atomCAT2 - Anal cancer outcome modelling using distributed learning
Research type
Research Study
Full title
atomCAT2 - A multicentre study of overall survival, locoregional control and distant metastasis in anal cancer utilising distributed learning
IRAS ID
303103
Contact name
Stelios Theophanous
Contact email
Sponsor organisation
University of Leeds
Duration of Study in the UK
2 years, 0 months, 1 days
Research summary
Anal cancer is a rare disease, where individual hospitals typically only treat a few patients per year. In order to improve and personalise treatment, we need to better understand the factors which determine how well patients do after treatment. Due to anal cancer’s rarity, it has been hard to establish and access sufficiently large datasets to explore this. The published evidence primarily consists of single-institution studies of small numbers of patients treated over multiple decades, often using outdated treatment regimens and techniques for many patients. Results of these studies are not representative for patients treated today.
Performing larger studies have traditionally depended on sharing data between hospitals. However, this comes with data protection and patient privacy concerns. "Distributed learning" is an approach which may alleviate this, while still facilitating multi-centre studies. "Distributed learning" is an analysis technique which allows us to learn from patients treated at many different institutions, without ever exchanging individual patient-level data - only summary statistics will be shared. In this way, patient data privacy will always be preserved. Ultimately, the insight gained from such analyses could help us design better trials for anal cancer and improve treatment for future patients.
Specifically, this study aims to develop and validate prediction models for patients with anal cancer treated with modern radiotherapy techniques, using distributed learning, as well as identify key prognostic factors and their effect size. We will have data from patients treated in 10+ international centres, with the aim of analysing a cohort of at least 800 patients. This will make for the largest available cohort of patients treated with up-to-date radiotherapy, and will provide unique insights into outcomes and prognostic factors for anal cancer patients. This application concerns use of data in three participating UK centres.
REC name
Wales REC 7
REC reference
22/WA/0081
Date of REC Opinion
8 Mar 2022
REC opinion
Favourable Opinion