AI breast cancer treatment research is advancing with a new model developed by researchers at University College London (UCL) and the University of Cambridge. The system could help identify treatment options for triple-negative breast cancer, an aggressive disease that can be difficult to treat.
Published in Cell Reports Medicine, the study explores how artificial intelligence can create digital twins of individual tumours. These virtual models help researchers predict treatment responses and investigate alternative therapies.
The model correctly predicted treatment responses in 74% of cases. Researchers tested it using data from 31 patients. The findings offer promise, but further research must establish whether the technology can improve patient outcomes in clinical practice.
How AI breast cancer treatment models work
The researchers developed a system using symbolic AI. This approach applies explicit logical rules to analyse information and generate predictions.
Using genetic information from patients, the team created computer-generated versions of their tumours. Researchers call these models digital twins. They used the virtual tumours to simulate different treatments and examine possible outcomes.
The study focused on triple-negative breast cancer (TNBC). This subtype accounts for approximately 15% of breast cancer cases. It lacks three common treatment targets: oestrogen receptors, progesterone receptors and the HER2 protein.
Consequently, hormone therapies and HER2-targeted treatments generally do not work against this cancer subtype. Doctors therefore need additional ways to identify potentially effective treatment options.
AI model predicts treatment responses
The researchers evaluated the model using data from 31 participants in the PARTNER clinical trial. They simulated different treatment scenarios, including chemotherapy alone and chemotherapy combined with olaparib.
The team then compared the model’s predictions with the outcomes recorded in the original trial. The system correctly predicted treatment responses in 74% of cases.
The researchers also explored alternative treatments for patients whose cancers had not responded fully to standard therapies. The model identified at least one potential alternative treatment strategy for 12 of the 13 patients in this group.
For 10 patients, simulations suggested that an additional targeted therapy might eliminate detectable invasive cancer in the breast and lymph nodes.
However, these results represent computer-generated predictions. The study did not demonstrate that administering the suggested treatments would improve outcomes for these patients.
Why AI could change cancer research
AI breast cancer treatment models could help researchers explore more personalised approaches to cancer care. By analysing the biological characteristics of individual tumours, these systems may reveal treatment options that conventional approaches overlook.
The researchers used symbolic AI because its reasoning can be examined through explicit logical rules. This may help scientists and clinicians understand why a model recommends a particular treatment.
The system could also help researchers identify biological pathways and potential drug targets. These discoveries may guide future studies into why some tumours resist standard treatments.
However, AI recommendations must undergo careful evaluation before doctors can rely on them. Researchers need to confirm that suggested treatments are safe and effective in real patients.
What the findings mean for patients
The findings represent an early step towards more personalised cancer treatment. The model could eventually help doctors investigate additional options when standard therapies prove insufficient.
However, the study involved only 31 patients and used simulations based on existing clinical trial data. Larger studies must establish how accurately the system predicts outcomes across different patient populations.
Researchers must also test whether treatments selected with the model produce better results than current clinical approaches.
Patients should not change or delay prescribed cancer treatment based on these findings. Anyone considering alternative treatment options should discuss them with their oncology team.
What happens next?
The next stage involves validating the model with larger datasets and determining whether its predictions translate into better clinical outcomes.
If future studies confirm its reliability, the technology could support doctors as they evaluate treatment options for difficult-to-treat cancers.
For now, the research highlights how AI could help scientists investigate the complex biology of triple-negative breast cancer. It also demonstrates the importance of distinguishing promising computer simulations from proven medical treatments.
