Abstract
Background
Perfluoroalkyl substances (PFASs) are possible carcinogens with widespread population exposure. Mammographic density (MD) is a biomarker of breast cancer risk that may offer important insights into potential mechanisms of breast carcinogenesis. Examining associations between PFAS exposure and MD may elucidate pathways by which PFASs could affect breast cancer risk. However, associations between PFAS exposure and MD remain poorly studied. We sought to evaluate associations of individual PFASs and their chemical mixture with percent mammographic density and dense breast area.
Methods
We analyzed data from 186 postmenopausal, cancer-free women not currently using hormone therapy from the Susan G. Komen Tissue Bank. Percent mammographic density (PMD) and dense breast area (DBA) were measured from digital mammography images taken near the time of blood sample donation. Serum PFAS concentrations were measured using online solid-phase extraction followed by high-pressure liquid chromatography with tandem mass spectrometry (HPLC-MS/MS). Associations were examined using multivariable linear regression and Quantile g-computation.
Results
Participants had a mean age of 58.7 years and mean body mass index of 29.6 kg/m², and most were White (72.6%). The vast majority of mammograms were performed within 2 years of blood sample donation (88%). No statistically significant associations were observed with individual PFASs and either PMD or DBA in analyses adjusted for confounding variables. Quantile G-computation analyses did not identify a statistically significant effect of the PFAS mixture on PMD or DBA. Similar results were observed in analyses stratified by body mass index or breastfeeding history.
Conclusions
Overall, our study did not identify strong and significant associations between a panel of major PFASs and MD.
Supplementary Information
The online version contains supplementary material available at https://doi.org/10.1186/s13058-026-02340-6.
Keywords: Per- and polyfluoroalkyl substances (PFASs), Mammographic density (MD), Postmenopausal women
Background
Per- and polyfluoroalkyl substances (PFASs) are man-made chemicals frequently used in commercial and industrial products. Due to their widespread presence and persistence in the environment, nearly 98% of the U.S. population has measurable PFASs in their blood. [1] Humans are exposed to PFASs through multiple pathways, including through consumption of contaminated food products and drinking water. [2] The PFAS chemical class includes thousands of compounds, with perfluorooctanoic acid (PFOA) and perfluorooctane sulfonic acid (PFOS) the most studied. PFASs accumulate in human tissues and may have half-lives of many years, [3] prompting major concerns about their potential for long-term health effects. The International Agency for Research on Cancer (IARC) has classified PFOA as carcinogenic to humans, and PFOS as possibly carcinogenic. [4] However, epidemiologic studies exploring links between PFASs and breast cancer risk have been inconsistent, and a recent meta-analysis reported a small and non-significant increase in breast cancer risk associated with pre-diagnostic serum PFOA concentrations (HR 1.16, 95% CI 0.96–1.40) [5].
Examining associations between PFAS exposure and biomarkers of breast cancer risk, such as mammographic density (MD), may elucidate pathways by which PFASs could affect breast cancer risk. The extent of MD is determined by the amount of connective and epithelial tissue relative to fat tissue as viewed on a mammogram. [6,7] MD is strongly and consistently associated with about a four-fold increased breast cancer risk (comparing 4th vs. 1st MD quartiles) regardless of age, menopausal status, or race. [6] Laboratory data suggest that some PFASs activate the estrogen receptor (ER); [8–13] MD is well-established to be sensitive to other exposures activating ER (e.g. postmenopausal hormone therapy [HT] [14]). Based on this prior work, effects of PFASs on breast cancer risk, if any, may proceed via mechanisms that would be apparent in effects on MD. However, associations between PFASs and MD remain poorly studied. One prior study reported no significant associations between PFASs and MD among a population of postmenopausal women, [15] while a recent study in premenopausal women observed positive associations between two PFASs (PFHxS and PFOS) and MD. [16] Further research is needed to address these inconsistent findings.
We evaluated associations between PFASs and MD in a population of cancer-free, postmenopausal women who participated in the Komen Tissue Bank. Our analysis included measuring a panel of seven PFASs and utilizing statistical approaches for analyzing mixtures of environmental chemicals.
Methods
Study population
The Susan G. Komen for the Cure Tissue Bank (KTB) was established in 2007 to profile the molecular histology of normal breast tissue. [17] Volunteer donors gave background health data, blood samples, and breast tissue specimen through biopsies. Eligible participants were biologically female, at least 18 years old, had no breast implants, no allergy to local anesthetics, were not taking blood thinners except aspirin, and had no history of chest radiation. Researchers can access publicly available participant data from the KTB through the Virtual Tissue Bank. (https://virtualtissuebank.iu.edu).
For this study, we included KTB participants who: (1) had no prior diagnosis of breast cancer or ductal carcinoma in situ, (2) were postmenopausal, (3) were not currently using hormone therapy, (4) had available serum sample, and (5) had the postmenopausal craniocaudal mammogram(s) available through the Virtual Tissue Bank. A total of 186 KTB participants met these criteria and were included in our analysis.
All participants gave written informed consent at the time of tissue donation. The KTB protocol was approved by the Institutional Review Board at Indiana University. Because this study used only deidentified data and previously collected samples, it was considered exempt from human subjects research by the Institutional Review Board at the University of Massachusetts Amherst.
PFASs measurement
At the time of breast tissue donation (2009–2012), venous blood samples were collected into 9 mL tubes, centrifuged at 2000 rcf for 15 min, and the serum fraction was aliquoted into multiple cryogenic vials and stored at − 80 °C until analysis. For this study, 750 µL serum aliquots from eligible KTB participants were retrieved and shipped on dry ice via overnight courier to the analytical laboratory. Serum PFASs concentrations were quantified at the Department of Clinical Pharmacology, Pharmacy, and Environmental Medicine, University of Southern Denmark. All laboratory personnel were blinded to participant information. Given that biological half-lives of PFAS are often 2–5 years, [18, 19] a single measurement of serum PFAS concentrations is generally viewed as representing exposure over this period.
A panel of seven PFAS compounds were analyzed using automated online solid-phase extraction followed by high-performance liquid chromatography coupled with tandem mass spectrometry (HPLC–MS/MS): [20] perfluorooctanoate (PFOA), perfluorooctane sulfonate (PFOS), perfluorohexane sulfonate (PFHxS), perfluoroheptane sulfonate (PFHpS), perfluorononanoate (PFNA), perfluorodecanoate (PFDA), and perfluoroundecanoate (PFUnDA). The limit of detection (LOD) was 0.03 ng/mL for all analytes. Values below the LOD were set to 0.015 ng/mL for all PFASs. The proportion of values below the LOD was low across PFASs: 0% for PFOA, PFNA, PFDA, and PFOS; 1% for PFHxS; 5% for PFHpS; and 6% for PFUnDA.
Analytical precision was assessed using quality control samples analyzed in each batch. Coefficients of variation (CVs) were 4.1% for PFOA, 9.1% for PFOS, 4.7% for PFNA, 6.3% for PFDA, 10.1% for PFHxS, 13.7% for PFHpS, and 5.5% for PFUnDA. Samples were randomly distributed across batches, and no significant batch effects were detected (ANOVA p > 0.9 for all PFASs). Method accuracy and interlaboratory comparability were ensured through participation in the German External Quality Assessment Scheme (G-EQUAS), organized by the University of Erlangen-Nuremberg.
Mammographic density (MD) measurement
KTB participants provided permission for KTB researchers to contact their mammogram facility to obtain recent and future mammogram images. We retrieved all processed digital mammograms available through the KTB Virtual Tissue Bank and selected the mammogram closest in time to the year of KTB donation. For some participants mammograms were available before and after the KTB donation with the same absolute interval of time between mammogram and KTB donation. In these cases, we preferentially selected the mammogram taken after KTB donation (e.g., if mammograms from one year before and one year after were available, the mammogram taken one year after KTB donation was selected). The craniocaudal view of the left breast was selected.
The majority of mammograms (N = 164, 88%) was taken ± 2 yrs from the time of blood and tissue donation. We measured PMD (PMD, %) and dense breast area (DBA, cm2) from the digital mammogram images using software developed by Laboratory for Individualized Breast Radiodensity Assessment (LIBRA). [21,22] Briefly, LIBRA utilizes computer algorithms to differentiate between the breast and pectoral muscle and then to identify areas of MD. LIBRA is appropriate for raw and processed mammographic images.
Covariate data
At the time of sample donation, additional information was collected from questionnaires completed by KTB participants. Participants self-reported information on age, race, education, tobacco use, pregnancy and live birth history, breastfeeding history, and postmenopausal HT use. A research nurse measured each participant’s height and weight during the enrollment visit to calculate body mass index (BMI).
Statistical analysis
Descriptive statistics were calculated as frequency and percentage for categorical variables and means and standard deviation (SD) for continuous variables. All PFASs and breast density variables were natural log-transformed to improve normality. The distribution of each PFAS was reported as the geometric mean, geometric standard deviation, and range. We calculated age-adjusted least-square mean PMD and DBA by quartile of each measured PFAS. Quartile cutpoints for PFASs are reported in Supplemental Table 1.
Based on prior knowledge and causal diagrams, we included the following variables as covariates in our analysis: age (continuous), race (White, non-White), BMI (continuous), ever pregnant (no, yes), parity (continuous), ever used tobacco (no, yes), ever breastfeed (no, yes), blood relatives with breast cancer (no, yes), time since last menstrual period (continuous), ever used postmenopausal HT (no, yes), time between blood sample donation and mammogram (continuous), and education (less than high school, associate’s/bachelor’s degree, graduate/professional degree).
We first examined associations between individual PFAS concentrations (continuous and categorized as quartiles) and both PMD and DBA using generalized linear regression models adjusting for age (Model A). Subsequent multivariable models additionally adjusted for covariates from the list above that were associated with percent density in bivariate models (Model B: race, years since menopause, ever use of postmenopausal HT, education, and time between blood sample and mammogram). A third multivariable model added adjustment for BMI (Model C). These analyses were conducted using SAS version 9.4 (SAS Institute, Cary, NC).
Additionally, we used Quantile G-computation to estimate the overall effect of the PFASs mixture separately on PMD and DBA, utilizing the R package qgcomp in R version 4.4.2. After adjusting for the previously mentioned covariates, we calculated the individual positive and negative contributions of each PFAS to the total mixture effect with the same sets of covariates (Model C) for both outcomes.
We repeated the above analyses restricted to parous women and then with stratification on breastfeeding history. Analyses also were repeated with stratification on BMI (< 25 kg/m2, ≥ 25 kg/m2). We also repeated analyses excluding participants (N = 22) whose blood sample was obtained > 2 years from the date of their mammogram.
Results
Descriptive statistics
A total of 186 postmenopausal women were included in the analysis, with an average age of 58.7 years (SD 7.8); most (72.6%) identified as White (Table 1). The mean BMI was 29.6 kg/m2. A majority (85.0%) had been pregnant, with average parity of 2.2 (SD 1.2) among those ever pregnant. Among parous women, 63.3% reported ever breastfeeding. One-third had previously used postmenopausal HT (33.9%), and 25.3% had a history of tobacco use. On average, participants were 12.8 years (SD 10.0) past menopause, and the average time between blood collection and mammogram was 0.3 years (SD 2.3). Moderate to strong positive correlations were observed between log-transformed PFAS concentrations (Supplemental Fig. 1). Correlations between PFNA and PFDA (r = 0.84), PFHpS and PFOS (r = 0.77), and PFDA and PFUnDA (r = 0.74) were especially high. Age-adjusted analyses (Table 2) identified no statistically significant associations between quartile of individual PFASs and PMD or DBA.
Table 1.
Baseline characteristics of study participants, selected from the Komen Tissue Bank (N = 186)
| Characteristics | |
|---|---|
| Age, years; Mean (SD) | 58.7 (7.8) |
| White; N (%) | 135 (72.6) |
| Body mass index, kg/m2; Mean (SD) | 29.6 (6.4) |
| Ever pregnant; N (%) | 158 (85.0) |
| Number live birthsa; Mean (SD) | 2.2 (1.2) |
| Ever breastfeda; N (%) | 100 (63.3) |
| Ever used postmenopausal hormone therapy; N(%) | 63 (33.9) |
| Ever used tobacco; N (%) | 47 (25.3) |
| Blood relatives with cancer; N (%) | 100 (53.8) |
| Education; N (%) | |
| Less than High School/High School | 60 (32.3) |
| Associate’s/Bachelor’s Degree | 64 (34.4) |
| Graduate/Professional Degree | 62 (33.3) |
| Time between KTB donation and last menstrual period, years; Mean (SD) | 12.8 (10.0) |
| Time between blood sample donation and mammogram, years; Mean (SD) | 0.3 (2.3) |
| Percent mammographic density, %; Geometric Mean (SD) | 15.1 (2.5) |
| Dense breast area, cm2; Geometric Mean (SD) | 24.6 (2.2) |
a Among those ever pregnant
Table 2.
Age-adjusted least-square mean (LSM) percent mammographic density and dense breast area by quartile of serum perfluoroalkyl substances concentrations (N = 186).
| Quartile 1 LSM (95%) | Quartile 2 LSM (95%) | Quartile 3 LSM (95%) | Quartile 4 LSM (95%) | P value | |
|---|---|---|---|---|---|
| Percent mammographic density, % | |||||
| PFDA | 21.6 (15.4–27.7) | 21.2 (15.2–27.3) | 23.8 (17.7–29.8) | 23.3 (17.1–29.5) | 0.92 |
| PFHpS | 24.3 (18–30.7) | 23.5 (17.4–29.5) | 21.7 (15.6–27.7) | 20.4 (14.1–26.8) | 0.85 |
| PFHxS | 21.7 (15.3–28.1) | 21 (15–27.1) | 26.1 (20.1–32.1) | 21 (14.5–27.4) | 0.59 |
| PFNA | 25.3 (19–31.5) | 21.8 (15.8–27.8) | 20.9 (14.9–27) | 21.9 (15.8–28.1) | 077 |
| PFOA | 21.4 (15.2–27.5) | 23.5 (17.4–29.5) | 21.9 (15.8–27.9) | 23.2 (17–29.4) | 0.96 |
| PFOS | 21.1 (14.9–27.3) | 25.9 (19.9–31.9) | 20.1 (14.1–26.1) | 22.8 (16.4–29.1) | 0.55 |
| PFUnDA | 16.9 (10.8–23) | 25.5 (19.5–31.5) | 23.4 (17.5–29.4) | 23.9 (17.9–30) | 0.22 |
| Dense breast area, cm 2 | |||||
| PFDA | 34.8 (24.4–45.1) | 32.1 (21.9–42.2) | 39.2 (29.1–49.3) | 29.6 (19.2–40.1) | 0.61 |
| PFHpS | 35.6 (24.9–46.2) | 40.3 (30.1–50.4) | 27.2 (17–37.4) | 32.7 (22.1–43.3) | 0.36 |
| PFHxS | 35.1 (24.3–46) | 36.2 (26–46.4) | 36.5 (26.3–46.7) | 27.8 (17–38.7) | 0.64 |
| PFNA | 38.8 (28.2–49.3) | 32.6 (22.4–42.7) | 34.1 (23.9–44.3) | 30.3 (19.9–40.7) | 0.72 |
| PFOA | 34.4 (24.1–44.7) | 33.6 (23.4–43.7) | 37.7 (27.5–47.8) | 30 (19.6–40.5) | 0.78 |
| PFOS | 32 (21.5–42.4) | 41.6 (31.5–51.7) | 30.9 (20.8–41) | 31.2 (20.5–41.8) | 0.39 |
| PFUnDA | 30.9 (20.5–41.3) | 38.7 (28.6–48.9) | 32.5 (22.4–42.7) | 33.5 (23.2–43.8) | 0.73 |
a Quartile cutpoints for PFAS are reported in Supplemental Table 1
Multivariable linear regression analysis
Overall, individual PFAS were not significantly associated with PMD (Fig. 1 & Supplemental Table 2). While PFUnDA concentrations were positively associated with PMD in the age-adjusted model (Model A), the associations were attenuated and non-significant when adjusted for confounders (Models B and C).
Fig. 1.
Estimated regression coefficient and 95% CI for associations between PFAS and natural log-transformed percent mammographic density estimated using multivariable generalized linear regression models (N = 186).a Model A: Adjusted for age; Model B: adjusted for age race, years since menopause, ever use of postmenopausal hormone therapy, education, time between blood sample and mammogram; Model C: adjusted for Model B + body mass index
No associations were observed between any of the PFASs and DBA (Fig. 2).
Fig. 2.
Estimated regression coefficient and 95% CI for associations between PFAS and log-transformed dense breast area estimated using multivariable generalized linear regression models (N = 186). a Model A: Adjusted for age; Model B: adjusted for age race, years since menopause, ever use of postmenopausal hormone therapy, education, time between blood sample and mammogram; Model C: adjusted for Model B + body mass index
Results were similar in analyses restricted to parous women, stratified by breastfeeding history, stratified by BMI category, and excluding observations where the interval between blood sample and mammogram was > 2 years.
Quantile-G computation analysis
Among the full study population, Quantile G-computation analysis indicated no overall PFASs mixture effect for either density measure (Table 3). For PMD, the mixture effect was β = − 0.04 (95% CI: − 0.17, 0.10), and for DBA the overall mixture effect was β = − 0.02 (95% CI: − 0.14, 0.11). PFNA and PFHxS contributed most positively to the mixture effect, while PFOS and PFUnDA contributed most negatively. These findings remained consistent when restricted to parous women and with stratification on breastfeeding history, stratifying on BMI category, or restricting analyses to participants with less than two years between blood collection and mammogram (Supplemental Table 4). Among parous women who had never breastfed, there was a suggestion of a negative association between the PFAS mixture and DBA (β = − 0.24; 95% CI: − 0.52, 0.04; p = 0.09), though this association was not statistically significant.
Table 3.
Summary of estimates of overall mixture effect and individual contribution to the overall mixture effect for each outcome using Quantile-G computation (N = 186) a
| Percent mammographic density | Dense breast area | |
|---|---|---|
| Overall mixture effect beta (95% CI); p | ||
| − 0.04 (− 0.17, 0.10); 0.61 | − 0.02 (− 0.14, 0.11); 0.81 | |
| Individual weights and direction of effect | ||
| PFDA | 0.16 | 0.30 |
| PFOA | 0.002 | − 0.09 |
| PFOS | − 0.73 | − 0.56 |
| PFHxS | − 0.005 | 0.11 |
| PFNA | 0.33 | 0.25 |
| PFUnDA | − 0.26 | − 0.35 |
| PFHpS | 0.50 | 0.36 |
a Adjusted for age, race, years since menopause, ever use of postmenopausal hormone therapy, education, time between blood sample and mammogram, and body mass index (Model C)
Discussion
In this study of 186 postmenopausal women, we found limited evidence for associations between serum PFASs concentrations and MD. No statistically significant associations were observed with individual PFASs, and Quantile G-computation analyses did not identify a statistically significant effect of the PFASs mixture on either measure of MD.
We carefully considered the role of BMI in our analysis. While BMI may act as a confounder, there also is some concern that BMI could represent a prior or parallel step on a causal pathway linking PFASs to MD; substantial evidence supports positive associations between PFAS and BMI [23] as well as negative associations between BMI and PMD. [24, 25] We fit models to estimate associations both with and without adjustment for BMI as well as models stratifying on BMI. Our results were robust across these different models. Including or excluding BMI from the models did not meaningfully change the estimates for most of the PFASs, suggesting that BMI did not substantially mediate or confound the association between PFAS exposure and the outcome.
Given that prior work has reported negative associations between circulating PFASs concentrations and breastfeeding duration, [26, 27] we also carefully considered the role of parity and breastfeeding history in our analysis. Our results were generally consistent in models with and without adjustment for parity and breastfeeding as well as in models restricted to parous women and with stratification on breastfeeding status. However, we observed a suggested negative association between the PFAS mixture and DBA among women who had never breastfed.While statistical power was insufficient for analysis within this small subgroup (N = 58), this result is consistent with our recent analysis in the same parent population in which we observed a suggested positive association between the PFASs mixture and breast involution among women who had never breastfed. [28] Our measure of breast involution also is negatively associated with breast cancer risk. [29] Further research is needed to confirm and fully understand these surprising results.
Our overall null results are similar to those observed within a population of 116 non-Hispanic white postmenopausal women participating in a substudy of the California Teachers Study. [15] That study reported no associations between a panel of 8 PFASs with neither PMD nor DBA, including no effect modification by either parity or BMI. While our study population was younger, on average, the time periods of sample donation were very similar. Together, our results and the results of this prior study do not support important associations between PFASs exposure and MD in postmenopausal women. However, a recent study in a large (N = 705) population of premenopausal women reported positive associations between PFOS and PFHxS and volumetric PMD. [16] Whether the discordance in results is due to the difference in menopausal status and/or the use of volumetric measurements, or some other factor, is unclear. However, further research will be useful in understanding the associations, if any, between PFASs and MD.
Our analysis was strengthened by several unique methodological features. First, we used mammographic images from a volunteer population of cancer-free women, providing a well-defined sample for assessing breast tissue characteristics in the absence of disease. We measured a comprehensive panel of PFASs, including those most commonly detected in the U.S. population. Given that different PFASs may exert distinct biological effects, it is important to examine compounds beyond PFOA and PFOS, which have been the primary focus of most epidemiologic research. Additionally, we employed statistical methods designed to model PFAS concentrations as a chemical mixture, providing a more accurate representation of real-world human exposure. These approaches enabled us to assess both the individual effects of each PFAS while accounting for co-exposure to others, as well as the overall impact of the PFASs mixture.
Our study has several limitations. The majority of KTB donors are White and reside in Indiana. Although non-White participants represented nearly 30% of our study population, future research should include more racially and ethnically diverse populations. However, it is unlikely that mechanisms linking PFASs to MD would vary by race/ethnicity. Additionally, our analysis was limited to postmenopausal women who were not using HT; different associations may be apparent among premenopausal women or postmenopausal women currently using HT. PFASs exposure was assessed at a single time point concurrent with the mammogram. However, serum PFASs levels are reasonable indicators of long-term exposure (likely several years, depending on the compound). [3] The vast majority of mammograms were taken within 2 years after the blood sample, thus supporting the relevance of our serum PFAS measurements for reflecting exposures that occurred prior to the assessment of MD. We acknowledge that our cross-sectional analysis cannot assess causal relationships, and we recommend future prospective studies explore associations between PFAS exposure and MD. Lastly, while our study was reasonably powered for analyses in the full sample, stratified results should be viewed with caution due to smaller subgroup sizes and reduced statistical power.
Conclusion
Overall, our study did not identify strong, significant associations between a panel of PFASs and MD. Together with our prior results reporting no substantial associations between PFASs and breast tissue involution, our research suggests that potential associations between postmenopausal PFASs exposure and future risk of breast cancer are not mediated by these breast cancer biomarkers. Future research employing prospective designs, larger sample sizes, and exploring various windows of susceptibility will be useful for understanding associations between PFAS and MD.
Supplementary Information
Below is the link to the electronic supplementary material.
Acknowledgements
Samples from the Susan G. Komen Tissue Bank at the IU Simon Comprehensive Cancer Center were used in this study. We thank contributors, including Indiana University who collected samples used in this study, as well as donors and their families, whose help and participation made this work possible. The authors thank Mr. Kevin Casey for his contributions to data management and analysis.
Abbreviations
- ANOVA
Analysis of variance
- BIRADS
Breast imaging reporting and data system
- BKMR
Bayesian kernel machine regression
- BMI
Body mass index
- DBA
Dense breast area
- ER
Estrogen receptor
- HPLC/MS-MS
High-performance liquid chromatography with tandem mass spectrometry
- HT
Hormone therapy
- KTB
Komen Tissue Bank
- LIBRA
Laboratory for Individualized Breast Radiodensity Assessment
- LOD
Limit of detection
- MD
Mammographic density
- PFAS
Perfluoroalkyl substance
- PFDA
Perfluorodecanoate
- PFHpS
Perfluoroheptane sulfonate
- PFHxS
Perfluorohexane sulfonate
- PFNA
Perfluorononanoate
- PFOA
Perfluorooctanoic acid
- PFOS
Perfluorooctane sulfonic acid
- PFUnDA
Perfluoroundecanoate
- PMD
Percent mammographic density
- SD
Standard deviation
Author contributions
SN performed statistical analyses, interpreted data, and wrote the initial manuscript draft. KWR conceived the research topic, designed the study, acquired data, interpreted data, and edited manuscript drafts. YO designed the study, interpreted data, and edited manuscript drafts. JH analyzed data and edited manuscript drafts. PG and FN performed laboratory analyses, interpreted data, and edited manuscript drafts. AL acquired data. All authors read and approved the final manuscript.
Funding
R21ES034801, National Institute of Environmental Health Sciences, PI Katherine W. Reeves.
Data availability
Data on Komen Tissue Bank participants are publicly available to researchers through the Virtual Tissue Bank ( [https://virtualtissuebank.iu.edu](https:/virtualtissuebank.iu.edu) ).
Declarations
Ethics approval and consent to participate
All Komen Tissue Bank participants provided written informed consent at the time of donation. All procedures involving human participants were conducted in accordance with the Declaration of Helsinki. Human subjects approval for the Komen Tissue Bank was provided by the Institutional Review Board at Indiana University. The present research utilized fully deidentified data and previously collected biospecimens and thus was determined to not qualify as human subjects research by the Institutional Review Board at the University of Massachusetts Amherst.
Consent for publication
Not applicable.
Competing interests
PG has served as a health expert for PFAS-exposed populations in legal cases. KR, YO, SN, JH, AL and FN declare that they have no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Calafat AM, et al. Legacy and alternative per- and polyfluoroalkyl substances in the U.S. general population: Paired serum-urine data from the 2013–2014 National Health and Nutrition Examination Survey. Environ Int. 2019;131:105048. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Sunderland EM, et al. A review of the pathways of human exposure to poly- and perfluoroalkyl substances (PFASs) and present understanding of health effects. J Expo Sci Environ Epidemiol. 2019;29:131–47. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Medicine Division; Division on Earth and Life Studies; Board on Population Health and Public Health Practice, Health M. Board on Environmental Studies and Toxicology; Committee on the Guidance on PFAS Testing and Health Outcomes. Guidance on PFAS Exposure, Testing, and Clinical Follow-Up. Washington (DC): National Academies Press (US); 2022. [PubMed] [Google Scholar]
- 4.IARC. Perfluorooctanoic Acid (PFOA) and Perfluorooctanesulfonic Acid (PFOS). [PubMed]
- 5.Chang C-J, et al. Exposure to per- and polyfluoroalkyl substances and breast cancer risk: a systematic review and meta-analysis of epidemiologic studies. Am J Epidemiol. 2024;193:1182–96. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Pettersson A, et al. Mammographic density phenotypes and risk of breast cancer: a meta-analysis. J Natl Cancer Inst. 2014;106:dju078. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Boyd NF, et al. Mammographic density as a marker of susceptibility to breast cancer: a hypothesis. IARC Sci Publ. 2001;154:163–9. [PubMed] [Google Scholar]
- 8.Cook JC, Murray SM, Frame SR, Hurtt ME. Induction of Leydig cell adenomas by ammonium perfluorooctanoate: a possible endocrine-related mechanism. Toxicol Appl Pharmacol. 1992;113:209–17. [DOI] [PubMed] [Google Scholar]
- 9.Biegel LB, Liu RC, Hurtt ME, Cook JC. Effects of ammonium perfluorooctanoate on Leydig cell function: in vitro, in vivo, and ex vivo studies. Toxicol Appl Pharmacol. 1995;134:18–25. [DOI] [PubMed] [Google Scholar]
- 10.Yang C, Tan YS, Harkema JR, Haslam SZ. Differential effects of peripubertal exposure to perfluorooctanoic acid on mammary gland development in C57Bl/6 and Balb/c mouse strains. Reprod Toxicol Elmsford N. 2009;27:299–306. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Zhao Y, Tan YS, Haslam SZ, Yang C. Perfluorooctanoic acid effects on steroid hormone and growth factor levels mediate stimulation of peripubertal mammary gland development in C57BL/6 mice. Toxicol Sci Off J Soc Toxicol. 2010;115:214–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Zhao B, Chu Y, Hardy DO, Li X, Ge R-S. Inhibition of 3beta- and 17beta-hydroxysteroid dehydrogenase activities in rat Leydig cells by perfluorooctane acid. J Steroid Biochem Mol Biol. 2010;118:13–7. [DOI] [PubMed] [Google Scholar]
- 13.Benninghoff AD, et al. Promotion of hepatocarcinogenesis by perfluoroalkyl acids in rainbow trout. Toxicol Sci Off J Soc Toxicol. 2012;125:69–78. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Boyd NF, et al. Mammographic density, response to hormones, and breast cancer risk. J Clin Oncol Off J Am Soc Clin Oncol. 2011;29:2985–92. [DOI] [PubMed] [Google Scholar]
- 15.Lee E, et al. Serum levels of commonly detected persistent organic pollutants and per- and polyfluoroalkyl substances (PFASs) and mammographic density in postmenopausal women. Int J Environ Res Public Health. 2020;17:E606. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Ning YS, et al. PFAS levels, early life factors, and mammographic breast density in premenopausal women. Environ Health Perspect. 2024;132:97008. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Sherman ME, et al. The Susan G. Komen for the cure tissue bank at the iu simon cancer center: a unique resource for defining the molecular histology of the breast. Cancer Prev Res (Phila Pa. 2012;5:528–35. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Li Y, et al. Half-lives of PFOS, PFHxS and PFOA after end of exposure to contaminated drinking water. Occup Environ Med. 2018;75:46–51. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Pérez F, et al. Accumulation of perfluoroalkyl substances in human tissues. Environ Int. 2013;59:354–62. [DOI] [PubMed] [Google Scholar]
- 20.Nielsen F, Fischer FC, Leth PM, Grandjean P. Occurrence of major perfluorinated alkylate substances in human blood and target organs. Environ Sci Technol. 2024;58:143–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Keller BM, Chen J, Daye D, Conant EF, Kontos D. Preliminary evaluation of the publicly available Laboratory for Breast Radiodensity Assessment (LIBRA) software tool: comparison of fully automated area and volumetric density measures in a case-control study with digital mammography. Breast Cancer Res BCR. 2015;17:117. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Keller BM, et al. Estimation of breast percent density in raw and processed full field digital mammography images via adaptive fuzzy c-means clustering and support vector machine segmentation. Med Phys. 2012;39:4903–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Qi W, Clark JM, Timme-Laragy AR, Park Y. Per- and polyfluoroalkyl substances and obesity, Type 2 diabetes and non-alcoholic fatty liver disease: a review of epidemiologic findings. Toxicol Environ Chem. 2020;102:1–36. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Reeves KW, et al. Longitudinal association of anthropometry with mammographic breast density in the Study of Women’s Health Across the Nation. Int J Cancer. 2009;124:1169–77. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Hart V, et al. The effect of change in body mass index on volumetric measures of mammographic density. Cancer Epidemiol Biomark Prev Publ Am Assoc Cancer Res Cosponsored Am Soc Prev Oncol. 2015;24:1724–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Timmermann CAG, et al. Shorter duration of breastfeeding at elevated exposures to perfluoroalkyl substances. Reprod Toxicol Elmsford N. 2017;68:164–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Romano ME, et al. Maternal serum perfluoroalkyl substances during pregnancy and duration of breastfeeding. Environ Res. 2016;149:239–46. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Reeves KW, Oulhote Y, Grandjean P, Nielsen F. Associations of perfluoroalkyl substances (PFAS) with terminal ductal lobular unit involution of the normal breast. Breast Cancer Res BCR. 2025;27:148. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Figueroa JD, et al. Standardized measures of lobular involution and subsequent breast cancer risk among women with benign breast disease: a nested case-control study. Breast Cancer Res Treat. 2016;159:163–72. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Data on Komen Tissue Bank participants are publicly available to researchers through the Virtual Tissue Bank ( [https://virtualtissuebank.iu.edu](https:/virtualtissuebank.iu.edu) ).