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

Abstract

Methods for extrapolating measured in vitro intrinsic clearance to a whole-body biotransformation rate constant (k B ) have been developed to support modeled bioaccumulation assessments for fish. The inclusion of extrapolated k B values into existing bioaccumulation models improves the prediction of chemical bioconcentration factors (BCFs), but there remains a tendency for these methods to overestimate BCFs relative to measured values. Therefore, a need exists to evaluate the extrapolation procedure to assess potential sources of error in predicted k B values. We examined how three different approaches (empirically based, composition based, and polyparameter linear free energy relationships [ppLFERs]) used to predict chemical partitioning in vitro (liver S9 system; K S9W ), in blood (K BW ), and in whole fish tissues (K FW ) impact the prediction of a chemical's hepatic clearance binding term (f U ) and apparent volume of distribution (V D ), both of which factor into the calculation of k B and the BCF. Each approach yielded different K S9W , K BW , and K FW values, but resulted in f U values that were of similar magnitude and remained relatively constant at log octanol-water partition ratios (K OW ) greater than 4. This is because K BW and K S9W values predicted by any given approach exhibit a similar dependence on log K OW (i.e., regression slope), which results in a cancelation of "errors" when f U is calculated. In contrast, differences in K BW values predicted by the three approaches translate to differences in V D , and by extension k B and the BCF, which become most apparent at log K OW greater than 6. There is a need to collect K BW and V D data for hydrophobic chemicals in fish that can be used to evaluate and improve existing partitioning prediction approaches in extrapolation models for fish.

Bioconcentration Assessment in Fish Based on In Vitro Intrinsic Clearance: Predictivity of an Empirical Model Compared to In Vitro–In Vivo Extrapolation Models

“…The mechanistic IVIVE models may be considered preferable from a conceptual point of view to the purely empirical regression model aimed at optimized predictivity. However, significant uncertainties in the model parameters do exist for the IVIVE models such as the volume of distribution ( V D ) and k 1 . , Until now, it has not been clear which of the IVIVE models gives the best predictivity and which should be preferred for regulatory application when using in vitro data from the OECD TG 319B. Our data provides the most comprehensive set of in vivo BCF and associated CL IN VITRO,INT data, which were generated according to the OECD TG 319B and equally the most comprehensive set of predictions using the four iterations of the IVIVE models.…”

“…17 The mechanistic IVIVE models may be considered preferable from a conceptual point of view to the purely empirical regression model aimed at optimized predictivity. However, significant uncertainties in the model parameters do exist for the IVIVE models such as the volume of distribution (V D ) 52 and k 1 . 17,19 Until now, it has not been clear which of the IVIVE models gives the best predictivity and which should be preferred for regulatory application when using in vitro data from the OECD TG 319B.…”

An amendedin vitro–in vivoextrapolation model that accounts for first pass clearance effects on chemical bioaccumulation in fish

“…Recent work has shown that f U values can be estimated with relatively high confidence, 46,47 but there remains considerable uncertainty in the calculation of V D . 48 In particular, uncertainty in the prediction of K BW directly translates to the calculation of V D (eqn (6)) and by extension modelled k B and BCF values. The impact of this uncertainty increases with chemical K OW and is especially apparent at log K OW ≥ 6.…”

“…The impact of this uncertainty increases with chemical K OW and is especially apparent at log K OW ≥ 6. 48 Based on an evaluation of several K BW prediction models for neutral organic chemicals, Nichols et al 13 reported that existing data are best described by an empirical equation given by Fitzsimmons et al 49 For this reason, we have retained the use of empirical equations to estimate K S9W and K BW , but stress that there is a need for directed studies to evaluate K BW and V D and, if necessary, refine estimation approaches for IVIVE model partition ratios.…”

In Vitro–In Vivo Extrapolation of Hepatic Biotransformation Data for Fish. III. An In-depth Case Study with Pyrene

“…Alternatively, this binding may be predicted using polyparameter linear free‐energy relationships (ppLFERs; Krause & Goss, 2020). A comparison of these approaches indicates, however, that they yield different binding predictions, which translates to differences in predicted k B , and by extension the BCF (Saunders & Nichols, 2023). With only one known exception (Laue et al, 2020), all reported evaluations of the IVIVE procedure performed to date have employed binding terms estimated by one of these methods.…”