Risk assessment of individuals with screen-detected lung nodules
Principal Investigator
Name
Stephen Duffy
Degrees
BSc MSc CStat
Institution
Queen Mary University of London
Position Title
Professor of Cancer Screening
Email
About this CDAS Project
Study
NLST
(Learn more about this study)
Project ID
NLST-1007
Initial CDAS Request Approval
Jan 12, 2023
Title
Risk assessment of individuals with screen-detected lung nodules
Summary
A systematic review and meta-analysis of predictive factors for lung cancer in patients with lung nodules identified by low-dose CT screening will be performed as a first step of this project. The MEDLINE and Embase databases will be searched from the year 2000 onwards. Studies of patients with LDCT-screen detected nodules and LC outcomes will be included. Studies only including predicted or simulated data will be excluded. It is anticipated that a major contribution to this will be the intervention groups of randomised LDCT screening trials, but observational follow-up studies of patients with screen-detected lung nodules will also be included. A data extraction framework will be developed to ensure consistent extraction across studies. Individual-level data, both nodule-specific and patient-specific, will be used where available. Both logistic and time-to-event analyses will be carried out where possible. Meta-analysis will be performed using random effects models. Heterogeneity will be assessed using the I² statistic. The PROBAST (Prediction model Risk Of Bias ASsessment Tool) will be used to assess study quality. The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines will be followed.
As a subsequent step, a model which predicts risk of lung cancer in individuals with screening detected lung nodules will be built. A multivariable Cox regression model or a multivariable logistic regression model will be built, depending on whether time-to-event outcome variables have been successfully obtained. The factors identified in the systematic review will comprise the potential predictor variables in the model.
As a result of this project, it will be possible to more accurately identify the subset of screenees who are safe to be left without any further investigation or surveillance until the next screening round. The meta-analysis team will be advised and overseen by a group of multidisciplinary collaborators who are active in LC-screening research.
As a subsequent step, a model which predicts risk of lung cancer in individuals with screening detected lung nodules will be built. A multivariable Cox regression model or a multivariable logistic regression model will be built, depending on whether time-to-event outcome variables have been successfully obtained. The factors identified in the systematic review will comprise the potential predictor variables in the model.
As a result of this project, it will be possible to more accurately identify the subset of screenees who are safe to be left without any further investigation or surveillance until the next screening round. The meta-analysis team will be advised and overseen by a group of multidisciplinary collaborators who are active in LC-screening research.
Aims
- To identify risk factors predictive of LC, which are known at the time of the scan, in patients with LDCT screen-detected lung nodules
- To more accurately identify the subset of screenees who are safe to be left without any further investigation or surveillance until the next screening round
Collaborators
Professor Harry de Koning, Erasmus MC
Professor John Field, University of Liverpool
Dr Christine Berg, US NCI (retired)
Dr Arjun Nair, University College London Hospitals NHS Foundation Trust
Professor Robert Rintoul, University of Cambridge
Professor Sam Janes, UCL
Dr Phil Crosbie, University of Manchester
Professor Matthew Callister, Leeds Teaching Hospitals
Dr Mark Hammer, Harvard Medical School
Dr Carolyn Horst, Kings College London