Using in vitro data to derive acceptable exposure levels: A case study on PBDE developmental neurotoxicity

Sherri Bloch, Laura Lévêque, Irva Hertz-Picciotto, Birgit Puschner, Ellen Fritsche, Jördis Klose, Nynke I. Kramer, Maryse F. Bouchard, P. Chandrasekera, Marc André Verner*

*Corresponding author for this work

Research output: Contribution to journalArticleAcademicpeer-review


Background: Current acceptable chemical exposure levels (e.g., tolerable daily intake) are mainly based on animal experiments, which are costly, time-consuming, considered non-ethical by many, and may poorly predict adverse outcomes in humans. Objective: To evaluate a method using human in vitro data and biological modeling to calculate an acceptable exposure level through a case study on 2,2′,4,4′-tetrabromodiphenyl ether (BDE-47) developmental neurotoxicity (DNT). Methods: We reviewed the literature on in vitro assays studying BDE-47-induced DNT. Using the most sensitive endpoint, we derived a point of departure using a mass-balance in vitro disposition model and benchmark dose modeling for a 5% response (BMC05) in cells. We subsequently used a pharmacokinetic model of gestation and lactation to estimate administered equivalent doses leading to four different metrics of child brain concentration (i.e., average prenatal, average postnatal, average overall, and maximum concentration) equal to the point of departure. The administered equivalent doses were translated into tolerable daily intakes using uncertainty factors. Finally, we calculated biomonitoring equivalents for maternal serum and compared them to published epidemiological studies of DNT. Results: We calculated a BMC05 of 164 μg/kg of cells for BDE-47 induced alteration of differentiation in neural progenitor cells. We estimated administered equivalent doses of 0.925–3.767 μg/kg/day in mothers, and tolerable daily intakes of 0.009–0.038 μg/kg/day (composite uncertainty factor: 100). The lowest derived biomonitoring equivalent was 19.75 ng/g lipids, which was consistent with reported median (0.9–23 ng/g lipids) and geometric mean (7.02–26.9 ng/g lipids) maternal serum concentrations from epidemiological studies. Conclusion: This case study supports using in vitro data and biological modeling as a viable alternative to animal testing to derive acceptable exposure levels.

Original languageEnglish
Article number108411
JournalEnvironment International
Publication statusPublished - Jan 2024


  • Biomonitoring equivalent
  • In vitro toxicology testing
  • Mass-balance modeling
  • Pharmacokinetic modeling
  • Risk assessment
  • Tolerable daily intake


Dive into the research topics of 'Using in vitro data to derive acceptable exposure levels: A case study on PBDE developmental neurotoxicity'. Together they form a unique fingerprint.

Cite this