Developing an All-cancer Risk Model

Principal Investigator

Name
Paul Pinsky

Degrees
Ph.D.

Institution
National Institutes of Health

Position Title
Chief, EDRG

Email
pp4f@nih.gov

About this CDAS Project

Study
PLCO (Learn more about this study)

Project ID
PLCO-2040

Initial CDAS Request Approval
Apr 13, 2026

Title
Developing an All-cancer Risk Model

Summary
Multi-cancer detection tests are a new paradigm in screening. There is a need to identify persons at high risk for any cancer. We will use PLCO questionnaire data to develop models of cancer incidence and cancer mortality to identify population subsets at high risk for cancer incidence or mortality. We will use the baseline questionnaire, supplemental questionnaire and dietary questionnaires.
We will develop logistic regression models for cancer incidence, with different time frames, as well as for cancer mortality. Statistical methods will be used to build the optimal models. We will compute AUCs of the models as well as other statistics.

Aims

Develop a model of all-cancer incidence
Develop a model of all-cancer mortality
Develop models of all-cancer incidence for a selected group of cancer types
Develop models of all-cancer mortality for a selected group of cancer types

Collaborators

Paul Pinsky National Institutes of Health
Eric Miller National Institutes of Health
Elyse Leevan National Institutes of Health