NIH develops AI tool to better pair cancer patients with drugs

The proof-of-concept study published in Nature Cancer reveals how researchers are using tumor cells to predict whether patients will respond to specific drugs.
By Trevor Dermody
06:10 pm
Share

Photo: sanjeri/Getty Images

Editor's note: This article has been updated with a comment from the NIH.

Researchers at the National Cancer Institute (NCI), part of the National Institutes of Health (NIH), have developed an artificial intelligence tool that predicts whether a patient will respond to cancer-treatment drugs using individual tumor cells, according to a study published in Nature Cancer.

Traditional approaches to drug-patient matching focus on bulk sequencing of tumor DNA and RNA. The new AI approach, named PERCEPTION, takes advantage of single-cell RNA sequencing, which provides better-resolution data to fine-tune predicted drug responses. 

The new approach showed promise in two clinical trials for myeloma and breast cancer. 

The study created a numerical value for the effectiveness of clinical trial drug combinations and ranked them best on individual tumor cell responses to the treatments, allowing researchers to decide the most effective treatment for a specific patient. 

"PERCEPTION demonstrates the feasibility of using single cell measurements as a promising new way to optimize cancer treatment while considering the tumor’s heterogenic composition – it is my hope that future studies will build on it to further advance precision oncology," Dr. Eytan Ruppin of the National Cancer Institute Center for Cancer Research, who supervised the NIH study, told MobiHealthNews in an email. 

THE LARGER TREND

NIH announced in October that a study it funded found a tablet application could be used for early screening for autism. By accurately identifying patients at an early age, the study aims to give patients and parents the support they need.  

In February, the NIH launched the Cancer Screen Research Network, which evaluates emerging technologies in cancer screening with the goal of identifying cancer earlier. The network plans to launch a pilot study this year to determine the effectiveness of using multi-cancer detection in clinical trials. 

On April 11, the NIH released a study in Nature adapting an AI program to predict signs of PTSD following childbirth. The tool examines patient statements who have given birth and will be used to identify patients who are at risk for the disorder. 

Artificial intelligence is being widely used within the life sciences sector for drug discovery. 

Israel-based AION Labs is an AI-enabled drug discovery company built out of a government tender. The organization was formed thanks to a partnership between global pharma and tech companies AstraZeneca, Merck, Pfizer and Teva, alongside venture capital firm Israel Biotech Fund, strategic partner Amazon Web Services and German independent research institute Bio Med X.

AION has launched numerous AI-enabled drug discovery companies, including  OMEC.AI, DenovAI, CombineAble.AI and TenAces

Share