Endometrial (Uterine) cancer awareness: Photomicrograph of uterine biopsy showing Endometrial cancer or Endometrial carcinoma.
Credit: Md Saiful Islam Khan/Getty Images

Researchers at the University of British Columbia (UBC) say they have used AI to spot distinct patterns in cancer cell images of endometrial cancer that indicate a greater risk of recurrence and death that would otherwise go undetected via traditional diagnostic methods. The study, published today in Nature Communications, can help physicians treating these high-risk patients to identify and develop more comprehensive, beneficial treatment strategies.

“Endometrial cancer is a diverse disease, with some patients much more likely to see their cancer return than others,” said Jessica McAlpine, MD, a professor at UBC and surgeon-scientist at BC Cancer and Vancouver General Hospital. “It’s so important that patients with high-risk disease are identified so we can intervene and hopefully prevent recurrence. This AI-based approach will help ensure no patient misses an opportunity for potentially lifesaving interventions.”

Endometrial cancer (EC) has four molecular subtypes with strong prognostic value and therapeutic implications, as demonstrated in 2013 by The Cancer Genome Atlas (TCGA). The most common subtype (NSMP; No Specific Molecular Profile) is assigned after exclusion of the defining features of the other three molecular subtypes and includes patients with heterogeneous clinical outcomes. But as the study authors note: “The clinicopathological parameters used for decades to classify endometrial cancers and guide management have been sub-optimally reproducible, particularly in high-grade tumors. Specifically, inconsistency in grade and histotype assignment has yielded an inaccurate assessment of the risk of disease recurrence and death.”

While McAlpine and her team have previously developed an endometrial cancer molecular test called ProMiSE that can accurately discern the four subtypes of the disease, the most common of them is a veritable catch-all called NSMP—No Specific Molecular Profile—that accounts for roughly half of all cases. Patients classified with NSMP show a wide range of outcomes.

“There are patients in this very large category who have extremely good outcomes, and others whose cancer outcomes are highly unfavorable. But until now, we have lacked the tools to identify those at-risk so that we can offer them appropriate treatment,” said McAlpine.

For the new research, McAlpine turned to long-time collaborator and machine learning expert Ali Bashashati, PhD, a professor of biomedical engineering, and pathology and laboratory medicine at UBC. The team in the Bashashati lab developed an AI model to analyze images of samples collected from cervical cancer patients. The model was then trained using 2,300 cancer tissue images to discern between the different ECs. From this the new AI model identified the new EC subgroup that showed significantly worse survival rates.

“The power of AI is that it can objectively look at large sets of images and identify patterns that elude human pathologists,” said Bashashati. “It’s finding the needle in the haystack. It tells us this group of cancers with these characteristics are the worst offenders and represent a higher risk for patients.”

The UBC team is now studying how best to deploy this tool into the clinic and for it to be used in combination with existing molecular testing methods. They note that a benefit of the AI approach is that it is cost-efficient and relatively easy to deploy across geographies, potentially bringing it to the community setting where the vast majority of cancer cases are treated.

“What is really compelling to us is the opportunity for greater equity and access,” Bashashati noted. “The AI doesn’t care if you’re in a large urban center or rural community, it would just be available, so our hope is that this could really transform how we diagnose and treat endometrial cancer for patients everywhere.”

Also of Interest