The purpose of the present work is to translate an existing physiologically based pharmacokinetic (PBPK) model for perfluorooctanoic acid (PFOA) to R, which will be used to study human exposure of PFOA in blood from diet and cosmetics. The model is for the adult general population.

PFOA is a part of a large group of compounds called per- and polyfluoroalkyl substances (PFASs), which are widely used chemically synthesized compounds that are water, stain, and oil repellent. They are highly persistent and several of the compounds, including PFOA, accumulates in the food chain. The health effects of PFOA in humans are still being studied, but exposure are associated with a range of health effects, including developmental and reproductive toxicity, immune system effects, and cancer. Increased knowledge of human exposure and internal dose in humans are important to link exposure of PFOA to the adverse health effects.

Several PBPK model for PFOA have been published, but the model that have been used the most was first published by Loccisano et al in 2011 (https://www.sciencedirect.com/science/article/abs/pii/S0273230010002242?via%3Dihub). This model has been widely used and was partly modified and used by the European Food Safety Authority (EFSA) in 2020 (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7507523/). EFSA modeled PFOA using the Berkeley Madonna software in adults and breastfeeding children and the code was published in the EFSA opinion.

We have only used the adult part of the model and translated it from Berkeley Madonna to R. The model was modified by including an dermal exposure rout, a fecal excretion of PFOA, and a few other modifications to the kidney compartment. The first work using the PBPK model is published open access in the paper "Comparison of aggregated exposure to PFOA from diet and personal care products with concentrations in serum using a PBPK model – results from the Norwegian biomonitoring study in EuroMix" by Husøy et al 2023, where you can find information on the use of the model and the first results. Challenges and weaknesses in the PBPK model is discussed in the paper.

The R code to run the exposure assessment from PCPs
PFOA_PCP_version1.R
The data files as input to the exposure estimates from PCPs
2_SumPCPPFAS_LB_050122.csv
2_SumPCPPFAS_MB_050122.csv
2_SumPCPPFAS_UB_050122.csv
AmountsPerApplication_female.csv
AmountsPerApplication_male.csv
EEF_distributions_female.csv
EEF_distributions_male.csv
PCP_frequency_dummy.csv
EuroMix_dummy_sex_weight.csv

The R code to run the dietary exposure assessment
PFOA_food_version1.R
The data files as input to the dietary exposure
3-SumPFAS_food_conc_LB.csv
3-SumPFAS_food_conc_MB.csv
3-SumPFAS_food_conc_UB.csv
foodintake_dummy_day1.csv
foodintake_dummy_day2.csv
EuroMix_dummy_sex_weight.csv

The R code to run the PBPK model
PBPK_PFOA_version1_150323.R
The data files as input to the PBPK modeling from the exposure
SumPFOA_LB_food_PCP.csv

PBPK_PFOA_pksensi_version1.R

Establish the folders "Code", "Data" and "Results" in your directory and copy the relevant files to the appropriate folders. Open the R or Rmd code in RStudio and insert your work directory in the code. The data files will be uploaded and create a new folder under "Results" with the date of the day to store the results. Run the exposure assessment for the PCP first as the output will be merged with the results from the food later. This can be input to the PBPK model.

Normally I would create a R project linked to github as a start of my work, but you can also run the PBPK model just by open it in R. Upload the dummy data file from the Data folder and run the model. The model includes exposure from both diet and personal care products, and you can run one at the time by setting the exposure to zero for one of them. Run the model from ###START PBPK## to ###END PBPK###.

The PBPK model in the sensitivity analysis is presented separately for now. Here only one concentration for the diet and PCP is tested in each run, and the values have to be keyed in manually. We have used R package pksensi developed by Hsieh et al 2020 (https://pubmed.ncbi.nlm.nih.gov/33426260/), which is a tool for global sensitivity analysis of PBPK models. The repository for the published version of pksensi you can find here here https://github.com/ElsevierSoftwareX/SOFTX_2020_29. We have included the parameters from the PBPK model of PFOA that we believe is the most important ones, and the output will give you an uncertainty evaluation and a sensitivity analyse of all the included parameters.