The derivation of a Reference Dose (RfD) for perfluorooctane sulfonate (PFOS) based on immune suppression

Highlights

• PFOS causes decreased plaque forming cell response in mice.

• PFOS exposure in humans is associated with decreased vaccine response.

• PFOS exposure in humans is associated with increased risk of childhood infections.

• The animal and human data suggest that PFOS causes immunosuppression.

• Based on immunosuppression in mice, an RfD of 1.8 × 10⁻⁶ mg/kg/day was calculated.

Abstract

Exposure to perfluorooctane sulfonate (PFOS) is ubiquitous in populations and environments worldwide. Its long half-life in humans, indefinite persistence in the environment, and awareness of its widespread presence in drinking water make the human health assessment of PFOS a priority. While developmental, endocrine, and hepatic effects, and increased serum cholesterol are among the outcomes resulting from PFOS exposure, immunosuppression has also consistently emerged as an adverse effect. An in-depth review of the relevant scientific literature on the toxicology of PFOS has identified immunosuppression as a sensitive endpoint for PFOS toxicity. Here, we focus specifically on that endpoint and provide a detailed derivation of a Reference Dose (RfD) of 1.8 × 10⁻⁶ mg/kg/day for chronic human exposure to PFOS. This RfD is based on decreased plaque-forming cell (PFC) response in mice, an endpoint that reflects suppression of the immune response to a foreign antigen. We additionally identify two endpoints in the epidemiology literature, decreased vaccine response and increased incidence of childhood infections, that are associated with PFOS exposure and that are consistent with and support the decreased PFC response endpoint from animal studies. We provide a weight of evidence analysis integrating the evidence from animal and epidemiology endpoints. Finally, we compare this RfD to the PFOS RfD derived by the United States Environmental Protection Agency (USEPA) Office of Water based on a developmental endpoint. Based on this comparison, and given our assessment, the USEPA RfD does not provide sufficient protection against the adverse health effects of PFOS. The RfD derived herein is intended to be public health protective and appropriately minimizes PFOS exposure based on available evidence.

Introduction

Perfluorooctane sulfonate (PFOS) is a widely occurring environmental contaminant of public health concern. The chemistry of PFOS (e.g., its carbon-fluorine bonds) led to its use in a wide array of commercial and industrial applications, such as a stain/water repellant for fabrics, in metal plating and finishing, photograph development, and food packaging (USEPA, 2016a). Notably, PFOS also has been a constituent of aqueous film forming foam (AFFF) used in extinguishing Class B fires (i.e., involving flammable liquids) (Seow, 2013). Although the production and use of PFOS and related chemistries (e.g., perfluorooctanesulfonyl fluoride) was phased out in the United States in 2002 (USEPA, 2016a), the persistence of PFOS (see below) results in its continuing presence in the environment. PFOS is detected in ground and surface water, fish and other biota, soil, and house dust both near sources of contamination and globally (USEPA, 2016a). PFOS has been found in drinking water at numerous locations throughout the U.S. and worldwide, particularly near sites where AFFF was used (Hu et al., 2016, Post et al., 2017).

The chemistry of PFOS also has important implications for its distribution in environmental media and biota. Due to its resistance to physical and biological degradation, PFOS persists indefinitely in the environment. PFOS has both hydrophilic and lipophilic characteristics, and strongly, but non-covalently binds to protein, including in fish (Conder et al., 2008), while not accumulating in lipid-rich tissues.

The presence of PFOS and its precursors in multiple environmental media allows for multiple sources of exposure by humans. In contrast to other well-known persistent and bioaccumulative compounds such as PCBs and dioxins, PFOS is water soluble and drinking water is an important exposure route. Additionally, infants may be exposed to PFOS through breast milk (ATSDR, 2018).

Since 1999, the National Health and Nutrition Examination Survey (NHANES) has measured PFOS in the serum of a representative sample of the U.S. general population. As of 2013–2014, the median and 95th percentile of serum PFOS concentrations were 5.2 and 18.5 μg/L, respectively (ng/ml; CDC, 2017). Although the level of PFOS in human serum has been declining, the human half-life of 5.4 years (Olsen et al., 2007) raises a particular concern for adverse health effects in humans.

Health hazard assessments of PFOS have generally identified increased serum cholesterol, liver effects, decreased thyroid hormone levels, immunotoxicity, and developmental effects such as offspring mortality, decreased body weight, and neurotoxicity. While human cancer data are inconsistent, liver tumors have been observed in rats (reviewed in ATSDR, 2018; USEPA, 2016b).

Quantitative assessments of identified health effects have developed daily oral intake values (ng/kg/day) intended to be protective for chronic exposure (e.g., Tolerable Daily Intake [TDI], Reference Dose [RfD]) to PFOS (Dong et al., 2017, Lilienthal et al., 2017). Over time, these values have trended lower (Dong et al., 2017). To date, the bases for the derivation of these values have primarily been decreased serum triiodothyronine (T3) levels in monkeys, liver effects in rats, or decreased offspring body weight in rats (reviewed in Dong et al., 2017). Although consistently identified as an effect of PFOS, immunotoxicity has not been used as the primary basis (i.e., the critical effect) for daily intake values despite strong evidence that this effect can result from exposure to low levels PFOS.

As part of an independent quantitative assessment of PFOS (DWQI, 2018), a comprehensive literature search and screening was conducted to identify relevant human and laboratory animal information for the identification of potential health hazards from PFOS exposure. Immune suppression was selected as the critical effect and the basis for quantitative risk assessment. Specifically, the immunotoxic effect selected for the RfD was decreased plaque forming cell (PFC) response (Jerne and Nordin, 1963, Cunningham and Szenberg, 1968), in mice following inoculation with a foreign antigen (sheep red blood cells [SRBCs]) as reported in Dong et al. (2009). Here, we focus and expand on the qualitative (Hazard Identification) and quantitative (Exposure-Response) rationale for using immune suppression from Dong et al. (2009) as the basis for development of an RfD for PFOS.