A Machine Learning Model to Estimate Toxicokinetic Half-Lives of Per- and Polyfluoro-Alkyl Substances (PFAS) in Multiple Species

Abstract

Per- and polyfluoroalkyl substances (PFAS) are a diverse group of man-made chemicals that are commonly found in body tissues. The toxicokinetics of most PFAS are currently uncharacterized, but long half-lives (t½) have been observed in some cases. Knowledge of chemical-specific t½ is necessary for exposure reconstruction and extrapolation from toxicological studies. We used an ensemble machine learning method, random forest, to model the existing in vivo measured t½ across four species (human, monkey, rat, mouse) and eleven PFAS. Mechanistically motivated descriptors were examined, including two types of surrogates for renal transporters: (1) physiological descriptors, including kidney geometry, for renal transporter expression and (2) structural similarity of defluorinated PFAS to endogenous chemicals for transporter affinity. We developed a classification model for t½ (Bin 1: <12 h; Bin 2: <1 week; Bin 3: <2 months; Bin 4: >2 months). The model had an accuracy of 86.1% in contrast to 32.2% for a y-randomized null model. A total of 3890 compounds were within domain of the model, and t½ was predicted using the bin medians: 4.9 h, 2.2 days, 33 days, and 3.3 years. For human t½, 56% of PFAS were classified in Bin 4, 7% were classified in Bin 3, and 37% were classified in Bin 2. This model synthesizes the limited available data to allow tentative extrapolation and prioritization.

A Comparison of In Vitro Points of Departure with Human Blood Levels for Per- and Polyfluoroalkyl Substances (PFAS)

“…The model does not provide a numeric half-life, but instead a class, which is one of <0.5 days, <1 week, <2 months, or >2 months. This model predicts that all but one of the 31 PFAS analyzed here have half-lives > 2 months [ 66 ]. The one exception is PFPeA, with a half-life of <1 week.…”

“…Only a few PFAS have measured half-lives, but a recent paper by Dawson et al uses a QSAR model to predict human The approach used in this analysis assumes that chemicals have long half-lives, so that blood concentrations are relatively stable over time. Only a few PFAS have measured half-lives, but a recent paper by Dawson et al uses a QSAR model to predict human plasma half-lives for a large collection of PFAS [66]. The model does not provide a numeric half-life, but instead a class, which is one of <0.5 days, <1 week, <2 months, or >2 months.…”

Risk-based prioritization of PFAS using phenotypic and transcriptomic data from human induced pluripotent stem cell-derived hepatocytes and cardiomyocytes

“…Additional data, including that from in vitro studies, may provide a biological dimension for informing class-based approaches for read-across and/or within-group prioritization for additional testing. Indeed, toxicity and toxicokinetic data are being generated using in vitro testing and computational methodologies to inform PFAS hazard characterization (US EPA, 2019; Patlewicz et al, 2019Patlewicz et al, , 2022Kreutz et al, 2023;Dawson et al, 2023;Carstens et al, 2023). In addition, gene expression data offer a comprehensive understanding of chemical-induced bioactivity, delivering not only mechanistic insights but also dose-response information (Reardon et al, 2023;Johnson et al, 2022).…”

A Scoping Assessment of Implemented Toxicokinetic Models of Per- and Polyfluoro-Alkyl Substances, with a Focus on One-Compartment Models

“…For PFOS, current half-life estimates vary from 2.91 years [ 20 ] to 6.3 and 22 years [ 18 ] for females and males, respectively. A compilation of half-lives for these and other PFAS for multiple species is recently provided by Dawson et al [ 21 ]. Some short chain perfluoroalkyl acids have short elimination half-lives in humans and thus lower potential for bioaccumulation; for example, perfluorobutanoic acid (PFBA, DTXSID4059916) is eliminated with an estimated half-life of 72 days [ 22 ] and perfluorobutane sulfonate (PFBS, DTXSID5030030) is eliminated with an estimated half-life ranging from 0.12 days to 26 days for females and 28 days for males [ 23 ].…”

“…It is recognized that organic anion transport proteins play a key role in renal tubular reabsorption of certain PFAS [ 27 , 29 , 30 ]. However, the parameterization of mechanistic models of PFAS toxicokinetics (e.g., physiologically based toxicokinetic models (PBTK)) remains a challenge due to the relative lack of experimental toxicokinetic data for many PFAS and the large number of species-, sex-, and age-specific differences that have been observed [ 21 ]. Use of parsimonious models (i.e., one-compartment) is common in the literature for both commonly studied PFAS [ 31 ] and data poor scenarios, for which data is not readily available for model parameterization [ 15 ].…”