ESMO Congress 2023
Medical innovation, artificial intelligence, and the potential for a new era of oncology care
The past few years have seen AI tools begin a transition from research into clinical practice, with the FDA and other regulatory bodies now actively evaluating these tools for approved use, where they may support shared decision-making. Performance data from several different AI tools were presented during ESMO 2023, demonstrating their potential benefit and highlighting any possible pitfalls of use and areas for improvement. Types of AI tools discussed included imaging tools, prognostic assistance, and predictive models, with a few key take-home messages captured here.
To facilitate improved treatment targeting for prostate cancer, a pathology-based MMAI prognostic test (ArteraAI Prostate) was developed and assessed using final data from three STAMPEDE trials. The results showed that ArteraAI was successfully validated, demonstrating stronger prognostic associations when compared with individual variables, and emphasizing the value of AI in prognostication.1
AI may also be used to support the development of personalized therapies, as discussed during a proffered paper session by Borgers et al. presenting results on T-cell product therapy (BNT221). Predicted using a bioinformatics platform, the immunogenic neoantigen product for each patient was identified and then used to prime, activate, and expand memory and de novo T-cell responses. Initial results showed tumor shrinkage in 4/9 patients, but it is too early to detect the extent of durability of response.2
Another suggested use of AI is in predicting relapse, as shown in one poster presentation of an AI model combining clinical data with radiomic results from CT imaging in patients with CRC. The results demonstrated the improved ability to predict relapse with the model containing radiomic data compared with the model including clinical data alone. Further work is needed to improve performance and ensure clinically meaningful results, but this is certainly an encouraging advancement.3
AI, artificial intelligence; CRC, colorectal cancer; CT, computed tomography; ESMO, European Society for Medical Oncology; FDA, Food and Drug Administration; MMAI, multimodal artificial intelligence.
1. Parker CTA et al. ESMO Congress 2023. Abstract 1767MO; 2. Borgers JSW et al. ESMO Congress 2023. Abstract 1017O; 3. Bueno Gómez A et al. ESMO Congress 2023. Poster 579P.