Determinants of serum per- and polyfluoroalkyl substances (PFAS) in the Prostate, Lung, Colorectal and Ovarian Cancer (PLCO) Screening Trial

Principal Investigator

Name
Rena Jones

Degrees
Ph.D., M.S

Institution
National Cancer Institute

Position Title
Senior Investigator

Email
rena.jones@nih.gov

About this CDAS Project

Study
PLCO (Learn more about this study)

Project ID
PLCO-2089

Initial CDAS Request Approval
Sep 28, 2026

Title
Determinants of serum per- and polyfluoroalkyl substances (PFAS) in the Prostate, Lung, Colorectal and Ovarian Cancer (PLCO) Screening Trial

Summary
We are proposing an investigation of the determinants (e.g., demographic, dietary, lifestyle, temporal, geographic, reproductive, clinical, and biospecimen-related factors) of PFAS within a large population-based cohort. While PFAS contamination remains a growing public health concern, determinants of the variability in PFAS exposures in the U.S. are poorly understood.

Several nested case-control studies of cancer in PLCO have already measured PFAS concentrations: perfluorooctanesulfonic acid (PFOS), (2-(N-methyl-perfluorooctane sulfonamido) acetic acid, 2-(N-ethyl-perfluorooctane sulfonamide) acetic acid, perfluorohexane sulfonic acid (PFHxS), perfluoromethylheptanesulfonic acid (Sm-PFOS), perfluorooctanoic acid (PFOA), perfluorononanoic acid (PFNA), perfluorodecanoic acid (PFDA), and perfluoroundecanoic acid (PFUnDA) in pre-diagnostic serum samples. The participants for this proposed investigation will be sourced from the controls in these studies, including ovarian and endometrial (N=472 unique controls), renal (324), thyroid (378), and prostate (750) cancers and non-Hodgkin lymphoma (706). We propose to leverage available measurements from these prior studies as an efficient way to investigate determinants of serum PFAS concentrations, particularly across different geographic locations, without depleting biospecimen samples through new laboratory analyses. We will also evaluate composite scores summarizing dietary and lifestyle patterns to assess their contribution to serum PFAS concentrations. Moreover, pooling the controls from these studies will enhance statistical power for analyses stratified by participant characteristics that have not been possible before. We will also consider incorporating cases from each cancer study: ovarian (318), endometrial (429), renal (324), thyroid (189), prostate (750), and non-Hodgkin lymphoma (706) to increase the available sample size and leverage their pre-diagnostic PFAS measurements, using analytic methods that appropriately account for the underlying case-control sampling design.

Multiple linear regression analysis will be used to identify significant predictors of continuous, quartile, and ranked (e.g., high vs. low) PFAS serum concentrations. Determinants will be derived from self-reported survey data (e.g., demographics, dietary, lifestyle, reproductive factors) and secondary data linkages (e.g., urban/rural classification, drinking water source, residential proximity to military bases, airports, and industrial discharge sites). All models will include a covariate reflecting the study of origin for the control participant. Serum PFAS concentrations were measured by the CDC for most studies, although analyses for the NHL and thyroid studies were conducted by the Finnish National Institute of Health and Welfare. To account for laboratory differences, we anticipate sensitivity analyses adjusting for lab or excluding samples measured by the Finnish lab and potentially performing a meta-regression analysis across studies.

Three additional PLCO studies of breast (N=621 controls), pancreatic (360), and liver (175) cancers, relied on PFAS derived from metabolomic signatures measured by Metabolon, Inc. to assess PFOA and PFOS. Metabolon and CDC measures were highly correlated (p>0.70), indicating they could be informative to relative ranking of exposures. Therefore, we propose to also include controls from these three studies in determinants analysis limited to predicting high vs. low exposures to PFOA and PFOS; in addition to a sensitivity analysis where we will conduct a separate pooled analysis restricted to the three metabolomic studies. Again, we will also consider incorporating cases from each cancer study: breast (N=621), pancreatic (360), and liver (175) to leverage the pre-diagnostic samples.

Aims

Aim 1: Evaluate demographic, dietary, lifestyle, temporal, geographic, reproductive, clinical, and biospecimen-related determinants of serum PFAS concentrations, including the influence of composite indices that capture broader dietary and lifestyle patterns, among participants in the Prostate, Lung, Colorectal and Ovarian Cancer Screening Trial.

Aim 2: Assess the comparability of case and control groups across contributing nested case-control studies within the Prostate, Lung, Colorectal and Ovarian Cancer Screening Trial by systematically evaluating differences in study eligibility criteria, matching factors, participant characteristics, specimen-collection periods, biospecimen characteristics, laboratory platforms, and PFAS measurement methods that may contribute to heterogeneity in measured PFAS concentrations across studies.

Collaborators

Jared Fisher National Cancer Institute
Vicky Chang National Cancer Institute
Rena Jones National Cancer Institute
Eun Mi Jung National Cancer Institute
Wen-Yi Huang National Cancer Institute
Steve Moore National Cancer Institute
Katherine McGlynn National Cancer Institute
Jongeun Rhee National Cancer Institute
Rachael Stolzenberg-Solomon National Cancer Institute
Sonja Berndt National Cancer Institute
Luz Villanueva National Cancer Institute
Jonathan Hofmann National Cancer Institute
Mark Purdue National Cancer Institute