MULTI-COUNTRY HEOR LANDSCAPE ASSESSMENT FOR MULTIPLE SCLEROSIS USING A MULTI-AGENT GENAI SYSTEM
Author(s)
Sumeyye Samur, PhD1, Jag Chhatwal, PhD2, Mine Tekman, PhD1, Ismail Fatih Yildirim, MSc1, Klas Bergenheim, PhD3.
1Value Analytics Labs, Boston, MA, USA, 2Associate Professor and Director of the Institute for Technology Assessment, Harvard Medical School / Massachusetts General Hospital, Boston, MA, USA, 3AstraZeneca, Molndal, Sweden.
1Value Analytics Labs, Boston, MA, USA, 2Associate Professor and Director of the Institute for Technology Assessment, Harvard Medical School / Massachusetts General Hospital, Boston, MA, USA, 3AstraZeneca, Molndal, Sweden.
OBJECTIVES: Generative AI can accelerate HEOR and market access tasks, but commonly used models often lack the accuracy, depth, and traceability required for rigorous evidence synthesis. We used a HEOR-specific multi-agent AI system, ValueGen.AI, to address these limitations. This study evaluated the system’s ability to autonomously generate a comprehensive, multi-country HEOR landscape assessment for multiple sclerosis (MS).
METHODS: We used a hierarchical “Deep Agent” architecture using the LangGraph framework to orchestrate over 1,000 specialized sub-agent invocations across three layers: 1) A Main Orchestrator decomposing queries and evaluating information sufficiency; 2) A Deep Agent Layer executing parallel retrieval; and 3) A Tool Layer utilizing the Model Context Protocol (via FastMCP) to interface with heterogeneous data sources (NICE, G-BA/IQWiG, ICER, INAHTA, PubMed, clinical trial registries, and web repositories). The platform was tasked to generate a landscape report for MS covering epidemiology; clinical management across all MS phenotypes (RRMS, SPMS, PPMS); outcomes evidence; economic evidence; payer and reimbursement environment; competitor and pipeline analysis; stakeholder mapping; and evidence gaps, spanning US, UK, Germany, France, Italy, Spain, and China. Content and references were verified via human review.
RESULTS: ValueGen.AI generated a 300-page, seven-country HEOR landscape assessment for MS in under 48 hours. The report synthesized over 380 verifiable references across 12 sections with 126 country-specific subsections. Key findings included cross-country prevalence estimates ranging from 2.2 per 100,000 (China) to 335 per 100,000 (Germany); over 15 approved DMTs with anti-CD20 antibodies dominating the approximately $16.1 billion global market; a late-stage pipeline centered on BTK inhibitors with fenebrutinib showing positive Phase 3 results; and progressive MS as the principal unmet need.
CONCLUSIONS: Multi-agent AI workflows can dramatically improve HEOR productivity by compressing months of work into days. This study demonstrates that structured AI architectures can perform complex, multi-country HEOR landscape assessments without compromising the scientific rigor and traceability essential for decision-making.
METHODS: We used a hierarchical “Deep Agent” architecture using the LangGraph framework to orchestrate over 1,000 specialized sub-agent invocations across three layers: 1) A Main Orchestrator decomposing queries and evaluating information sufficiency; 2) A Deep Agent Layer executing parallel retrieval; and 3) A Tool Layer utilizing the Model Context Protocol (via FastMCP) to interface with heterogeneous data sources (NICE, G-BA/IQWiG, ICER, INAHTA, PubMed, clinical trial registries, and web repositories). The platform was tasked to generate a landscape report for MS covering epidemiology; clinical management across all MS phenotypes (RRMS, SPMS, PPMS); outcomes evidence; economic evidence; payer and reimbursement environment; competitor and pipeline analysis; stakeholder mapping; and evidence gaps, spanning US, UK, Germany, France, Italy, Spain, and China. Content and references were verified via human review.
RESULTS: ValueGen.AI generated a 300-page, seven-country HEOR landscape assessment for MS in under 48 hours. The report synthesized over 380 verifiable references across 12 sections with 126 country-specific subsections. Key findings included cross-country prevalence estimates ranging from 2.2 per 100,000 (China) to 335 per 100,000 (Germany); over 15 approved DMTs with anti-CD20 antibodies dominating the approximately $16.1 billion global market; a late-stage pipeline centered on BTK inhibitors with fenebrutinib showing positive Phase 3 results; and progressive MS as the principal unmet need.
CONCLUSIONS: Multi-agent AI workflows can dramatically improve HEOR productivity by compressing months of work into days. This study demonstrates that structured AI architectures can perform complex, multi-country HEOR landscape assessments without compromising the scientific rigor and traceability essential for decision-making.
Conference/Value in Health Info
2026-11, ISPOR Europe 2026, Vienna, Austria
Value in Health, Volume 29, Issue 12S
Code
P19
Topic
Methodological & Statistical Research
Topic Subcategory
Artificial Intelligence, Machine Learning, Predictive Analytics
Disease
Neurological Disorders, No Additional Disease & Conditions/Specialized Treatment Areas