Emad Shash, Leading Cancer Researcher and Director at National Cancer Institute, shared on LinkedIn:
“AI in Cancer Care Needs Both Quixotes and Sanchos!
One of the most memorable messages I heard at ESMO 2026 in Madrid came from Dr. Francisco Sánchez-Vega, computational oncologist at Memorial Sloan Kettering Cancer Center.
His analogy was as insightful as it was practical: the Quixotes bring vision and bold ideas; the Sanchos bring clinical judgment, patient perspectives, real-world data and the discipline of validation. Real progress needs both.
The science is exciting.
Multimodal AI can connect what oncology has long analyzed separately: pathology slides, tumor genomics, imaging, clinical records and patient outcomes. Research tools are exploring how to classify tumors, refine recurrence risk, anticipate treatment response and update predictions as a patient’s disease evolves.
But one slide carried a powerful warning: A tumor-subtyping model achieved 90% top-1 accuracy on internal MSK data, compared with 71% on an external dataset.
Promising AI is not necessarily transferable AI. We need diverse data, independent validation and evidence that these tools improve clinical decisions.
For me, particularly from the perspective of low- and middle-income countries, the greatest opportunity is to make precision oncology more accessible, not more exclusive.
Imagine using AI to prioritize molecular testing where resources are scarce, support multidisciplinary decisions and enable institutions to learn collaboratively while protecting patient data.
The future I hope we build is not AI-led oncology. It is human-led, AI-enabled, patient-centered oncology, where technology strengthens clinical judgment, equity and trust.
Thank you, Dr. Sánchez-Vega, for reminding us that bold vision changes cancer care only when grounded in science, people and reality.”
You can also read: Mario Balsa: Key Insights From ESMO 2026
