David Pesántez, Guest Lecturer at Centro de Estudios Biosanitarios – CEB, Medical Director, Early Global Development, Oncology R&D at AstraZeneca, shared a post on LinkedIn:
“Excited to share our latest publication in ESMO Real World Data and Digital Oncology:
‘Development and Evaluation of a Large Language Model-Based, Retrieval-Augmented Generation Application for Query Response in Early Oncology Clinical Trials.’
Clinical trial protocols have become extraordinarily complex — hundreds of pages of documents that research teams must navigate every day, under pressure, in real time.
To address this challenge, we developed the Study Document Assistant (SDA), a Retrieval-Augmented Generation (RAG) platform designed to provide rapid, evidence-grounded answers directly from clinical trial documentation.
In a controlled study involving sponsor clinical trial personnel, SDA demonstrated:
- 43.8% reduction in query response time
- Faster resolution of protocol-related questions
- Preserved answer quality and accuracy, as independently confirmed by blinded subject matter experts and AI-based evaluations
- High user satisfaction and successful implementation
Beyond the technology itself, what excites me most is the potential impact on clinical research operations. Every minute spent searching across complex study documents is time that cannot be invested in advancing research and supporting patients.
Our findings suggest that thoughtfully designed AI systems can help clinical teams access critical information more efficiently while maintaining scientific rigor, transparency, and traceability.
This work represents an important step toward integrating trustworthy AI solutions into oncology drug development and clinical trial execution.
I am proud to have collaborated with an outstanding multidisciplinary team on this project and excited about the potential of trustworthy AI solutions to accelerate drug development and ultimately improve outcomes for patients.”
Title: Development and evaluation of a large language model-based, retrieval-augmented generation application for query response in early oncology clinical trials
Authors: D. Pesántez 1, G. Fucà, A. Magrì, L. Koulai, K. Bliznashki, A. Soltani, M. Khan, F. Multari, M. Moschetta
Other articles featuring David Pesántez on OncoDaily.