Using population-based healthcare data to investigate the risk of mortality or relapse after radiotherapy for prostate cancer
Principal Investigator
Name
Bill Nailon
Degrees
PhD
Institution
University of Edinburgh
Position Title
Clinical Scientist
Email
About this CDAS Project
Study
PLCO
(Learn more about this study)
Project ID
PLCO-1601
Initial CDAS Request Approval
Jun 24, 2024
Title
Using population-based healthcare data to investigate the risk of mortality or relapse after radiotherapy for prostate cancer
Summary
To date, methods to personalise an individual’s radiotherapy treatment have been based on biological information and developments in technology. In general, little or no account is taken of the vast array of data that is available for each patient within their healthcare record, the extensive digital imaging data collected at diagnosis, and data collected during a course of radiotherapy. In Scotland, I am involved in the PROSECCA project (Improving radiotherapy in PROState cancer using EleCtronic population-based healthCAre data), which aims to use artificial intelligence (AI) to analyse this data from 10,000+ prostate cancer patients and in doing so identify new scientific relationships between a patient’s medical history and how well they respond to radiotherapy. By establishing what factors in a patient’s complex healthcare record indicate that they may have a poor response to treatment, or an increased risk of side effects from radiation, it will be possible to identify these patients earlier and adapt their treatment accordingly.
In advance of the full Scottish data set arriving later this year, which will take time to curate for analysis, I would like to access the PLCO data to investigate the impact of a patient's clinical and demographic characteristics on the risk of mortality or biochemical relapse in prostate cancer patients post radiotherapy, through building predictive AI and ML models. The PLCO trial contains similar data to the PROSECCA project and the opportunity to use it will significantly benefit my understanding of using data of this nature. Utilising the PLCO will help build approaches and workflows that will be used with the Scottish PROSECCA data to allow us to improve radiotherapy treatment for prostate cancer patients in the future. The PLCO data will help identify features (e.g. demographic characteristics, comorbidities, etc.) of clinical significance in relation to the occurrence of a poor outcome, which may help direct the approaches used for analyses going forward.
In advance of the full Scottish data set arriving later this year, which will take time to curate for analysis, I would like to access the PLCO data to investigate the impact of a patient's clinical and demographic characteristics on the risk of mortality or biochemical relapse in prostate cancer patients post radiotherapy, through building predictive AI and ML models. The PLCO trial contains similar data to the PROSECCA project and the opportunity to use it will significantly benefit my understanding of using data of this nature. Utilising the PLCO will help build approaches and workflows that will be used with the Scottish PROSECCA data to allow us to improve radiotherapy treatment for prostate cancer patients in the future. The PLCO data will help identify features (e.g. demographic characteristics, comorbidities, etc.) of clinical significance in relation to the occurrence of a poor outcome, which may help direct the approaches used for analyses going forward.
Aims
- To evaluate the utilisation of radiotherapy as a prostate cancer treatment within this popualation, compared to other treatment options
- To predict the probability of mortality after radiotherapy for prostate cancer patients
- To predict the probability of biochemical relapse after radiotherapy for prostate cancer patients
- To identify the important factors from a patients healthcare history that may influence radiotherapy response
Collaborators
Zhuolin Yang (University of Edinburgh)
Professor Bill Nailon (University of Edinburgh)