Image courtesy:

Organizing an AACC Annual Scientific Meeting session on a topic of interest is a running tradition for each year’s chair of the Annual Meeting Organizing Committee (AMOC). This year, the chair’s invited session at the 69th AACC Annual Scientific Meeting & Clinical Lab Expo will examine the emerging use of artificial intelligence (AI) in precision oncology treatment—matching the right treatment for the right patient at the right time.

“From the big perspective, I’d heard about AI and seen it implemented to certain extents in different places,” AMOC chair Stan Lo, PhD, associate professor of pathology at the Medical College of Wisconsin, told CLN Stat. “I don’t know a lot about it myself, and thought this would be a good way to find out more, to learn what AI really means in the context of medical care and its implications for clinical labs.”

AI is cutting edge, on the front lines of science, he continued. Many questions surround this technology: whether clinical labs need to prepare for it and start planning for it in different ways, and whether it will replace current informatics and procedures in pathology.

Is Artificial Intelligence in Genomics Ready for Prime Time? (32416), a late afternoon symposium taking place on July 31 from 12:30 p.m. to 2 p.m., will feature three highly regarded speakers, who will describe how their institutions/organizations are using AI-aided precision therapeutics in oncology, Lo said, using Watson Genomics as an example. Studies looking at this computing system to assess the value of AI-led treatment plans have reported varying results.

“They’ll give us the nuts and bolts of what’s happening now and also a glimpse at where AI in genomics is going,” Lo added.

To get the latest on Watson Genomics and its role in precision therapeutics, attendees will hear from Evan Leibovitz, a Watson for Genomics machine learning and natural language processing associate/business analyst at IBM in Cambridge, Massachusetts. 

Another speaker Michael Berger, PhD is the associate director of the Marie-Josée and Henry R. Kravis Center for Molecular Oncology at Memorial Sloan Kettering Cancer Center (MSKCC) in New York City, which has been performing comprehensive genomic testing for patients with metastatic cancer since 2014.

MSKCC’s in-house panel, MSK-IMPACT, currently includes 468 cancer-associated genes and has the capacity to detect many classes of genomic alterations. This encompasses sequence mutations, copy number amplifications and deletions, and chromosomal rearrangements, Berger told CLN Stat.

By sequencing DNA from both tumor and healthy cells, clinicians can identify acquired somatic and inherited germline mutations. Both mutations can have clinical significance, Berger said. “Somatic mutations represent attractive drug targets because they only occur in tumor cells. Germline mutations associated with cancer susceptibility have implications for patients’ family members who may be at an increased risk for developing cancer,” he explained.

MSKCC to date has tested more than 17,000 patients using MSK-IMPACT. Oncologists have used this information to select the most appropriate and personalized therapies and enroll patients in the clinical trials they are most likely to benefit from, according to Berger.

Not all mutations found in a patient’s tumor are biologically or clinically relevant, he continued. “Distinguishing functional ‘driver’ mutations from benign ‘passenger’ mutations is an area where AI may prove useful. We can leverage our large database to determine which mutations are recurrent or have similarities to mutations that are recurrent, in order to identify the mutations that are most responsible for a patient’s cancer progression,” explained Berger.

MSK-IMPACT also reveals more complex genomic patterns. “The presence of certain mutation signatures can serve as evidence that a patient is likely to respond to immunotherapy,” Berger said. U.S. Food and Drug Administration recently approved the drug Keytruda (pembrolizumab) for patients with a particular mutation signature, microsatellite instability, which MSKCC clinicians detect using MSK-IMPACT.

MSKCC has also used machine learning techniques to build and train a classifier based on molecular features in tumors MSK-IMPACT reveals to help predict the tissue of origin in challenging cases, including cancers of unknown primary. “Finally, by integrating genomic data from MSK-IMPACT with patient-matched clinical data, we can also discover biomarkers predicting therapeutic response and disease outcome,” Berger said.

The session’s third speaker, Nirali Patel, MD, assistant professor of pathology and laboratory medicine at University of North Carolina at Chapel Hill, will discuss using advanced analytics to enable precision medicine in oncology.

Participants who attend this session earn 1.5 CE hours. Register today for the for the 69th AACC Annual Scientific Meeting & Clinical Lab Expo in San Diego July 30-Aug. 3 to learn more about AI and its potential role in treating patients with metastatic cancer.