Cambridge startup inspires AI-led approach to cancer trials
Cambridge Cancer Genomics is leveraging its AI platform in a new partnership designed to more accurately target tumours in precision clinical trials. It is working with Precision-Panc, a UK-wide network of over 20 hospitals that can offer such trials to patients with pancreatic cancer.
The broad-based collaboration sets out to improve and standardise data analysis, distribution and sharing in platform these vital diagnostic programmes.
Startup CCG.ai is building software for data-driven precision oncology and its technology forms the basis of the piloted standardisation infrastructure and will provide insights into Precision-Panc’s large-scale, multi-centre, oncology trials.
CCG.ai’s platform – OncOS – automates the analysis of cancer DNA samples and matches suitable clinical trials based on the underlying molecular features of a tumour, simplifying and streamlining data sharing to accelerate translation into the clinic.
The company will also use machine learning to help uncover complex biomarkers of response to individual, or combinations of, therapies. This builds on work published at NeurIPS 2018 and NeurIPS 2019, outlining the company’s ML-led advances in identifying cancer mutations and predicting tumour growth.
The partnership specifically aims to build on and improve Precision-Panc’s adaptive clinical trials by incorporating CCG.ai’s AI technologies to help test and predict therapy response and building a framework to share and distribute data and analytics.
If successful, this pilot could form the framework for clinical and genomic data sharing throughout the world and validation of both Precision-Panc as the gold-standard platform for clinical trials and for CCG.ai’s clinical decision support tools.
Dr Harry Clifford, CTO of CCG.ai, who will lead the project says: “Whilst cancer is one of the greatest challenges facing healthcare systems around the world, and cancer-genomics data is hugely rich and widespread, there has been surprisingly little progress on applying AI based methods to the oncology clinical trial space.
“This collaboration is the first step to using AI-supported clinical decision support in cancer care and a huge step forward for the design of clinical trials.”
Work presented at CSBIO 2019 showcased preliminary results from the partnership, with CCG.ai presenting a new workflow for more accurate identification of cancer sub-types, and better analysis of raw data, on publicly available pancreatic cancer datasets.
This work will be built on by applying CCG.ai’s data analysis techniques to Precision-Panc’s rich clinical trial datasets – ultimately leading to development of AI-led analytics and frameworks for use in clinical trials.
Dr John Cassidy, CEO of CCG.ai says: “As next generation sequencing becomes an ever more important part of the standard clinical workflow, ensuring data standardisation and ease of data sharing amongst accredited researchers is of paramount importance.
“We are thrilled to be working with Precision-Panc on this project; they are among the world leaders in platform clinical trials.”