Clinical Pharmacokinetics
`
`Concepts and Applications
`
`third edition
`
`
`
`
`
`
`
`
`
`lM Roi...-'LA:N_ ;
`
`
`
`
`
`AstraZeneca Exhibit 2170 p. 1
`InnoPharma Licensing LLC V. AstraZeneca AB
`IPR2017-00905
`
`

`

`Clinical Pharmacokineiics
`
`Concepts and Applications
`
`AstraZeneca Exhibit 2170 p. 2
`
`

`

`
`
`
`
`Clinical Pharmacokinetics
`
`Conceptsand Applications
`
`third edition
`
`MAlCOl/Vl ROWLAND, PhD.
`
`Department at Pharmacy
`
`University at Manchester
`
`Manchester, England
`
`THOMAS N. TOZER, PhD.
`
`School at Pharmacy
`
`University at California
`
`San Francisco, California
`
`A Lea & Febiger Book
`
`Williams & Wilkins
`BALTIMORE ' PHILADELPHIA - HONG KONG
`LONDON - MUNICH 0 SYDNEY - TOKYO
`A WAVERLY COMPANY
`
`AstraZeneca Exhibit 2170 p. 3
`
`

`

`
`
`1
`
`Executive Editor: Donna Balaclo
`Developmental Editors.- Fraiices Klass, Lisa Stead
`Production Manager: Laurie Forsyth
`Project Editor: Robert D. Magee
`
`Copyright © 1995
`Williams & Wilkins
`Rose Tree Corporate Center
`1400 North Providence Road
`Building 11, Suite 5025
`Media, PA 19063—2045 USA
`
`
`
`All lights reserved. This book is protected by copyright, No part of this book may be reproduced in any
`form or by any means, including photocopying, or utilized by any information storage and retrieval system
`without written permission from the copyright owner.
`
`Accurate indications, adverse reactions, and dosage schedules for drugs are provided in this book, but it is
`possible they may change. The reader is urged to review the package information data of the manufacturers
`of the medications mentioned.
`
`Printed in the United States of America
`
`First Edition 1980
`
`Library of Congress Cataloging—in-Publication Data
`
`Rowland, Malcolm.
`Clinical Pharmacokinetics : concepts and applications / Malcolm
`Rowland, Thomas N. Tozer. — 5rd ed.
`p.
`cm.
`“A Lea 8: Febiger Book.”
`Includes bibliographical references and index.
`ISBN 0—683-07404—0
`1. Pharmacokinetics.
`H. Title.
`
`2. Chemotherapy.
`
`I. Tozer. Thomas N.
`
`2. Ding Therapy.
`
`QV 58 R883c 1994]
`
`[DNLM: 1. Pharmacokinetics.
`RM301.5.R68
`1994
`615.7—dc20
`DNLM/DLC
`
`for Library of Congress
`
`94—26305
`CIP
`
`The Publishers have made every efirort to trace the copyright holdersfor borrowed material. If they have in—
`adverteritly overloo/eed any, they will be pleased to malee the necessary arrangements at thefits! opportunity,
`
`969798
`2545678910
`
`Reprints of chapters may be purchased from Williams 8; Wilkins in quantities of 100 or more. Call Isabella Wise, Special Sales Department, (800) 358—3583.
`
`
`
`AstraZeneca Exhibit 2170 p. 4
`
`

`

`
`
`PREFACE
`
`PURPOSE OF TEXT
`
`The third edition, in keeping with the first two editions, is a primer in pharmacokinetics
`with an emphasis on clinical applications. The book should be useful to any student, prac-
`titioner, or researcher who is interested or engaged in the development, evaluation, or use
`of medicines. Such persons include pharmacists, physicians, veterinarians, pharmaceutical
`scientists, toxicologists, analytical chemists, biochemists, and clinical chemists. It is an in-
`troductory text and therefore presumes that the reader has little or no experience or knowl-
`edge in the area. Previous exposure to certain aspects of physiology and pharmacology
`would be helpful, but it is not essential. Some knowledge of calculus is also desirable.
`Our intent is to help the reader learn to apply pharmacokinetics in therapeutics. To this
`end, we emphasize concepts through problem solving with only the essence of required
`mathematics. In this respect, the book is a programmed learning text. At the beginning of
`each chapter, objectives are given to identify the salient points to he learned. To further
`aid in learning the material, examples are worked out in detail in the text. At the end of
`each chapter, except the first, there are problems that allow the reader to grasp the concepts
`of the chapter and to build on material given in previous chapters. The order of the prob—
`lems in each chapter reflects consideration of both difficulty and how well the problems
`apply to chapter principles. The questions start with the less difficult ones and those that
`emphasize the principles.
`
`ORGANIZATION AND CONTENT
`
`As in the second edition, the book is divided into five sections: Absorption and Disposition
`Kinetics, Therapeutic Regimens, Physiologic Concepts and Kinetics, Individualization, and
`Selected Topics. Those wishing to gain a general overview of the subject need only study
`Sections One and Two, together with Chapter 13, Variability, and Chapter 18, Monitoring.
`Section Three deals with the physiologic concepts relevant to an understanding of the
`processes of absorption, distribution, and elimination. This section forms the basis for an
`appreciation of the material in Section Four, which is concerned with the identification,
`description, and accounting of variability in patients’ responses to drugs. Covered here are
`general aspects of variability, followed by considerations of genetics, age and weight, dis-
`ease, interacting drugs, and monitoring of drug concentrations.
`Section Five contains selected topics. These are intended for those readers who wish to
`gain a more detailed insight into various aspects of clinical pharmacokinetics. The topics
`are distribution kinetics, pharmacologic response, metabolite ldnetics, dose and time de—
`pendencies, turnover concepts, and dialysis. Each topic is generally self—contained; they
`have not been arranged in any particular sequence.
`
`CHANGES IN THIRD EDITION
`
`The 6—year gap between this third edition and the second, published in 1989, is shorter
`than the 9 years between the second and first editions. This shortening of the time span 1
`
`AstraZeneca Exhibit 2170 p. 5
`
`

`

`
`
`viii
`
`PREFACE
`
`between editions reflects the ever—gathering pace of progress and application of clinical
`pharmacokinetics. Despite this growth, which has required the inclusion of much new
`material, every effort has been made to contain the overall size of the book. This, in turn,
`has meant that some material has had to be condensed or deleted. It has also resulted in
`a much greater use of abbreviations, especially for units.
`The number, topic, and sequence of chapters have been kept essentially the same as in
`the second edition. However, each chapter has been extensively revised and updated to
`ensure that the examples relate to currently prescribed drugs. A particular effort has been
`made to include stereochemistry, recognizing that isomers may have different kinetics and
`activity. There is also consideration of the increasing number of polypeptide and protein
`drugs emerging from advances in molecular biology and biotechnology. Although the ki—
`netic concepts are the same, the physiologic handling of macromolecular compounds is
`quite distinct from that of typical small molecular weight drugs.
`The presentation of the book has also been markedly improved through the use of color.
`The more important equations are now highlighted by means of color. Chapter number
`and section heading now appear at the top of each page layout to assist in cross-referencing.
`A table of frequently used symbols has been placed before Chapter 1 to facilitate redefining
`symbols, when necessary.
`The range and number of problems at the end of each chapter and Appendix I (total of
`87 new problems) have been substantially extended to assist in learning problem solving
`in pharmacoldnetics. Most of the additional problems are taken from literature, rather than
`simulated, data.
`The third edition contains 102 new figures and 20 new tables, reflecting, in large part,
`the advances made in recent years in our knowledge of the pharmacokinetics of drugs. The
`mateiial on “Small Volume of Distribution” that comprised the last chapter of the second
`edition has been incorporated into Chapter 10, Distribution, and Appendix I—F.
`We continue to adopt a uniform set of symbols and to use milligrams/liter (mg/L) as
`the standard measure of concentration. We do recognize, however, the increasing trend
`toward the adoption of molar units and have provided a factor for conversion between the
`two units of measurement in the pertinent figure captions. We shall only be convinced of
`the virtue of solely using the molar system of measurement when drugs are prescribed in
`such units.
`
`ACKNOWLEDGMENTS
`
`We wish to thank all the many students and readers who provided input that helped us
`shape this third edition. Their enthusiasm and encouragement have been a continual source
`of satisfaction. To the new reader, we hope that the book will succeed in helping you
`develop kinetic reasoning that will be of personal value in your professional practice.
`We have been enormously gratified by the wide and diverse readership of the first two
`editions of the book. We would like to believe that the book has been instrumental in
`furtherng rational management of drug therapy. We sincerely hope that the third edition
`will continue to do so.
`
`Manchester, England
`San Francisco, California
`
`Malcolm Rowland
`Thomas N. Tozer
`
`AstraZeneca Exhibit 2170 p. 6
`
`

`

`
`
`CONTENTS
`
`.......................................................................... .. xi
`Detinitions ot Symbols
`. Why Clinical Pharmacokinetics?
`............................................................. ..T
`
`T
`
`SECTION I. ABSORPTION AND DISPOSITION KINETICS
`2. Basic Considerations
`....................................................................... ..T T
`3.
`Intravenous Dose ............................................................................. ..T8
`4. Extravascular Dose .......................................................................... ..34
`
`THERAPEUTIC REGIMENS
`SECTION II.
`5. Therapeutic Response and Toxicity ...................................................... ..53
`o. Constant-Rate Regimens .................................................................... ..oo
`7. Multiple-Dose Regimens .................................................................... ..83
`
`PHYSIOLOGIC CONCEPTS AND KINETICS
`SECTION III.
`8. Movement Through Membranes
`....................................................... ..TOQ
`9. Absorption ................................................................................... ..T T9
`0. Distribution
`.................................................................................. ..T37
`T T. Elimination ................................................................................... .. T 56
`2.
`Integration With Kinetics
`................................................................. .. T 84
`
`INDIVIDUAIJZATION
`SECTION IV.
`1 3. Variability .................................................................................... ..203
`4. Genetics
`..................................................................................... ..220
`'5. Age and Weight
`.......................................................................... ..230
`6. Disease ....................................................................................... ..248
`7.
`Interacting Drugs
`........................................................................... .267
`8. Concentration Monitoring ............................................................... ..290
`
`
`
`SEIECTED TOPICS
`SECTION V.
`9. Distribution Kinetics ........................................................................ ..3T 3
`20. Pharmacologic Response ................................................................ ..340
`2T . Metabolite Kinetics
`.................................... .5................................... ..3o7
`............................. H394
`22. Dose and Time Dependencies
`...................
`
`....................................... ..‘Z.::'I ........................... ..42—-
`23. Turnover Concepts
`................... .......—-43
`24. Dialysis
`............................................................
`
`SELECTED READING .
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`. ..—~O3
`
`APPENDIX I. ADDITIONAL CONCEPTS AND DERIVATIONS
`A. Assessment otAUC ....................................................................... ..Z-OQ _
`B. EstimatiOn ot Elimination Halt-lite From Urine Data
`................................ ..473 '
`
`
`
`AstraZeneca Exhibit 2170 p. 7
`
`

`

`X
`
`CONTENTS
`
`.......... ..—-78
`C. Estimation of Absorption Kinetics From Plasma Concentration Data
`D. Mean Residence Time .................................................................... ..—-85
`E. Amount ot Drug in Body on Accumulation to Plateau ............................. “1-90
`F. Distribution at Drugs Extensively Bound to Plasma Proteins
`...................... ..—-QZ-
`(3. Blood to Plasma Concentration Ratio ................................................. .502
`H. Estimation ol Creatinine Clearance Under Nonsteady—State Conditions ..... _.50—-
`
`
`
`
`
`APPENDIX II. ANSWERS T0 PROBLEMS
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`. .607
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`. .586
`
`INDEX .
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`.
`
`
`
`AstraZeneca Exhibit 2170 p. 8
`
`

`

`
`
`VARIABILITY
`
`OBJECTIVES
`
`The reader will be able to:
`
`i . List six major sources of variability in drug response.
`2. Evaluate whether variability in drug response is caused by a variability in pharmacokinetics,
`pharmacodynamics, or both, given response and pharmacokinetic data.
`3. State why variability around the mean and shape of the frequency distribution histogram of
`a parameter are as important as the mean itself.
`4, Explain how variability in hepatic enzyme activity manifests itself in variability in both phar-
`macokinetic parameters and plateau plasma drug concentrations for drugs of high and low
`hepatic extraction ratios.
`
`5. Suggest an approach for initiating a dosage regimen for an individual patient, given patient
`population pharmacokinetic data and the individuals measurable characteristics,
`
`Thus far, the assumption has been made that all people are alike. True, as a species, humans
`are reasonably homogeneous, but differences among people do exist including their re—
`sponsiveness to drugs. Accordingly, there is a frequent need to tailor drug administration
`to the individual patient. A failure to do so can lead to ineffective therapy in some patients
`and toxicity in others.
`This section of the book is devoted to individual drug therapy. A broad overview of the
`subject is presented in this chapter. Evidence for and causes of variation in drug response,
`and approaches toward individualizing drug therapy are examined. Subsequent chapters
`deal in much greater detail with genetics (Chap. 14), age and weight (Chap. 15), disease
`(Chap. 16), interactions between drugs within the body (Chap. 17), and monitoring of
`plasma concentration of a drug as a guide to individualizng drug therapy (Chap. 18).
`Before proceeding, a distinction must be made between an individual and the popula—
`tion. Consider, e.g., the results of a study designed to examine the contribution of an acute
`disease to variability in drug response. Suppose, of 30 patients studied during and after
`recovery, only 2 showed a substantial difference in response; in the remainder the differ—
`ence was insignificant. Viewed as a whole, the disease would not be considered as a sig-
`nificant source of variability, but to the two affected patients itwould. Moreover, to avoid
`toxicity, the dosage regimen of the drug may need to be reduced in these two patients
`during the disease. The lesson is clear: Average data are useful as a guide; but ultimately,
`information pertaining to the individual patient is all—important.
`On a similar but broader point, substantial differences in response to most drugs exist
`among patients. Such interindividual variability is often reflected by a variety of marketed
`dose strengths of a drug. Because variability in response within a subject (intmindividual)
`is generally much smaller than interindividual variability, once well-established, there is
`
`203
`
`AstraZeneca Exhibit 2170 p. 9
`
`

`

`204
`
`VARIABILITY
`
`CHAPTER 1 3
`
`usually little need to subsequently adjust an individual’s dosage regimen. Clearly, if int/rain—
`dividual variability were large and unpredictable, trying to titrate dosage for an individual
`would be an extremely difficult task, particularly for drugs with narrow therapeutic win—
`dows. Stated differently, a drug that exhibits a high intmindividual variability in pharma—
`cokinetics can be prescribed only if it has a wide therapeutic window.
`
`EXPRESSIONS OF INDIVIDUAL DIFFERENCES
`
`Evidence for interindividual differences in drug response comes from several sources.
`Variability in the dosage required to produce a given response is illustrated in Figure 1—5
`(Chap. 1), which shows the wide range in the daily dose of warfarin needed to produce a
`similar degree of anticoagulant control. Variability in the intensity of response with time
`to a set dose is seen with the neuromuscular agent doxacurium (Fig. 13—1). As illustrated
`in Figs. 13—2, and 13—3, which Show frequency distribution histograms of the plateau
`plasma concentration of the antidepressant drug nortriptyline, to a defined daily dose of
`the drug and the plateau unbound plasma concentration of warfarin required to produce
`a similar degree of anticoagulant control, variability exists in both pharmacolcinetics and
`pharmacodynamics. Variability in pharmacokinetics was also illustrated by the wide scatter
`in the plateau plasma concentration of phenytoin seen following various daily doses of this
`drug (see Fig. 1—6, Chap. 1).
`
`The Need for Models
`
`The magnitude and relative contribution of pharmacokinetics and pharmacodynamics to
`variability in response to a given dosage within a patient population vary with the drug and,
`to some extent, the condition being treated. For example, with a nonsteroidal anti-inflam—
`matory drug, the relative contribution of pharmacodynamic variability may be different
`when the endpoint is the relief of a headache than when it is the relief from chronic aches
`and pains associated with inflamed joints. In clinical practice, attempts to assign the relative
`contribution to pharmacokinetics and pharmacodynamics may be made based on direct
`observations of plasma concentration and response. The assignment could be strongly in-
`fluenced, however, by the timing of the observations and the magnitude of the response,
`
`100 —
`
`Fig. 13—1. The degree of neuromus— x
`cular blockage with time after an iv b0-
`8
`lus dose of 0.04 mg/kg doxacurium to pa— 5
`tients varies widely. (1 mg/L = 0.97 uM) E
`(Modified from Schmith, V.D., Fiedler— §
`Kelly,
`Abou—Donia, M., Huffman,
`6-:
`C.S., and Grasela, T.H.: Population
`pharmacodynamics of doxacumin. Clin.
`Pharmacol. Ther., 52:528e536, 1992.)
`
`75 -
`
`_
`
`
`
`25 ‘
`
`0
`
`_I_
`
`‘I—
`
`I
`
`—I
`
`0
`
`80
`
`160
`Minutes
`
`240
`
`320
`
`
`
`AstraZeneca Exhibit 2170 p. 10
`
`

`

`CHAPTER 1 3
`
`VARIABILITY
`
`205
`
`as illustrated in Fig. 13—4. Here, a drug that displays little interpatient variability in Cm“,
`tmx and in maximum effect, but large variability in half—life and concentration needed to
`produce 50% maximum response, is given orally at two doses, one that achieves close to
`maximal response in all patients and one that does not. At the higher dose, observations
`made at Cmax would suggest little variability in either concentration or pharmacodynamics,
`with perhaps a greater assignment of variability to the former, as variation in plasma con—
`centration produces relatively little change in response. At later times after this higher
`dose, substantial variability is observed in both concentration and response. In contrast, for
`
`99
`
`
`
`3 90
`a
`a
`E 50
`e
`E
`g 10
`§
`
`1
`
`30
`
`:2
`E
`3 20
`‘5
`32E”
`g 10
`
`0
`
`
`
`i
`
`l
`
`r
`
`l
`
`r
`l
`
`1
`
`l
`
`0
`
`0.1
`
`0.2
`
`0.3
`
`0.01
`
`0.05
`
`0.1
`
`0.5
`
`Plasma Nortriptyline Concentration (mg/L)
`
`Plasma Nortriptyline Concentration
`(mg/L, log scale)
`
`Fig. 13—2. A, The plateau plasma concentration of nortriptyline varies widely in 263 patients receiving a regimen
`of 25 mg nortriptyline orally three times daily. B, The concentrations are log—normally distributed, as seen from
`the straight line, when the percentiles of the cumulative number of patients are plotted on probit scale against
`the logarithm of the concentration. (1 mg/L = 3.8 uM) (Redrawn and calculated from Sjoqvist, F., Borga, 0.,
`and Orme, M.L.E.: Fundamentals of clinical pharmacology. In Drug Treatment. Edited by 0.8. Avery. Edinburgh,
`Churchill Livingstone, 1976, pp. 1—42.)
`
`0
`
`'
`
`2
`
`4
`
`Unbound Plasma S—Warfarin
`
`Concentration (Mg/L)
`
`Fig. 13—3. The unbound plateau concentration of the predominately active S-warfarin associated with a similar
`degree of anticoagulation, varies widely among a group of 38 patients receiving racemic warfarin. (1 mg/L = 3.3
`11M) (Adapted from Chan, 13., McLachlan, A.]., Pegg, M., Mackay, A.D., Cole, R. B., and Rowland, M.: Disposition
`of warfarin enantiomers and metabolites in patients during multiple dosing. Br. J. Clin. Pharmacol, 37:563—569,
`1994.
`
`L——
`
`AstraZeneca Exhibit 2170 p. 11
`
`

`

`206
`
`VARIABILITY
`
`CHAPTER l 3
`
`0
`
`6
`
`12
`Hours
`
`18
`
`24
`
`15
`
`10
`
`5
`
`0
`
`C
`.2
`4.,
`E
`E);
`8
`3‘
`=—
`3 0.08
`E
`W
`5—5 0.06
`
`0.04
`
`0.02
`
`0
`
`1
`
`0.75
`
`0'5
`
`0
`
`A
`a 0.25
`E
`D.
`3:3
`E
`E
`"E
`E
`E 0.15
`3
`o
`E
`I 0'1
`
`0'05
`
`0
`
`0
`
`6
`
`12
`Hours
`
`18
`
`24
`
`Large
`Dose
`
`Small
`Dose
`
`0
`
`6
`
`12
`Hours
`
`18
`
`24
`
`0
`
`6
`
`18
`
`24
`
`12
`Hours
`
`Fig. 13—4. The inten'ndividualvaiiability in concentration and response varies with dose and time of observation.
`Shown are plasma concentrations (left) and responses (right) following large and small doses of a drug that displays
`little interpatient variability in Cum, tum and maximum response, but large interpatient variability in half—life and
`concentration needed to produce 50% maximum response. High dose (top): at tum, the maximum response in all
`patients is produced with little variability in either CW”. or response. Greater variability in concentration and
`response is seen at later times. Low dose (bottom): at tmm, variability in CW”. is still low, but that in response is
`now considerable.
`
`NIHX’
`
`but now there is
`the lower dose, at tum. there is still little interpatient variability in C
`considerable variability in response. This dependence on close and time in the assignment
`of variability is minimized by expressing variability not in terms of observations but rather
`in terms of the parameter values defining pharmacoldnetics and pharmacodynamics, that
`is, in F, ka, CL, and V for pharmacokinetics, and in maximal response, concentration to
`achieve 50% of the maximum response, and the factor defining the steepness of the con-
`centration—response relationship for pharmacodynamics (Chap. 20, Pharmacologic Re—
`sponse). Once variability in these parameters is defined, the expected variability in con—
`centration and response within the patient population associated with given dosage
`regimens can be estimated. The accuracy of the models defining pharmacokinetics and
`pharmacodynamics is obviously critical to an understanding of variability in patient re—
`sponse. Where appropriate, these models should incorporate such factors as protein bind—
`ing, active metabolites, and tolerance.
`
`DESCRIBING VARIABILITY
`
`Knowing how a particular parameter varies Within the patient population is important in
`therapy. To illustrate this statement consider the frequency distributions in clearance of
`
`;
`
`—————‘
`
`AstraZeneca Exhibit 2170 p. 12
`
`

`

`CHAPTER 1 3
`
`VARIABILITY
`
`207
`
`the three hypothetical drugs shown in Fig. 13—5. The mean, or central tendency, for all
`three drugs is the same, but the variability about the mean is very different. For Drugs A
`and B, the distribution is unimodal and normal; here the mean represents the typical value
`of clearance expected in the population. As variability about the mean is much greater for
`Drug B than for Drug A, one has much less confidence that the mean of Drug B applies
`to an individual patient. For Drug C, distribution in clearance is bimodal, signifying that
`there are two major groups within the population: those with high and low clearances.
`Obviously, in this case, the mean is one of the most unlikely values to be found in this
`population.
`Generally, distributions of pharmacokinetic parameters or observations are unimodal
`rather than polymodal, and they are often skewed rather than normal, as seen, e.g., in the
`frequency distribution of plateau plasma concentrations of nortriptyline (Fig. 13—2A). A
`more symmetrical distribution is often obtained with a logarithmic transformation of the
`parameter; such distributions are said to be log-normal. A common method of examining
`for log—normal distribution is to plot the cumulative frequency, or percentile, on a probit
`scale against the logarithm of the variable. The distribution is taken to be log-normal if the
`points lie on a straight line. As can be seen in Fig. 13—2B, this is the case for the plateau
`plasma concentration of nortriptyline. In such cases the median, or value above and below
`which there are equal numbers, differs from the mean. For nortriptyline, examination of
`Fig. 13—2B indicates that the median concentration is 0.05 mg/L, which is less than the
`average value of 0.069 mg/L.
`A comment on the quantitation of variability is needed here. Variance is a measure of
`the deviations of the observations about the mean; it is defined as the sum of the squares
`of these deviations. While useful to convey variability within a particular set of observations,
`variance does not allow ready comparison of variability across sets of observations of dif-
`ferent magnitude. Suppose, e.g., clearance in an individual is 50 mL/min and the mean is
`100 mL/min; the squared deviation is 2500 (mL/min)? If instead clearance had been
`quoted in L/min, the squared deviation would be (0.05 — 0.1)2, or 0.0025 (L/min)? Coeffi-
`cient of variation, which expresses variability with respect to the mean value, overcomes this
`problem. Specifically, it is the square root of variance (the standard deviation) normalized to
`the mean. In the example above, the deviation normalized to the mean is 0.5 and is independ—
`ent of the units of clearance. Furthermore, a large coefficient ofvariation now always signifies
`a high degree of variability. Subsequently, in the book, high and low variability refer to dis-
`tributions that have high and low coefficients of variation, respectively.
`
`Fig. 13—5. As the frequency distributions for the
`clearance of three hypothetical drugs (A, B, C) show,
`it is as important to define variability around the mean
`and the shape of the frequency distribution curve as
`it is to define the mean itself.
`
`
`
`Frequency
`
`Clearance (arbitrary units)
`
`
`
`AstraZeneca Exhibit 2170 p. 13
`
`

`

`208
`
`VARIABIUTY
`
`CHAPTER 1 3
`
`WHY PEOPLE DIFFER
`
`The reasons why people differ in their responsiveness to drugs in medicinal products are
`manifold and include, in general order of importance, genetics, disease, age, drugs given
`concomitantly, and a variety of environmental factors. Although inheritance accounts for a
`substantial part of the differences in response among individuals, much of this variability
`is largely unpredictable. Increasingly, however, this source of variability, particularly that
`related to drug metabolism, is being understood and made more predictable using the tools
`of molecular biology (Chap. 14, Genetics).
`Disease can be an added source of variation in drug response. Usual dosage regimens
`may need to be modified substantially in patients with renal function impairment, hepatic
`disorders, congestive cardiac failure, thyroid disorders, gastrointestinal disorders, and other
`diseases. The modification may apply to the drug being used to treat the specific disease
`but may apply equally well to other drugs the patient is receiving. For example, to prevent
`excessive accumulation and so reduce the risk of toxicity, the dosage of the antibiotic
`gentamicin used to treat a pleural infection of a patient must be reduced if the patient also
`has compromised renal function. Similarly, hyperthyroidic patients require higher than
`usual doses of digoxin, a drug used to improve cardiac efficiency. Moreover, a modification
`in dosage may arise not only from the direct impairment of a diseased organ but also from
`secondary events that accompany the disease. Drug metabolism, e. g., may be modified in
`patients with renal disease; plasma and tissue binding of drugs may be altered in patients
`with uremia and hepatic disorders.
`Age, weight, and concomitantly administered drugs are important because they are
`sources of variability that can be taken into account. Gender-linked differences in hormonal
`balance, body composition, and activity of certain enzymes manifest themselves in differ—
`ences in both pharmacokinetics and responsiveness, but overall, the effect of gender is
`small.
`
`Table 13—1 lists examples of additional factors known to contribute to variability in drug
`response. Perhaps the most important factor is noncompliance. N oncompliance includes
`the taking of drug at the wrong time, the omission or supplementation of prescribed dose,
`and the stopping of therapy, either because the patient begins to feel better or because of
`development of side-effects that the patient considers unacceptable. Whatever the reason,
`these problems lie in the area of patient counselling and education. Occasionally, plasma
`concentration data are used as an objective measure of noncompliance.
`Pharmaceutical formulation and the process used to manufacture a product can be
`important as both can affect the rate of release, and hence entry, into the body (Chap. 9).
`
`Table 1 3-1 . Additional Factors Known to Contribute to Variability
`
`in Drug Response
`
`
` fACTORS OBSERVATlONS AND REMARKS
`
`Noncompliance
`Route of administration
`
`Food
`
`Pollutants
`
`Time of day and season
`
`Location
`
`A major problem in clinical practice; solution lies in patient education,
`Patient response can vary on changing the route of administration. Not only
`pharmacokinetics of drug but also metabolite concentrations can change.
`Rate and occasionally extent of absorption are affected by eating. Effects
`depend on composition of food. Severe protein restriction may reduce the
`rate of drug metabolism.
`Drug effects are often less in smokers and workers occupationally exposed to
`pesticides; a result of enhanced drug metabolism.
`Diurnal variations are seen in pharmacokinetics and in drug response. These
`effects have been sufficiently important to lead to the development of a new
`subject, chronopharmacology.
`Dose requirements of some drugs differ between patients living in town and in
`
`the country.
`
`———4
`
`AstraZeneca Exhibit 2170 p. 14
`
`

`

`CHAPTER 1 3
`
`VARIABILITY
`
`209
`
`A well—designed formulation diminishes the degree of variability in the release character—
`istics of a drug in. viva. Good manufacturing practice, with careful control of the process
`variables, ensures the manufacture of a reliable product. Drugs are given enterally, topi-
`cally, parenterally, and by inhalation. Route of administration not only can affect the con—
`centration locally and systemically but also can alter the systemic concentration of metab—
`olite compared with that of drug (Chap. 21). All these factors can profoundly affect the
`response to a given dose or regimen.
`Food, particularly fat, slows gastric emptying and so decreases the rate of drug absorp—
`tion. Oral bioavailability is not usually affected by food, but there are many exceptions to
`this statement. Food is a complex mixture of chemicals, each potentially capable of inter-
`acting vvith drugs. Recall from Chap. 9, e.g., that the oral bioavailability of tetracycline is
`reduced when taken with milk, partly because of the formation of an insoluble complex
`with calcium. Recall also that a slowing of gastric emptying may increase the oral bioavail—
`ability of a sparingly soluble drug, such as griseofulvin. Diet may also affect drug metab—
`olism. Enzyme synthesis is ultimately dependent on protein intake. When protein intake
`is severely reduced for prolonged periods, particularly because of an imbalanced diet, drug
`metabolism may be impaired. Conversely, a high protein intake may cause enzyme induc—
`tion.
`Chronopharmacology is the study of the influence of time on drug response. Many
`endogenous substances, e.g., hormones, are known to undergo cyclic changes in concen-
`tration in plasma and tissue with time. The amplitude of the change in concentration varies
`among substances. The period of t