AI-based computer model offers new weapon against hard-to-treat cancers
In a new paper published in Cell Reports Medicine, the UCL team outlines how they, in collaboration with a University of Cambridge team, developed a new type of treatment-predicting computer program that uses an individual's detailed genetic fingerprint to explore which treatments, if any, may kill their cancer—especially when standard treatments have failed.
A new artificial intelligence (AI) computer model, developed by researchers from University College London (UCL) and the University of Cambridge, offers a promising tool to combat hard-to-treat cancers, particularly triple-negative breast cancer (TNBC). This type of aggressive and fast-growing breast cancer accounts for about 15% of all breast cancer cases.
Traditional AI, such as large language models, often operates as a "black box," making it difficult for medical professionals to understand how the AI arrives at its conclusions. This has raised skepticism within the medical community. To address this concern, the UCL team employed a different approach by utilizing "symbolic AI," which relies on logic and transparent reasoning that can be easily understood and interrogated by doctors.
For their study, the researchers built digital twins—computer-generated tumors—based on the genetic information of 31 participants in the Platinum and PolyAdenosine 5-Diphosphoribose Polymerisation Inhibitor for Research and Treatment of Neo-Adjuvant Early and Relapse-Risk Breast Cancer (PARTNER) clinical trial. These digital twins were used to simulate the effects of various treatments, including chemotherapy alone and in combination with olaparib, a drug known to be effective against TNBC in some instances.
By comparing the outcomes of these simulations with the results from the clinical trial, the AI model accurately predicted the effectiveness of treatment in 74% of cases. Furthermore, the model identified potential additional targeted treatments for 12 out of 13 patients who did not fully respond to the initial treatment. In 10 of these instances, the predictions suggested a high likelihood of eradication of the invasive cancer following the additional treatment.
Lead author Professor Jasmin Fisher, from UCL Cancer Institute, emphasized the broad and powerful insights that personalized virtual tumors can provide to cancer medicine through this symbolic AI-based computer model. The researchers view the computer model as a valuable aid to existing cancer clinical practice, offering new ideas particularly in relation to proteins or genes that could be targeted to restore a patient's sensitivity to existing drugs.
While the clinician may need to consider the potential toxicity of certain drugs, the model could ultimately lead to the exploration and adoption of more effective treatment options.
Written by urgent.news from Medical Xpress's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
