Presentation (CTI)

OBJECTIVES: Tumor-agnostic approvals represent a major shift in oncology, enabling treatment recommendations based on molecular biomarkers rather than tumor histology. However, health technology assessment (HTA) and reimbursement decisions are typically made within specific cancer types, creating uncertainty regarding the transferability of evidence from pooled multi-tumor populations to individual tumor settings. This study uses a Real-Time AI-Assisted Living Systematic Literature Review (REAL-SLR) to evaluate the evidence supporting tumor-agnostic treatment recommendations in pancreatic ductal adenocarcinoma (PDAC) and identify evidence gaps for HTA decision-making.
METHODS: A PRISMA-compliant, continuously updated REAL-SLR was conducted in PDAC. Interventional studies were mapped by biomarker target, treatment pathway, intervention category, regulatory status, and guideline inclusion. Regulatory approvals, treatment guidelines and HTA-relevant US/EU sources supplemented evidence mapping. Interventional clinical trials in PDAC were included and stratified based on clinical stage, performance status, biomarker target, treatment path, and intervention category. Studies were classified as Tumor-Specific Evidence when pancreatic cancer data were available and Supporting Evidence when recommendations were derived primarily from tumor-agnostic approvals.
RESULTS: As of June 15, 2026, the REAL-SLR contained 581 PDAC studies, including 191 in recurrent or ≥2nd line settings. Fourty-three studies evaluated biomarker-defined populations. US guidelines contained 10 biomarker-driven recommendations compared with three in Europe. 6/10 recommended biomarkers were supported by tumor-agnostic approvals including BRAF V600E, HER2-IHC-3+, NTRK, RET, MSI-H, and TMB-high populations, representing nine unique interventions. Most biomarker-driven recommendations were supported by tumor-agnostic evidence (7/9). When pancreatic cancer-specific evidence was available (HER2-positive, MSI-H), sample sizes were small (n~25) with lower overall response rates (12%-18%, respectively) compared to the overall tumor-agnostic population (N: 267; 155; ORR: 37%-38%, respectively).
CONCLUSIONS: The REAL-SLR database identified substantial gaps between tumor-agnostic approvals and pancreatic cancer-specific evidence. Living evidence stratification may help sponsors, HTA agencies, and payers assess evidence transferability, prioritize evidence generation, and support reimbursement decisions in biomarker-defined populations.