Challenges to acknowledge
Potential benefits of AI in the clinic are exciting, but there are many bench-to-bedside challenges.
A clinically obvious example of AI’s applications is radiographic image analysis. There is no biologic rationale for our RECIST “cut values” for partial response, minimal response, and stable disease.
If AI can measure subtle changes on imaging that correlate with tumor biology (i.e., radiomics), we stand a better chance of predicting treatment outcomes than we can with conventional measurements of shrinkage of arbitrarily selected “target lesions.”
A tremendous amount of work is needed to build the required large image banks. During that time, AI will only improve – and without the human risks of fatigue, inconsistency, or burnout.
Those human frailties notwithstanding, AI cannot substitute for the key discussions between patient and clinician regarding goals of care, trade-offs of risks and benefits, and shared decision-making regarding management options.
At least initially (but painfully), complex technologies like WGS and digital image analysis via AI may further disadvantage patients who are medically disadvantaged by geography or socioeconomic circumstances.
In the discussion period, AACR President Antoni Ribas, MD, of University of California, Los Angeles, asked whether AI can simulate crosstalk between gene pathways so that unique treatment combinations can be identified. Dr. Elemento said those simulations are the subject of ongoing investigation.
The theme of the opening plenary session at the AACR virtual meeting II was “Turning Science into Life-Saving Care.” Applications of AI to optimize personalized use of genomics, digital image analysis, and drug development show great promise for being among the technologies that can help to realize AACR’s thematic vision.
Dr. Elemento disclosed relationships with Volastra Therapeutics, OneThree Biotech, Owkin, Freenome, Genetic Intelligence, Acuamark Diagnostics, Eli Lilly, Janssen, and Sanofi.
Dr. Lyss was a community-based medical oncologist and clinical researcher for more than 35 years before his recent retirement. His clinical and research interests were focused on breast and lung cancers as well as expanding clinical trial access to medically underserved populations. He is based in St. Louis. He has no conflicts of interest.