An Agentic AI Approach to Predicting Immunotherapy Response

Immune checkpoint inhibitors have transformed cancer care, yet only about 1 in 5 patients respond to PD-1 blockade. Predicting who will respond remains one of oncology’s biggest unsolved challenges.

In this webinar, scientists from Mithrl and Elephas will share how combining functional tumor profiling with agentic AI can reduce that uncertainty, integrating real-time cytokine data from live tumor biopsies with large-scale genomic datasets to reveal the immune signatures that drive response. Attendees will get a firsthand look at how this approach was applied across 131 human tumor specimens and what it uncovered.

Key topics include:

• What these findings mean for biomarker strategy, patient selection, and immunotherapy drug development
• How Elephas’ elive platform captures real-time tumor responses to checkpoint inhibitors using live tumor fragments from patient biopsies
• How Mithrl’s AI Scientific Decision Engine integrates multiple data types to generate biological insights, without the manual data wrangling or pipeline engineering such analyses typically require
• The distinct immune profiles and cytokine signatures identified across patient subgroups, validated against public genomic datasets