Prediction of Unbound Fractions for in Vitro–in Vivo Extrapolation of Biotransformation Data

Abstract

For in vitro−in vivo extrapolation of biotransformation data, the different sorptive environments in vitro and in vivo need to be considered. The most common approach for doing so is using the ratio of unbound fractions in vitro and in vivo. In the literature, several algorithms for prediction of these unbound fractions are available. In this study, we present a theoretical evaluation of the most commonly used algorithms for prediction of unbound fractions in S9 assays and blood and compare prediction results with empirical values from the literature. The results of this analysis prove a good performance of "composition-based" algorithms, i.e. algorithms that represent the inhomogeneous composition of in vitro assay and in vivo system and describe sorption to the individual components (lipids, proteins, water) in the same way. For strongly sorbing chemicals, these algorithms yield constant values for the ratio of unbound fractions in vitro and in vivo. This is mechanistically plausible, because in these cases, the chemicals are mostly bound, and the ratio of unbound fractions is determined by the volume ratio of sorbing components in both phases. 8 f u,plasma : Saunders et al. 2020 8 2. Algorithms for Estimation of Fraction Unbound in S9 in Vitro Assays (f u,S9 ) 8 f u,S9 :

Validation of Methods for in Vitro–in Vivo Extrapolation Using Hepatic Clearance Measurements in Isolated Perfused Fish Livers

“…The obvious impacts of protein binding on clearances of all five chemicals, with the 10 g L −1 treatment consistently showing the lowest and the 1.0 g L −1 treatment consistently showing the greatest values measured values, are another clear indication that setting the term = f f f u blood unbound assay unbound to 1 is mechanistically inappropriate as it fails to correct for systemic bioavailability. 5,[20][21][22]32 Generally, the agreement between directly measured and extrapolated hepatic clearances was good for both in vitro input data based on hepatocytes (Figure 2A) and Figure S9 (Figure 2C), with RMSEs of 16.8 and 22.7 mL h −1 g −1 liver, respectively. Deviations of predicted from measured BCFs based on current IVIVE exercises may span several orders of magnitude, especially if the assumption of f u = 1.0 is not entertained.…”

“…As was observed for the impacts of protein binding on directly measured hepatic clearances in isolated perfused livers, hepatic clearance was predicted based on the in vitro data and partitioning taking into account the predicted binding in the presence of the different BSA concentrations and followed the same trend. The obvious impacts of protein binding on clearances of all five chemicals, with the 10 g L –1 treatment consistently showing the lowest and the 1.0 g L –1 treatment consistently showing the greatest values measured values, are another clear indication that setting the term to 1 is mechanistically inappropriate as it fails to correct for systemic bioavailability. ,− , Generally, the agreement between directly measured and extrapolated hepatic clearances was good for both in vitro input data based on hepatocytes (Figure A) and Figure S9 (Figure C), with RMSEs of 16.8 and 22.7 mL h –1 g –1 liver, respectively. Deviations of predicted from measured BCFs based on current IVIVE exercises may span several orders of magnitude, especially if the assumption of f u = 1.0 is not entertained .…”

“…20,21 More recently, however, a consensus has been reached that this practice was misleading and should be avoided, while instead more research is needed to identify the potential reasons for these discrepancies. 5,22 One such reason has been hypothesized to be the potential contribution of extrahepatic biotransformation in organs such as gill and gastrointestinal tract. 23,24 Additionally, the substantial variability of BCF measurements questions the robustness of its use as a comparator for IVIVE predictions.…”

“…One such factor is the influence of chemical binding to lipids and plasma proteins on freely dissolved chemical concentrations and consequently on in vitro biotransformation rates. Many researchers in this space found that incorporating the assumption that no difference in binding exists between in vitro assays and blood plasma in vivo , which is neither supported by theory nor empirical evidence, improved the apparent accuracy of models. , More recently, however, a consensus has been reached that this practice was misleading and should be avoided, while instead more research is needed to identify the potential reasons for these discrepancies. , One such reason has been hypothesized to be the potential contribution of extrahepatic biotransformation in organs such as gill and gastrointestinal tract. , Additionally, the substantial variability of BCF measurements questions the robustness of its use as a comparator for IVIVE predictions. , …”

Models Used to Predict Chemical Bioaccumulation in Fish from in Vitro Biotransformation Rates Require Accurate Estimates of Blood–Water Partitioning and Chemical Volume of Distribution

“…The ppLFER approach is appropriate for chemical‐specific assessments but is not suitable for generic analysis of a set of hypothetical chemicals because ppLFERs do not rely on K OW correlations (Krause & Goss, 2021). Therefore, to compare the three approaches, ppLFERs were used to estimate K S9W , K BW , and K FW values for specific neutral organic chemicals for which empirical V D data were available (Table 3).…”