Pharmacokinetics, Pharmacodynamics and Drug Metabolism
Tissue lipids and drug distribution: Dog versus rat

https://doi.org/10.1002/jps.23285Get rights and content

ABSTRACT

A key parameter in whole-body physiologically based pharmacokinetic models is the tissue-to-plasma water partition coefficient (Kpu), which is commonly assumed consistent across all species for all tissues for passively distributing drugs. Many drugs primarily bind to tissue lipids and although considerable tissue lipid concentration data exist in rodents, data on these and Kpu values in larger animals and humans are sparse to negligible. To test the above assumption, lipid levels were quantified in 13 dog tissues, then compared with the values in rat, and used to predict and compare Kpu values between these species. For many tissues, including muscle, lipid concentrations were comparable in dog and rat. However, spleen acidic phospholipid levels were sevenfold lower, skin neutral phospholipid threefold lower, and neutral lipids fivefold, 12-fold, and eightfold lower in brain, lung, and spleen, respectively, and fourfold higher in bone in dog than in rat. Such differences resulted in significant predicted Kpu differences. In contrast, unbound volume of distribution (Vuss), a global measure of distribution, showed generally good agreement (predictions and observations) between dog and rat for a diverse compound set, indicating tissues with large-predicted Kpu species differences tend either to contribute to Vuss to a limited extent, and/or occur in opposing directions tending to cancel each other out.

Section snippets

INTRODUCTION

A common assumption when scaling drug tissue distribution data from one species to another is that for any tissue the tissue-to-plasma water drug partition coefficient (Kpu) for passive processes is the same across all species. This in turn assumes that the composition of the constituents responsible for drug binding within a tissue is the same across all species. One major class of tissue constituents responsible for binding of many compounds is lipid, both neutral and acidic. For these

Lipid Analysis in Dog

Beagle dogs (n = 3) used for this study were 1-year-old naive male dogs, mean weight 17.7 kg, which were terminated via a lethal injection of phenobarbital. These animals were not terminated for the purposes of this study. Tissues were harvested and primarily used for safety studies but in accordance with the 3Rs, tissues were made available to other departments within Pfizer. Following termination, the dogs were exsanguinated before removal of the heart; the blood was centrifuged to prepare

Lipid Analysis in Dog

The dog tissue lipid concentrations obtained from MLA are presented in Table 1, with comparisons to rat tissue lipid concentrations given in Table 1 and Figure 1. For some tissues such as muscle, gastrointestinal tract, and heart, there was no material difference in lipid composition between dog and rat. However, concentrations differ by more than threefold in other tissues. For example, in dog, AP levels in spleen are sevenfold lower, NP levels are threefold lower in skin, whereas NL levels

DISCUSSION

A literature search revealed tissue distribution data in dog for some drugs but these are few and of limited value. At best, results, such as for those for amiodarone, digoxin, thiopental, and salicylate,34., 35., 36., 37. tend to come in the form of reported ratio of total tissue concentration to plasma concentration values that need to be corrected for plasma protein binding in the dog, which is often lacking, to allow an estimate of Kpu, the preferred measure of affinity of tissues for

CONCLUSIONS

Because species differences in tissue composition and Kpus has been highlighted, the use of species-specific tissue compositional data for Kpu predictions in organs of interest would be essential for predicting exposure in these species, or caution should be exercised in assuming constancy in Kpu values across all species. On the contrary, if whole-body distribution is the focus then species-specific tissue lipid data are not critical. When put into the context of physiologically based

Acknowledgements

One of the authors (Dr Trudy Rodgers) thanks the Centre for Applied Pharmacokinetic Research, University of Manchester for funding this research, and the reviewers of this article for their constructive comments.

REFERENCES (48)

  • L.I. Harrison et al.

    Physiologically based pharmacokinetic model for digoxin disposition in dogs and its preliminary application to humans

    J Pharm Sci

    (1977)
  • K.B. Bischoff et al.

    Thiopental pharmacokinetics

    J Pharm Sci

    (1968)
  • C.N. Chen et al.

    Pharmacokinetic model for simultaneous determination of drug levels in organs and tissues

    J Pharm Sci

    (1976)
  • T. Rodgers et al.

    Tissue distribution of basic drugs: Accounting for enantiomeric, compound and regional differences amongst beta-blocking drugs in rat

    J Pharm Sci

    (2005)
  • P. Poulin et al.

    Advancing prediction of tissue distribution and volume of distribution of highly lipophilic compounds from a simplified tissue-composition-based model as a mechanistic animal alternative method

    J Pharm Sci

    (2012)
  • W. Schmitt

    General approach for the calculation of tissue to plasma partition coefficients

    Toxicol In Vitro

    (2008)
  • T. Peyret et al.

    A unified algorithm for predicting partition coefficients for PBPK modeling of drugs and environmental chemicals

    Toxicol Appl Pharmacol

    (2010)
  • C.F. Baxter et al.

    Variations among vertebrates of lung phospholipid class composition

    Lipids

    (1969)
  • G. Rouser et al.

    Species variations in phospholipid class distribution of organs. 1. Kidney, liver and spleen

    Lipids

    (1969)
  • G. Simon et al.

    Species variations in phospholipid class distribution of organs. II. Heart and skeletal muscle

    Lipids

    (1969)
  • G.B. Spurr et al.

    Plasma and erythrocyte Na, K, Cl and water in hypothermic and hyperthermic dogs

    Am J Physiol

    (1959)
  • T. Rodgers et al.

    Mechanistic approaches to volume of distribution predictions: Understanding the processes

    Pharm Res

    (2007)
  • B.S. Shin et al.

    Physiologically based pharmacokinetics of bisphenol A

    J Toxicol Environ Health A

    (2004)
  • B. Davies et al.

    Physiological parameters in laboratory animals and humans

    Pharm Res

    (1993)
  • Cited by (26)

    • Application of the Tissue Composition–Based Model to Minipig for Predicting the Volume of Distribution at Steady State and Dermis-to-Plasma Partition Coefficients of Drugs Used in the Physiologically Based Pharmacokinetics Model in Dermatology

      2019, Journal of Pharmaceutical Sciences
      Citation Excerpt :

      The development of unified algorithms was accomplished by integrating the most accurate TCMs previously published to compute the Kp of drugs and environmental chemicals to structure the predictions of the distribution of organic compounds at the macrolevel (i.e., whole tissue) and microlevel (i.e., cells and fluids).7,18 Accordingly, different comparative assessments showed that the TCM of Poulin and Haddad showed the highest degree of accuracy with the highly lipophilic neutral drugs,8 whereas, the TCM of Rodgers and Rowland was the most accurate model with the other classes of drugs.13-21 Furthermore, it has recently been demonstrated that the TCM is comparably accurate or superior to empirical allometric approaches based on the extrapolation of in vivo data from preclinical species to predict the Vss in human, particularly while the TCM is able to predict Vss in preclinical species before the extrapolation to humans.21

    • Drug distribution to human tissues: Prediction and examination of the basic assumption in in vivo pharmacokinetics-pharmacodynamics (PK/PD) research

      2015, Journal of Pharmaceutical Sciences
      Citation Excerpt :

      As the in vivo Kp value equals to the ratio of total concentration between tissue and plasma, or the ratio between fup and fut, this parameter Kp would deviate from the unity. This is because there are diverse effects on both sides of the membrane, namely, the pH gradient, non-specific binding to lipids, and binding to plasma proteins or lipoproteins.1–12 In this situation, although the fup will differ to fut because of the dissimilarities between the proteins and lipids content on both sides of the membrane, and, hence, of the bound drug concentration, the free drug concentration in the aqueous phase would be the same in plasma and tissue at steady state but only for the neutral compounds.

    • Prediction of Drug Distribution in Subcutaneous Xenografts of Human Tumor Cell Lines and Healthy Tissues in Mouse: Application of the Tissue Composition-Based Model to Antineoplastic Drugs

      2015, Journal of Pharmaceutical Sciences
      Citation Excerpt :

      Accordingly, the tissue composition-based model is described in a way that it can be used to predict tumor tissue partitioning at both the total and free drug level (i.e., partitioning into the aqueous, lipid and protein compartments can be estimated separately) (Table 4), and, hence, could be related to toxicity and efficacy end points. Researchers have already successfully explored the interspecies differences in Kp values for healthy tissues, upholding the notion that the scaling of Kp values across species on the basis of differences in tissue composition and plasma protein binding (and the elimination effect in the case of liver) is accurate and scientifically justifiable.4,62 For rough estimates in humans, the tissue composition of tumors can be assumed to be relatively conserved across species, and, therefore, predictions of tumor Kp in any species under steady-state in vivo conditions can be made by using the tissue composition-based model.

    • Predicting passive and active tissue:Plasma partition coefficients: Interindividual and interspecies variability

      2014, Journal of Pharmaceutical Sciences
      Citation Excerpt :

      However, it did eliminate the need for an assumption used by Schmmit,1 Furthermore, this approach still defines the unbound chemical fraction in terms of concentration so it can be directly estimated from experiments. No previous study has compared interindividual and interspecies Kt:plPassive variability using the tissue composition Model described herein or the original,1 There are however, other Models4,11 that have compared species variability but not interindividual variability. These Models previously showed that interspecies differences exist for the rat, Mouse, rabbit, and human, and support the conclusion herein.

    • An algorithm for evaluating potential tissue drug distribution in toxicology studies from readily available pharmacokinetic parameters

      2013, Journal of Pharmaceutical Sciences
      Citation Excerpt :

      This strategy was successful in this study because relatively good estimates of Kp values in dog were obtained for three structurally diverse compounds (Table 4). Differences in fut across species could either be estimated from the corresponding differences in the tissue lipid content as shown in this study,14-17 or experimentally determined in vitro by using tissue homogenates from each species.19,24,30,31 Furthermore, it could be assumed that the correlation equations depicted inFigure 3 are conserved across species, and therefore, predictions of Kp in any species can only be made by using the input Vd for the respective species.

    View all citing articles on Scopus

    Trudy Rodgers’ present address is ICON Development Solutions, Manchester M15 6SH, UK.

    ICON Development Solutions, Manchester M15 6SH, UK.

    View full text