Curie Brief
Turn on cookies to sign in
Signing in saves your progress to your Curie account. We can only do that with cookies on — turn them on to continue.
NewsWhen cancer trials produce baffling results, AI might hold the answers. Oncologist Maurie Markman proposes using AI to mine medical records from past trials — like two surprising ovarian cancer studies — to uncover overlooked details that could explain unexpected outcomes and improve future drug development.
Sometimes cancer trials don't just fail — they produce results so surprising that researchers are left searching for answers for years. Oncologist Maurie Markman, MD, is proposing a novel idea: use AI to dig through medical records from past trials and surface the hidden clues that explain those unexpected outcomes.
He highlights two phase 3 ovarian cancer trials as prime examples. In one, the experimental drug canfosfamide — which showed real promise in early studies — performed significantly worse than standard chemotherapy. In another, a drug showed only a modest improvement in progression-free survival but a dramatic boost in overall survival, a pattern that defies typical expectations. Neither result has ever been fully explained. Markman argues that an AI-powered review of patient records could surface overlooked drug interactions, comorbidities, or population imbalances that human reviewers missed.
Key Takeaways:
Why it matters: Unexplained trial results aren't just academic curiosities — they can stall drug development and leave clinicians without answers. If AI can help decode these mysteries, it could reshape how oncology trials are designed, analyzed, and ultimately translated into better patient care.
