Ovarian Cancer Coverage from Every Angle

Using Artificial Intelligence Tool to Identify Aggressive Ovarian Cancer Cells

By: Kayci Reyer
Posted: Wednesday, January 30, 2019

Research published in Nature Communications found an association between morphologically diverse cancer cells and aggressive ovarian cancer by using a novel artificial intelligence tool to identify clusters of cells with amorphous nuclei within tumors. Yinyin Yuan, PhD, of the Institute of Cancer Research, London, and colleagues indicated that their novel computer test may be used along with genetic testing to identify women with aggressive ovarian cancer so treatment can be tailored to their needs.

Tissue samples from 514 women with ovarian cancer, amounting to nearly 150 million cells in total, were automatically analyzed for shape and spatial distribution by the computer tool. Tumor spatial zone mapping revealed that critical DNA repair genes, including BRCA1, had much lower activity in tumors containing clusters of cells with a range of misshapen nuclei. The presence of such clusters was also associated with increased disease mortality.

In addition, analysis found that these diversified clusters experienced lower lymphocytic infiltration within the immunoreactive subtype when compared with other tumor zones. This finding suggests that DNA repair dysregulated, morphologically diverse tumor cells may allow disease progression through immune evasion. “These results call for further investigations to elucidate the mechanism by which diversifying ovarian cancer cells mediate and influence immune escape,” noted Dr. Yuan and colleagues.

“By developing a systems approach to define morphological diversification of [the] cancer nucleus, we demonstrated that the presence of morphologically diverse cancer cells predicted poor overall survival, an observation that cannot be explained by known clinical and genetic factors in [high-grade serous ovarian cancer],” wrote the investigators.

Disclosure: The study authors reported no conflicts of interest.

By continuing to browse this site you permit us and our partners to place identification cookies on your browser and agree to our use of cookies to identify you for marketing. Read our Privacy Policy to learn more.