Summary
The upcoming ASCO 2026 meeting will focus on the clinical application of AI in cancer care, emphasizing multimodal AI models that integrate pathology, radiology, molecular data, and liquid biopsy. While AI shows promise in improving cancer diagnosis and treatment, its validation and implementation in clinical practice remain critical challenges. The meeting will also address the intersection of AI and circulating tumor DNA (ctDNA), highlighting their potential in early risk characterization and treatment response assessment.
Charna Albert
April 2026—Artificial intelligence, circulating tumor DNA, trial data: ASCO 2026 next month in Chicago will feature platforms, technologies, and therapeutic approaches once considered future concepts in cancer diagnosis and care.
What comes now is less flashy than a breakthrough advance perhaps, but no less significant. As Janice Lu, MD, PhD, puts it, “The real work now is disciplined clinical translation: validation, collaboration, and implementation, in ways that truly improve patient care.”
Dr. Lu, clinical professor of medicine in the Division of Oncology at Stanford University School of Medicine, will lead a session on the clinical application of AI in cancer care. The meeting’s scientific symposia, in her estimate, will be similarly clinically minded, featuring AI research anchored more in clinical practice than in years past. “We’re seeing a move away from purely technical performance metrics toward studies that tie AI outputs to clinically meaningful endpoints,” she says.
Multimodal AI models also are likely to feature prominently. “A major trend is integrating pathology, radiology, molecular data, and liquid biopsy rather than evaluating each in isolation,” Dr. Lu says. And AI validation and governance are likely to be major topics, as the current “rate-limiting steps for adoption.”
“In short, I expect this year’s news to reflect clinical translation, multimodal integration, and implementation discipline—not just excitement,” she says.
Ahead of the meeting, CAP TODAY spoke with presenters who will cover AI, emerging melanoma diagnostics, and validated new therapeutic targets in gastroesophageal cancer.
Meeting-goers will learn to distinguish AI hype from real help in “A New Era: Clinical Application of Evolving Technologies and Artificial Intelligence in Cancer Care,” the session Dr. Lu will lead.
“We want attendees to leave with a grounded view of AI in oncology,” she says. “Where it is genuinely ready, where it is still investigational, and how clinicians can evaluate claims responsibly while staying open to innovation.”
Her co-presenters are Lee Cooper, PhD, associate professor and director of computational pathology at Northwestern University, and Joseph A. Sparano, MD, professor in clinical cancer therapeutics and chief, Division of Hematology and Medical Oncology, Icahn School of Medicine at Mount Sinai. Dr. Sparano will address leveraging clinical trial data and biospecimens to develop and validate molecular, pathomic, and AI-based diagnostics.
Dr. Lu will focus on multimodal AI models. She cites as an example a model developed by Dr. Sparano and his colleagues, who sought to improve recurrence risk prognostication in patients with HR-positive, HER2-negative breast cancer. Their model, which integrates pathomic imaging, clinical, and expanded molecular models, outperformed the Oncotype DX 21-gene recurrence score for overall 15-year distant recurrence and late recurrence after five years in the training, cross-validation, and holdout validation data sets. It also provided statistically significant and clinically relevant prognostic stratification in the Oncotype low and high genomic risk groups, Dr. Sparano reported in the abstract (Sparano JA, et al. Clin Cancer Res. 2026;32[suppl 4]:GS1-08).