Nicholas Turner, MD, PhD, FRCP, FMedSci

Improving prediction of the risk of recurrence of ER-positive breast cancer.

Impact

Estrogen receptor-(ER) positive breast cancer is the most common subtype of breast cancer, affecting approximately 70 percent of all people diagnosed. Many hormone therapies have been developed to target the ER in cancer cells and are given for localized, early-stage cancers along with surgery and often radiation. With these treatments, most patients are cured but some will recur. The patients who are at risk of recurrence despite hormone therapies and surgery, are often offered chemotherapy to improve the chances of cure. Tests such as the Oncotype Dx™, Prosigna™, and EndoPredict™ that can predict the risk of recurrence measure gene expression in cancer as a whole—usually from a sample of thousands to millions of cancer cells—to determine whether the cancer is more aggressive (high risk) or less aggressive cancer (low risk). The main limitations of these tests are that the specimens contain complex mixtures of different cells including tumor, immune, and many other non-cancer cells. This means that information is missing about how neighboring cells communicate with each other, as immune and normal cells in the cancer can influence how a tumor responds to treatment. The tests will also miss small areas of cancer hiding among otherwise normal cells. Some cancers recur even though the test predicts low risk, and many patients with high-risk test results don’t have a recurrence. Dr. Turner aims to improve the prediction of risk of recurrence and seeks to better understand the biology of what causes a cancer to recur.

Progress Thus Far

Dr. Turner and his team have made progress in developing and testing spatial cell technologies to gain information about the complex nature of tumors. The team is using a new technology capable of analyzing many individual cells within each cancer to analyze approximately 1,000 breast cancer samples from the ATAC trial, which tested hormone therapy in postmenopausal women. In many cases, results from OncotypeDX and Prosigna analyses are known, allowing the team to directly compare their own single-cell analysis with established prediction methods. The team has now analyzed 424 tumor samples from 360 patients. This analysis showed that a single-cell approach can separate samples into high- and low-risk categories similarly to the OncotypeDX assay, while also revealing substantial differences between cancers in their immune and normal cell populations, differences the standard tests cannot detect. The team is now positioned to complete the full analysis of 1,000 tumor samples and define precisely how immune and normal cells are organized within each cancer.

What’s next

Next, Dr. Turner and his team will be using mathematical and computational methods to build a detailed map of how individual tumor cells interact with the immune and normal cells surrounding them, including how tumor cells signal to nearby cells and how these communication patterns vary from one cancer to another. The team will then test whether these interaction patterns can help predict a patient’s risk of recurrence, potentially revealing what the standard gene expression tests are missing, such as immune cells that fail to attack cancer cells rather than actively suppressing them—information that could lead to improved, more personalized recurrence-risk tests. For patients in the study whose cancer has already returned, the team will also analyze these recurrence samples using the same approach, to understand how the cancer’s biology changed when it recurred.