By Alex Keown


Artificial intelligence technology has an established home in drug design for many developers. Now, researchers are seeing it as a potential tool in driving therapeutic choices in the treatment of ovarian cancer.

Researchers in the United Kingdom used artificial intelligence technology to analyze tissue samples of 514 ovarian cancer patients. The AI technology was used to carefully identify and track the shape of about 150 million different cells in total, Forbes reported. Yinyin Yuan, the team leader in computational pathology at the Institute for Cancer Research, said the AI was used to identify the types of cancer cells found within the patients. The identification of the cancer cells will allow doctors to tailor the most precise treatment for the cancer patients.

As Forbes reported, artificial intelligence technology was able to scan the shape and content of the cells within the tumor. The technology particularly primarily focused on the nuclei and discovered some anomalies. In most cases, Forbes reported, the nuclei of cells are round, or a bit oblong. However, the artificial intelligence scan revealed there were “small patches with misshapen nuclei” in some of the ovarian tumor samples. The women who had these patches had an aggressive form of the disease with limited survivability. Those particular cancer patients had a 15 percent five-year survival rate, compared to 53 percent of the cancer patients without the misshapen nuclei.

Not only that, the Ai revealed that those cells with the misshapen nuclei “had lower levels of genes that repair DNA, including BRCA1-one of the genes,” Forbes said. Those particular genes are often the targets for various drugs in cases of hereditary breast and ovarian cancer. The defective genes were picked up by the AI test, but not typical gene sequencing data, Forbes added.

Yuan noted that the AI technology testing method provides the medical community with a “way of way of detecting tumors with hidden weaknesses in their ability to repair DNA that wouldn’t be identified through genetic testing.” Yuan also noted that the AI testing discovered that immune cells were unable to reach these clusters of misshapen cancer cells. That revelation suggests that the tumors have evolved in such a manner as to actually escape detection from the body’s immune response and could be a reason why there is such a poor prognosis for these patients.

“Our test also revealed that ovarian tumors with these clusters of misshapen nuclei have evolved a new way of evading the immune system, and it might be possible to target this mechanism with new forms of immunotherapy,” Yuan told Forbes.

With these AI-backed findings, the treatment paradigm for these patients will likely be impacted as doctors can select treatment options that are tailored for those particular patients. In particular, Yuan pointed to PARP inhibitors that have been approved for these particular kinds of mutations, such as AstraZeneca’s Lynparza. The PARP inhibitor, developed by AstraZeneca and Merck, has been approved by the U.S. Food and Drug Administration as an ongoing treatment for adult patients with recurrent, epithelial ovarian, fallopian tube or primary peritoneal cancer who are in a complete or partial response to platinum-based chemotherapy, regardless of BRCA status.



BioSpace source: