Designing a Targeted Lung Cancer (LC) Screening Trial in the U.S. Screening-Ineligible Population Using Synthetic Cohorts and Natural History Modelling
Principal Investigator
Name
Thomas Trikalinos
Degrees
M.D., Ph.D.
Institution
Brown University
Position Title
Professor and Director
About this CDAS Project
Study
PLCO
(Learn more about this study)
Project ID
PLCO-2047
Initial CDAS Request Approval
Jun 1, 2026
Title
Designing a Targeted Lung Cancer (LC) Screening Trial in the U.S. Screening-Ineligible Population Using Synthetic Cohorts and Natural History Modelling
Summary
We are developing a mathematical simulation model of the natural history of lung cancer among US adults aged 50-80 who smoke <20 pack years or who quit smoking more than 15 years ago with the goal of informing the design of a screening trial comparing screening based on novel biomarkers versus usual care. Towards this end, we want to use PLCO data to (i) characterize the joint distribution of key LC risk factors and (ii) generate a predictive model of LC risk based on key risk factors. We plan to report the analyses we propose here as supporting or prior analyses in manuscripts describing the mathematical simulation model or its derivative works.
Emerging liquid biopsies, including LC-specific assays and multi-cancer early detection (MCED) tests that analyze circulating cell-free DNA and related biomarkers, are increasingly used in oncology for tumor genotyping and disease monitoring/surveillance, but they are not currently recommended as a population screening test for asymptomatic LC, which relies on annual low-dose CT. Liquid biopsies have been evaluated in screening-like cohorts demonstrating feasibility and very high specificity in some settings, though evidence is still evolving. They could improve LC screening by offering a minimally invasive option that may increase participation, help risk-stratify individuals who are currently screening-ineligible, and potentially triage follow-up imaging.
This project will support a broader effort to design a large RCT to evaluate screening strategies with liquid biopsies to detect LC among 50-80 year olds in the US population who are not currently covered by the USPSTF-2021, namely, never smokers, current and former smokers with a lifetime-history of less than 20 pack years, or former smokers with 20+ pack years but who quit more than 15 years ago. For convenience, we henceforth refer to this population as the US screening-ineligible population. Such a trial has the potential to impact clinical guidelines and change clinical practice.
Aims
To support the overarching objective of designing a large RCT, the aim of this project is to develop a mathematical simulation model of LC risk factors and natural history to generate synthetic cohorts of screening-ineligible individuals in the U.S and use it to predict outcomes, e.g., incidence and stage distribution at detection, under hypothetical LC screening strategies with liquid biopsies and the status quo.
Achieving this aim will require completing two tasks that are necessary for downstream analyses.
Tasks requiring PLCO data as inputs:
1. Characterize the multivariate distribution of risk factors in the US screening-ineligible population
2. Characterize LC incidence (overall and by stage) across subgroups defined by the above risk factors.
Collaborators
THOMAS TRIKALINOS Brown University
Jonah Popp Brown University
Yuliia Sereda Brown University