A new data layer now sits inside Flatiron Health’s flagship lung cancer dataset. The Panoramic collection, built around non-small cell lung cancer, has grown to hold specimen-level molecular detail spanning more than 345,000 patients.
AI extraction built the new layer, covering 17-plus guideline-recommended and emerging markers in early and advanced disease. Drivers such as EGFR, BRAF, KRAS, ALK, ROS1, RET, MET, HER2, NTRK and PD-L1 are included. Emerging immunotherapy predictors like KEAP1, STK11, TP53, MTAP, CDKN2A, HRAS, NRAS and PIK3CA are tracked as well.
Each testing event records when specimens were collected and received, the test type, the lab and any mutation detail. Researchers can now trace not just whether a biomarker is present, but how and when testing happened.
The added depth opens the door to biomarker prevalence research in rare mutation-defined groups, work that has long been impractical at this scale. That depth matters to biopharma partners, who increasingly build precision therapies on real-world evidence rather than trial data alone.
The expansion fits a wider trend in oncology: real-world records are becoming a core input for precision medicine, with AI doing more of the work to turn messy clinical files into research-ready structure. Drug developers get a clearer view of testing patterns, mutation drivers and how treatment choices evolve.
