Research Articles
Prediction of Human Pharmacokinetics from Animal Data and Molecular Structural Parameters using Multivariate Regression Analysis: Oral Clearance

https://doi.org/10.1002/jps.10510Get rights and content

Abstract

The aim of the study reported here was to develop a regression equation for predicting oral clearance of various kinds of drugs in humans using experimental data from rats and dogs and molecular structural parameters. The data concerning the oral clearance of 87 drugs from rats, dogs, and humans were obtained from literature. The compounds have various structures, pharmacological activities, and pharmacokinetic characteristics. In addition, the molecular weight, calculated partition coefficient (c log P), and the number of hydrogen bond acceptors were used as possible descriptors related to oral clearance in human. Multivariate regression analyses, multiple linear regression analysis, and the partial least squares (PLS) method were used to predict oral clearance in human, and the predictive performances of these techniques were compared by allometric approaches, which have been used in interspecies scaling. Interaction terms were also introduced into the regression analysis to evaluate the nonlinear relationship. For the data set used in this study, the PLS model with the tertiary term descriptors gave the best predictive performance, and the value of the squared cross‐validated correlation coefficient (q2) was 0.694. This PLS model, using animal oral clearance data for only two species and easily calculated molecular structural parameters, can generally predict oral clearance in human better than the allometric approaches. In addition, the molecular structural parameters and the interaction term descriptors were useful for predicting oral clearance in human by PLS. Another advantage of this PLS model is that it can be applied to drugs with various characteristics. © 2003 Wiley‐Liss, Inc. and the American Pharmacists Association J Pharm Sci 92:2427–2440, 2003

Section snippets

INTRODUCTION

In the process of developing new drugs, it is very important to predict the pharmacokinetic profile of a drug in human for decisions on appropriate dosage and the best clinical trial design. In our previous paper,1 we reported a new regression equation for predicting human clearance after intravenous administration, using animal data for clearance and some molecular structural parameters, and showed that the regression equation had better predictive performance than allometric approaches that

Data

CLPO data for rats (CLPO,rat; range, 0.300–27778 mL/min/kg), dogs (CLPO,dog; range, 0.150–1598 mL/min/kg), and humans (CLPO,man; range, 0.150–3767 mL/min/kg) after oral administration of 87 drugs were obtained from the literature.6., 7., 8., 9., 10., 11., 12., 13., 14., 15., 16., 17., 18., 19., 20., 21., 22., 23., 24., 25., 26., 27., 28., 29., 30., 31., 32., 33., 34., 35., 36., 37., 38., 39., 40., 41., 42., 43., 44., 45., 46., 47., 48., 49., 50., 51., 52., 53., 54., 55., 56., 57., 58., 59., 60.

RESULTS

MLR analysis led to the selection of five descriptor variables as statistically significant in the best combination; they are, CLPO,rat × MW, CLPO,rat × Ha, CLPO,dog × Ha, c log P × c log P, and Ha × Ha. The values of q2 and rRMSE for the MLR model are 0.597 and 0.523, respectively.

PLS analyses were performed using five basic descriptors and 15 quadratic term descriptors. The standard PLS analysis with those 20 descriptors suggests that a one‐component PLS model is the best model according to

DISCUSSION

Predicting the pharmacokinetic profile of a drug in human is of great interest, and CLPO,man is one of the most important pharmacokinetic parameters. In the present study, we aimed to develop a method for predicting CLPO,man and tested the predictability of our method by comparing it with allometric approaches.

Judging from global indices such as q2 and RMSE, shown in Table 2, the predictive performance of the regression analysis methods is better than that of the allometric approaches, and the

REFERENCES (142)

  • Y. Sawada et al.

    Prediction of the disposition of β‐lactam antibiotics in human from pharmacokinetic parameter in animals

    J Pharmacokinet Biopharm

    (1984)
  • M.R. Feng et al.

    Disposition of the antipsychotic agent CI‐1007 in rats, monkeys, dogs, and human chytochrome P450 2D6 extensive metabolizers. Species comparison and allometric scaling

    Drug Metab Dispos

    (1998)
  • H. Matsushima et al.

    Pharmacokinetics and plasma protein binding of tamsulosin hydrochroride in rats, dogs, and humans

    Drug Metab Dispos

    (1998)
  • H.M. Siefert et al.

    Pharmacokinetics of the 8‐methoxyquinolone, moxifloxacin: A comparison in humans and other mammalian species

    J Antimicrob Chemother

    (1999)
  • K. Matsui et al.

    Correlation of the intrinsic clearance of donepezil (Aricept®) between in vivo and in vitro studies in rat, dog and human

    Xenobiotica

    (1999)
  • S. Higuchi et al.

    Comparative pharmacokinetics of nicardipine hydrochloride, a new vasodilator, in various species

    Xenobiotica

    (1980)
  • T. Sakai et al.

    Pharmacokinetics of an antiallergic agent, 1‐(2‐ethoxyethyl)‐2‐(hexahydro‐4‐methyl‐1H‐1, 4‐diazepin‐1‐yl)‐1H‐benzimidazole difumarate (KG‐2413) after oral administration: Interspecies differences in rats, guinea pigs and dogs

    J Pharmacobio‐Dyn

    (1989)
  • C. Joukhadar et al.

    Pharmacokinetics of emedastine difumarate, a new anti‐histaminic agent in patients with renal impairment

    Eur J Clin Pharmacol

    (2001)
  • Y. Yamada et al.

    Metabolic fate of the new angiotensin‐converting enzyme inhibitor imidapril in animals

    Arzneim‐Forsch/Drug Res

    (1992)
  • J.F. Hoogkamer et al.

    Pharmacokinetics of imidapril and its active metabolite imidaprilat following single dose and during steady state in patients with chronic renal failure

    Eur J Clin Pharmacol

    (1998)
  • T. Izumi et al.

    Prediction of the human pharmacokinetics of troglitazone, a new and extensively metabolized antidiabetic agent, after oral administration, with an animal scale‐up approach

    J Pharmacol Exp Ther

    (1996)
  • M.A. Sayed et al.

    Pharmacokinetics of ethosuximde in the dog

    Arch Int Pharmacodyn

    (1978)
  • K. Bachmann et al.

    Single sample estimate of ethosuximide clearance

    Int J Clin Pharmacol Ther Toxicol

    (1986)
  • K. Matsumoto et al.

    Absorption, distribution and excretion of 3‐(sulfamoyl[14C]methyl)‐1,2‐benzisoxazole (AD‐810) in rats, dogs, and monkeys and of AD‐810 in men

    Arzneim‐Forsch/Drug Res

    (1983)
  • G. Montay et al.

    Absorption, distribution, metabolic fate, and elimination of pefloxacin mesylate in mice, rats, dogs, monkeys, and humans

    Antimicrob Agents Chemother

    (1984)
  • Y. Tanaka et al.

    Toxicokinetics of zenerestat, an aldose reductase inhibitor in animals and man

    Xenobiotica

    (1994)
  • J.A. Scatina et al.

    Metabolic disposition and pharmacokinetics of pelrinone, a new cardiotonic drug, in laboratory animals and man

    Eur J Drug Metab Pharmacokinet

    (1990)
  • J. Kawakami et al.

    Prediction of brain delivery of ofloxacin, a new quinolone, in the human from animal data

    J Pharmacokinet Biopharm

    (1994)
  • Y. Kanai et al.

    Pharmacokinetics of SY5555, a novel oral penem antibiotics, in animals

    Chemotherapy (Tokyo)

    (1994)
  • Y. Kobayashi et al.

    Bacteriological and clinical studies on SY5555

    Chemotherapy (Tokyo)

    (1994)
  • K. Yamashita et al.

    Pharmacokinetics of loracarbef, a new oral carbacephem, in experimental animals

    Chemotherapy (Tokyo)

    (1993)
  • K. Shiba et al.

    Phase 1 study of loracarbef

    Chemotherapy (Tokyo)

    (1993)
  • H. Akiyama et al.

    Pharmacokinetics of grepafloxacin (1). Absorption, distribution and excretion after oral administration of grepafloxacin in animals as determined by HPLC

    Chemotherapy (Tokyo)

    (1995)
  • Nakashima M, Uematsu T, Nagashima S, Kosuge K, Kanamaru M. Phase 1 study of grepafloxacin. Chemotherapy (Tokyo)...
  • T. Nakagawa et al.

    Pharmacokinetics of the new fluoroquinolone balofloxacin in mice, rats and dogs

    Arzneim‐Forsch/Drug Res

    (1995)
  • O. Kozawa et al.

    Comparative study of pharmacokinetics of two new fluoroquinolones, balofloxacin and grepafloxacine, in elderly subjects

    Antimicrob Agents Chemother

    (1996)
  • A. Saito

    Pharmacokinetic study on cefixime

    Chemotherapy (Tokyo)

    (1985)
  • T. Yoshida et al.

    Absorption, excretion and metabolism of cefaclor in animals

    Chemotherapy (Tokyo)

    (1979)
  • T. Kamiki et al.

    Phase 1 study of cefaclor

    Chemotherapy (Tokyo)

    (1979)
  • H. Sakamoto et al.

    Pharmacokinetics of cefdinir, a new oral cephalosporin, in experimental animals

    Chemotherapy (Tokyo)

    (1989)
  • K. Shimada et al.

    Phase 1 study of cefdinir

    Chemotherapy (Tokyo)

    (1989)
  • T. Suwa et al.

    Metabolic fate of TE‐031 (A‐56268). (1) Comparative pharmacokinetics of TE‐031 and erythromycin stearate in rats and mice

    Chemotherapy (Tokyo)

    (1988)
  • Y. Mizushima et al.

    General clinical study on TE‐031 (A‐56268)

    Chemotherapy (Tokyo)

    (1988)
  • E. Vilmănyi et al.

    Clarithromycin pharmacokinetics after oral administration with or without fasting in crossbred beagles

    J Small Anim Pract

    (1996)
  • T. Yasuda et al.

    Absorption, distribution, metabolism and excretion of T‐3263 in experimental animals

    Chemotherapy (Tokyo)

    (1988)
  • M. Nakashima et al.

    Phase 1 study of T‐3262, a new pyridone‐carboxylic acid derivative

    Chemotherapy (Tokyo)

    (1988)
  • S. Kawamura et al.

    Pharmacokinetics of a novel quinolone, AT‐4140, in animals

    Antimicrob Agents Chemother

    (1990)
  • J. Shimada et al.

    Clinical pharmacokinetics of sparfloxacin

    Clin Pharmacokinet

    (1993)
  • D.K. Walker et al.

    Pharmacokinetics and metabolism of sildenafil in mouse, rat, rabbit, dog and man

    Xenobiotica

    (1999)
  • K.C. Beaumont et al.

    Pharmacokinetics and metabolism of darifenacin in the mouse, rat, dog and man

    Xenobiotica

    (1998)
  • Cited by (28)

    • PhRMA CPCDC initiative on predictive models of human pharmacokinetics, part 1: Goals, properties of the PhRMA dataset, and comparison with literature datasets

      2011, Journal of Pharmaceutical Sciences
      Citation Excerpt :

      These prediction methods use a variety of in vitro and/or in vivo preclinical data. A thorough comparison of these prediction methods using preclinical and clinical datasets has been performed by several authors.6–24 However, these datasets tend to include older compounds with properties differing from current discovery drug pipelines, with the exception of the novel dataset published by Hosea et al.17 Furthermore, there has been no initiative to our knowledge where all types of methods have been evaluated for both the prediction of human PK properties and the simulation of the corresponding plasma concentration–time profiles using a representative dataset.

    • Estimation of the extent of oral absorption in animals from oral and intravenous pharmacokinetic data in drug discovery

      2009, Journal of Pharmaceutical Sciences
      Citation Excerpt :

      These models are particularly useful in the lead optimization stage of drug discovery in order to avoid advancing drug candidates with poor oral bioavailability, which would complicate the development process. Many oral absorption models are now available including in vitro,5–10 in silico,11–18 in situ,19 physiological,20 and combinations thereof.21 These models are being used to screen for the potential of NMEs to be orally absorbed, as one of the two determinants of oral bioavailability.

    • Chapter 19 Human Exposure and Dose Projections

      2008, Annual Reports in Medicinal Chemistry
      Citation Excerpt :

      As oral exposure of drugs is governed by oral clearance (CL/F), several studies have evaluated the success of predicting oral clearance in humans. In a study by Wajima et al. [59], a partial least squares regression model, using animal oral clearance data from only two species and calculated molecular structural parameters (MW, c log P, and HA) were found to predict human oral clearance better than the allometric approaches. In another study by Feng et al. [60], human CL/F was more accurately estimated using unbound drug concentration and incorporating brain weight in the allometric relationship.

    View all citing articles on Scopus
    View full text