Seven studies examine cross-country evidence, AI-derived data quality, and treatment and safety patterns to help inform oncology research and access decisions
Flatiron Health today announced seven research abstracts to be presented at ISPOR—The Professional Society for Health Economics and Outcomes Research Europe 2026 conference, taking place November 8–11 in Vienna, Austria. The research examines how real-world evidence can support health technology assessment (HTA), how oncology findings may translate across countries, and how transparent validation can support the use of AI-derived data in research.
Flatiron will also host a speaking session on scaling AI-curated oncology data across countries, featuring clinical and scientific experts discussing how to maintain data accuracy, consistency, and clinical relevance as datasets expand across markets and disease areas.
“Answering whether a cancer treatment works is only part of the challenge. Researchers and health technology assessment bodies also need to understand how that evidence applies to patients in different healthcare systems,” said Nathan Hubbard, Chief Executive Officer, Flatiron Health. “Our research at ISPOR Europe examines what it takes to make those judgments with confidence, from understanding differences in treatment patterns to validating the quality of AI-derived data. This work can help strengthen the evidence that informs access to cancer treatments.”
Flatiron combines multinational oncology data from the US, UK, Germany, and Japan with clinical expertise and analytical methods to address questions facing researchers, payers, and regulators. This year’s studies examine both the opportunities and practical considerations involved in generating evidence across healthcare systems, including population comparability, subsequent treatment choices, and differences in routine monitoring.
Research highlights include:
- Evaluating when evidence can inform decisions across countries. Three studies examine the transportability of oncology evidence, including a structured approach to reviewing cross-country studies, benchmarking methods for comparing survival estimates, and a comparison of US and Finnish metastatic non–small cell lung cancer (NSCLC) cohorts. Together, the studies explore how researchers can assess whether international evidence is appropriate for local HTA questions.
- Understanding how subsequent care affects survival analyses. An advanced NSCLC analysis examines how second-line treatment patterns can influence estimates of first-line treatment effects on real-world overall survival. The methods offer a way to distinguish the effects of initial therapy from differences in the treatment patients receive next.
- Assessing AI-derived data quality and cross-country safety measurement. A UK prostate cancer study extends Flatiron’s VALID framework to evaluate data extracted using large language models. A separate study examines how laboratory monitoring practices in the UK, Germany, and Japan may influence detection of liver toxicity among patients receiving CDK4/6 inhibitors for advanced breast cancer.
- Examining variation in metastatic prostate cancer treatment. A US study describes differences in treatment selection and androgen receptor pathway inhibitor reuse across racial, ethnic, and area-level socioeconomic groups, highlighting questions for further research into access and clinical decision-making.
At booth #1018, attendees can explore Flatiron’s global Panoramic datasets through live demonstrations of Flatiron Telescope, its AI-powered oncology intelligence platform.
Schedule a meeting with Flatiron Health at ISPOR Europe 2026 and follow Flatiron Health on LinkedIn for conference updates.
Featured Speaking Session
Setting the global standard: What it takes to scale AI-curated oncology data across countries
Tuesday, November 10 | 15:15–15:45 | Exhibit Hall
Arun Sujenthiran, MD, Clinical Lead and Senior Medical Director, International, and Qianyi Zhang, MS, Director, Research Sciences, will discuss Flatiron’s approach to transparent data validation and its application across US and international oncology datasets.
Abstracts and Poster Presentations
Key Considerations for Synthesizing Cross-Country Oncology RWE Transportability Studies
This study identifies five recurring methodological dimensions across oncology transportability studies and proposes a practical checklist to help HTA bodies and regulators interpret international real-world evidence.
Poster Session 1 | SA8
Monday, November 9 | 10:30–13:30
Discussion period: 12:30–13:30
Benchmarking as a Tool for Assessing Cross-Country Comparability of Survival Evidence for HTA
Using exploratory case studies in metastatic NSCLC and multiple myeloma, this analysis examines how comparisons of predicted and reference survival estimates may help researchers assess the suitability of international evidence for HTA.
Poster Session 1 | HTA28
Monday, November 9 | 10:30–13:30
Discussion period: 12:30–13:30
Comparison of Patient Characteristics and First-Line Treatment Patterns in Metastatic NSCLC: US and Finland Cohorts to Inform Overall Survival Transportability Analyses
This study compares patient characteristics and first-line treatment patterns in US and Finnish cohorts to inform future analyses of whether survival findings can be applied across the two settings.
Poster Session 2 | SA30
Monday, November 9 | 16:00–19:00
Discussion period: 18:00–19:00
Controlling for What Comes Next: Estimating Generalized Direct Effects of First-line Therapy on Real-world Overall Survival Using Counterfactual Second-line Distributions in Advanced Non-small Cell Lung Cancer
This analysis explores a method for estimating first-line treatment effects on survival under alternative second-line treatment scenarios, helping researchers account for differences in subsequent care.
Poster Session 4 | MSR200
Tuesday, November 10 | 16:00–19:00
Discussion period: 18:00–19:00
Extending the Validation of Accuracy for Large Language Model (LLM)–/Machine Learning (ML)–Extracted Information and Data (VALID) Framework Globally: Quality Assessment of LLM-Derived Prostate Cancer Real-World Data in the UK
Applying the VALID framework to a UK prostate cancer dataset, this study evaluates LLM-derived data through variable-level assessments, clinical plausibility checks, and replication analyses.
Poster Session 5 | MSR263
Wednesday, November 11 | 09:00–11:30
Discussion period: 09:00–10:00
Real-World First-Line Treatment (tx) Patterns in Androgen Receptor Pathway Inhibitor (ARPI)-Experienced Metastatic Castration-Resistant Prostate Cancer (mCRPC) by Race/Ethnicity and Area-Level Socioeconomic Status (SES) in the US
This study describes variation in first-line treatment selection and ARPI reuse across racial, ethnic, and area-level socioeconomic groups among US patients with ARPI-experienced metastatic castration-resistant prostate cancer.
Poster Session 1 | HSD15
Monday, November 9 | 10:30–13:30
Discussion period: 12:30–13:30
Cross-Country Differences in Real-World Laboratory Monitoring Intensity and Hepatotoxicity Adverse Event Detection Among Initiators of CDK4/6 Inhibitors for Advanced Breast Cancer
This study examines how differences in laboratory monitoring across the UK, Germany, and Japan may influence observed liver toxicity rates and the interpretation of cross-country safety evidence.
Poster Session 5 | RWD173
Wednesday, November 11 | 09:00–11:30
Discussion period: 09:00–10:00
About Flatiron Health
Flatiron Health is a healthtech company expanding the possibilities for point of care solutions in oncology and using data for good to power smarter care for every person with cancer. Through machine learning and AI, real-world evidence, and breakthroughs in clinical trials, we continue to transform patients’ real-life experiences into knowledge and create a more modern, connected oncology ecosystem. Flatiron Health is an independent affiliate of the Roche Group.
Contact
Nina Toor
press@flatiron.com