Table 2.

An overview of the main challenges related to KMD and TK saturation data, examples from the text, and recommendations to address these challenges.

Challenges	Case studies/examples	Recommendations
Inflection point: No inflection point in the hyperbolic relationship between administered dose and systemic exposure of a chemical that exhibits nonlinear TK behavior.	• Computational modeling to simulate administered dose versus systemic exposure relationships • Collecting TK data using in vitro approaches • See Section 2.1	• The current definition of a KMD that uses the term “inflection point” or “onset of nonlinear behavior” is not appropriate. • Use in vitro and in vivo TK data, along with available hazard data and human exposure information, in a weight of evidence approach to achieve a scientifically defensible decision on dose selection or response interpretation. • Develop guidance regarding the use of statistical methods to analyze proportionality between administered dose and systemic exposure, given a criterion determined from the weight of evidence assessment. • Use TK modeling throughout the chemical safety testing program to organize and integrate available data, test hypotheses on the mode of action, and recommend targeted studies to inform the role of TK on toxicity pathways.
Exposure: Human exposure levels are hard to predict, making it difficult to ensure that a top dose selected using TK information is higher than human exposures by a sufficient margin.	• Occupational worker dermal and inhalation exposures to pesticides • Dermal and inhalation exposures to a wide range of environmental chemicals • Screening level exposure estimates for 448 chemicals • See Section 2.2	• Many valid and robust exposure data resources and approaches are available to estimate human exposure levels, and uncertainty analysis has been an integral part of exposure assessment. • Due to the conservative nature of exposure assessment approaches, there is more uncertainty in the lower end of possible exposure prediction, rather than in the higher end. In most cases, all dose levels used in repeated dose animal toxicity studies are likely to be higher than human exposure levels by a sufficient margin.
3Rs: Obtaining TK data requires additional animals, which does not meet the 3Rs principles.	• Microsampling within the main study • In vitro methods to obtain ADME parameters + TK modeling (rifampicin) • See Section 2.3	• Dose selection is best supported by a scientific weight of evidence approach that collectively considers the available body of data • Leverage alternative methods (such as in vitro bioassays), novel techniques (such as microsampling), and computational modeling to reliably collect and integrate TK data without additional satellite animal groups.
Limitations in TK studies: Animal TK studies are not always reflective of human exposure scenarios; TK differences between animals and humans cannot be determined; TK endpoints from animal studies may not be an appropriate internal dose metric for the toxic moiety.	• MPP+ (selective concentration of the toxic moiety) • Dichloropropane (glutathione depletion in the lungs) • See Section 2.4	• To improve the relevance and predictability of animal models to human hazard identification and risk assessment, we recommend designing fit-for-purpose animal studies and using in vitro approaches and computational modeling (such as PBK analysis) to understand TK and TD differences between test species and humans. • Use the internal dose metric, or an appropriate surrogate, that is most closely related to toxicity responses to understand the impact of nonlinear TK on the dose-response relationship.
Conflicts with some regulatory requirements: Lower dose studies may not meet requirements for some regulatory purposes, such as classification and labeling or endocrine disruption assessment.	• See Section 2.5	• Use a weight of evidence approach to optimize the study design of repeated dose animal studies. • TK data may have a different weight in the dose selection process across regulatory programs, but these data can still be critical for other risk assessment objectives.
Insufficiency of TK information for top dose setting: TK data are not sufficient on their own to inform study design.	• Saturation of absorption (florpyrauxifen-benzyl, fenpicoxamid) • Saturation of absorption and clearance (sulfoxaflor) • See Section 3	• A weight of evidence approach that incorporates available data, such as hazard, TK, mechanistic, and exposure information, should be used to inform dose setting.

Abbreviations: 3Rs, reduction, replacement, and refinement; ADME, absorption, distribution, metabolism, excretion; TK, toxicokinetics.