Researchers at Case Western Reserve University have found that the number of tubules in tumors may predict which women with estrogen receptor positive (ER+) breast cancer will benefit from hormone therapy alone and which require chemotherapy, providing a cheaper and faster alternative to genetic risk tests.
The study, “Automated Tubule Nuclei Quantification and Correlation with Oncotype DX risk categories in ER+ Breast Cancer Whole Slide Images,” published in the journal Scientific Reports, also describes a novel computer program that automatically counts tubules shown in whole slide images of breast cancer tissue specimens.
The conundrum in the treatment and management of early-stage ER-positive breast cancer is identifying which cases are candidates for adjuvant chemotherapy and which patients will respond to hormonal therapy alone. Oncotype DX (ODX) and other gene expression tests allow clinicians to distinguish these patients, but such tests are usually expensive, harmful to tissue, and require shipping tissue samples for testing.
Studies have shown that the grade of ER-positive breast cancer tumors highly correlates with the results of these genetic tests, suggesting that grading the tumors may be a possible approach to distinguish which patients should receive each type of therapy. However, when the grade is determined by pathologists, it can be highly variable.
Therefore, the research team developed a computer-aided system that cheaply and quickly provided similar answers to those obtained with genetic tests.
Currently, grading is performed using three criteria: mitotic index, which reflects the proportion of cells that are dividing; tubule formation; and variability in nuclear size. Elevation of levels in any of these three categories indicates an increased risk that the cancer is aggressive and requires chemotherapy.
The investigators used machine learning to allow a computer to automatically quantify the number of tubules in each tissue sample. This was achieved through a ratio of tubule nuclei (green dots) to overall number of nuclei (red dots).
Using slide images from 174 ER-positive patients, the investigators found that the number of tubules correlates with patients’ Oncotype DX gene expression test risk scores.
“This is the first large-scale validation that fundamental prognostic information is in the tissue data, and that it predicts the underlying genomics of the tumor,” Anant Madabhushi, F. Alex Nason Professor II of biomedical engineering at Case Western and study coauthor, said in a press release. “If we can mirror the genomics of the tumor, we can predict who responds to hormone therapy only and who doesn’t.”
The study is the first of three planned by Madabhushi’s lab in this area of research. The next will quantify the number of dividing cells in whole image slides and correlate the numbers with the genomic test. The final study in the series will quantify nuclear pleomorphism and architecture and correlate them with the genomic test.
If all the studies contribute to accurate predictions, Madabhushi plans to integrate them and develop a single integrated program for providing coherent breast cancer prognoses.