical setting.