T H R T E E N T H EDITION
`
`Editors
`KURT J. ISSELBACHER, A.B.,
`M.D.
`Mallinckrodt Professor of Medicine. Harvard Medical
`School; Physician and Director, Cancer Center,
`Massachusetts General Hospital, Boston
`
`EUGENE BRAUNWALD, A.B.,
`M.D., M.A. (Hon.), M.D. (Hon.)
`Hersey Professor of the Theory and Practice of Physic,
`Harvard Medical School ; Chairman, Department of
`Medicine, Brigham and Women's Hospital, Boston
`
`JEAN D. WILSON, M.D.
`Charles Cameron Sprague Distinguished Chair and
`Professor of Internal Medicine; Chief, Division of
`Endocrinology and Metabolism, The University of Texas
`Southwestern Medical Center, Dallas
`
`McGRAW-HILL, Inc.
`Health Professions Division
`
`JOSEPH B. MARTIN~ M.D., Ph.D.,
`F.R.C.P. (C), M.A. (Hon.)
`Professor of Neurology and Chancellor, University of
`California, San Francisco
`
`ANTHONY S. FAUCI, M.D.
`Director, National Institute of Allergy and Infectious
`Diseases; Chief, Laboratory of lmmunoregulation; Director,
`Office of AIDS Research, National Institutes of Health,
`Bethesda
`
`DENNIS L. KASPER, M.D.
`William Ellery Channing Professor of Medicine, Harvard
`Medical School; Chief, Division of Infectious Diseases,
`Beth Israel Hospital; Co-Director, Channing Laboratory,
`Brigham and Women's Hospital, Boston
`
`New York St. Louis San Francisco Colorado Springs Auckland Bogota Caracas Hamburg Lisbon London
`Madrid Mexico Milan Montreal New Delhi Paris San Juan Sao Paulo Singapore Sydney Tokyo Toronto
`
` DRL - EXHIBIT 1024
`
`

`

`Note: Dr. Fauci's work as editor and author was performed outside
`the scope of his employment as a U.S. government employee. This
`work represents his personal and professional views and not necessarily
`those of the U.S. government.
`
`HARRISON'S
`RINCIPLES OF INTERNAL MEDICINE
`Thirteenth Edition
`
`Copyright !0 1994. 1991. 1987. 1983. 1980, 1977. 1974. 1970, 1966, 1962, 1958 by McGraw(cid:173)
`Hill. Inc. All nghts reserved. Copyright 1954 , 1950 by McGraw-Hill , Inc. All rights reserved.
`Copyright renewed 1978 by Maxwell Myer Wintrobe and George W. Thorn. Printed in the
`Unhcd States of America. Except as pem1ittcd under the United Stutes Copyright Act of 1976,
`no part of this publication may be reproduced or distributed in any form or by any means. or
`stored in a data base or retrieval system. without the prior wrillcn permission of the publisher.
`
`l 2 3 4 5 6 7 8 9 0 DOW DOW
`
`987654
`
`Foreign Language Editions
`CHINESE (T welfrh Edition)- McGraw-Hill Book Company-Singapore,
`© 1994
`FRENCH (Twelfth Edition)- Fiammarion. © 1992
`GERMAN (Tenth Edition)- Schwabc and Company, Ltd., © 1986
`GREEK (Twelfth Edition)- Parissianos, © 1994 (est.)
`ITALIAN (Twelfth Edition)- McGruw-Hill Libri ltalia S.r.l. © 1992
`JAPANESE (Eleventh Edition)- Hirokawa. © 1991
`PORTUGUESE (Twelfth Edition)- Editora Guanabara Koogan, S.A ..
`© 1992
`SPANISH (Twelfth Edition)- McGruw-Hillllnteramericana de Espana,
`© 1992
`
`This book was set in Times Roman by Monotype Composition Company.
`The editors were J. Dereck Jeffers and Stuart D. Boynton. The indexer
`was Irving Tullar; the production supervisor was Roger Kasunic; the
`designer was Marsha Cohen; R. R. Donnelley & Sons Company was
`printer and binder.
`
`Library of Congress Cataloging-in-Publication Data
`
`Harrison's principles of internal medicine-13th ed./editors,
`Kurt J . lsselbacher . . . [el al.]
`p.
`em.
`Includes bibliographical references and index.
`ISBN 0·07 -032370-4 (1-vol. ed.) ; 98.00 -
`ISBN 0-07-911 169-6 (2
`vol. ed. set): 127.00 -ISBN 0-07-032371-2 (bk. 1). -
`ISBN
`0-07-032372-0 (bk. 2)
`1. Internal medicine.
`11. lsselbacher. Kurt J.
`medicine.
`[DNLM: 1. Internal Medicine. WB 115 P957 1994]
`RC46.H333 1994
`616-dc20
`DNLM/DLC
`for Library of Congress
`
`I. Harrison, Tinsley Randolph, 1900-
`Ill. Title: Principles of internal
`
`93-47393
`CIP
`
`

`

`

`

`

`

`

`

`396
`
`Plllll I'OUII CLINil.Al I'IIARMACoJI.Of'Y
`
`upon which to estimate changes in drug elimination with reduction
`in creatinine clearance in renaJ insufficiency. Half-life is not linearly
`related to clearance.
`The imponant relationship
`
`0.693V"
`Ita = _C_I_
`
`indicntcs clearly the dependency of half-life, a measure of rate of
`elimination , on the two physiologically independent variables of
`volume of distribution and clearance, which expresses the efficiency
`of elimination. Thus half-life is shortened when phenobarbital induces
`the enzymes responsible for hepatic cleurance of a drug, and half-life
`is lengthened when a drug's renal clearance is auenuated in renal
`failure. Also, the half-life of some drugs is shortened when their
`volume of distribution is reduced. If. as in the case of cardiac failure,
`the volume of disuibution is reduced at the same time that clearance
`is reduced, there may be little change in drug half-life to reflect the
`impaired clearance, but steady state plasma levels will be increased,
`as is the case with lidocaine. ln treating patients after an overdose.
`the effects of hemodialysis on a drug's elimination arc dependent on
`its volume of distribution . When the volume of distribution is large,
`as with tricyclic antidepressants (V,1 of desipramine equals more than
`2000 L) , the removal of drug, even with a high-clearance dialyzer.
`proceeds slowly.
`The extent to which a drug is bound to plasma protein also
`determines the fraction extracted by the organ(s) of elimination.
`Altered binding changes the extraction ratio significantly. however.
`only when elimination is limited to the unbound (free) drug in plasma.
`The extent to which binding influences elimination depends on the
`relative affinity of tbe plasma binding versus the affinity of the drug
`for the extraction process. Tbe high affinity of the renal tubular anion
`transport system for many drugs leads to extraction of bound and
`unbound drug, and the efficient process by which the liver removes
`propranolol extracts most of this highly bound drug from blood.
`However, in the case of drugs with low organ extraction ratios, only
`unbound drug is available for elimination.
`STEADY STATE With a constant infusion of drug, the infusion
`rate equals elimination rate at steady state. Therefore,
`
`response should lend to cQnsideration of bioavaiJnbility as a possible
`facror. Calculation of a dosage regimen should be corrected for
`bioavailability:
`
`I d
`0
`ra ose =
`
`Cp •• x Cl x dosage interval
`F
`
`DRUG ELIMINATION THAT IS NOT FIRST-ORDER The e]jmina(cid:173)
`tion of some drugs such as phenytoin. salicylate, and theophylline
`does not follow first-order kinetics when amounts of drug in the body
`are in the therapeutic range. For these drugs, the clearance changes
`us levcJs in the body fall during elimination or after alterations in
`dose. This pattern of elimination is said to be dose-dependem.
`Accordingly, the time for the concentration to fall to one-half becomes
`less as plasma levels fall: thls halving time is not truly a balf-life.
`because the term half-life applies to first-order kinetics and is a
`constant. The elimination of phenytoin is dose-dependent, and when
`very high levels are present (in the toxic range), tbe halving time
`may be longer than 72 h , whereas as the concentration in plasma
`declines, the clearance increases and the concentration in plasma Will
`halve in 20 to 30 h. When n drug is eliminated by first-order kinetics,
`the plasma level at steady state is directly related to the amount of
`the maintenance dose, and a doubling of the dose should lead to
`doubling of the steady state plasma level. However, for drugs with
`dose-dependent kinetics, increases in the dose may be accompanied
`by disproportionate increases in plasma level. Thus. if the daily dose
`of phenytoin is increased [rom 300 to 400 mg, plasma levels rise by
`mof'e than 33 percent. The extent of increase is not predictable because
`of the interpatient variability in the extent ro which clearance deviates
`from first order. Theophylline and salicylaies also are eliminated by
`dose-dependent kinetics, and in child1-en particular caution must be
`taken with the administrotion of salicylates in high doses. Changes
`in dosage regimens for such drugs should always be accompanied
`by survei llance for adverse effects and by measurement of the
`concentration of the drug in plasma after sufficient time to establish
`n new steady state. Ethanol metabolism also is dose-dependent, with
`obvious implications. The mechanisms involved in dose-dependent
`kinetics may include the saturation of the rate-limitjng step in
`metabolism or a feedback inhibition of the rate-limiting enzyme by
`a product of the reaction.
`
`INDIVIDUALlZATION OF DRUG THERAPY
`
`Optimal drug therapy requires administration of just the right amount
`of drug for the particular patient-
`too little and efficacy is not likely,
`whereas too large a dose increases the risk of undesirable effects.
`When the desired response is a readily determined clinical effect,
`such ns altered blood pressure or consolation time. then an optimal
`dosage requirement can be achieved 10 an empirical fashion. Dosage
`alterations should, however. involve modest changes in amount (50
`percent) and no more frequently than every two to three half-lives.
`Ln most cases, however, drug therapy must be guided by the concept
`of a "therapeutic window" within which drug concentrations must
`be achieved and maintained. lf this therapeutic window is large, i.e ..
`little dose-relnted toxicity. then maximal efflcacy, should this be
`desired and achievable. may be obtained by administering a supraeffec(cid:173)
`tive dose. Such a strategy is often used for penicillins and many
`beta-adrenoceptor blocking agent~. It is also possible under these
`circumstances to usefully extend the duration of action of the drug.
`especially when it is rapidly cljminated from rhe body. Thus 75 mg
`of captopril will result in reduced blood presurc for up to 12 h. even
`though the elimination half-life of the ACE-inhibitor is about 2 h.
`The therapeutic window for most drugs, however, is much narrower.
`and in certain instances (see Table 66-4), as little as a twofold
`difference distinguishes the. dose (concentration) of drug producing
`the desired response from that eliciting an udverse effecl. ln these
`cu~es , the application of pharmacokinetic principles is critical to
`ac:hlcving the defined therapeutic objective.
`
`Cl
`(voVunit time)
`
`Cp ..
`(amtlvol)
`
`X
`
`fnfusion rate
`(ami/unit time)
`when the units for amount, volume, and time are consistent.
`Thus, if clearance (CJ) is known, the infusion rate to achieve a
`given steady state plasma level c:an be calculated. Estimation of drug
`clearance is discussed in the section on renal disease.
`When the dose is given intermittently instead of by infusion.
`the above relationship between plasma concentration and the dose
`administered ar each dosage iotervai can be expressed as
`Dose = Cp,., x Cl x dosage interval
`
`The average plasma concentration (Cp.,) implies, as seen in Fig.
`66-2. that levels can be higher and lower tJum the average during the
`dosage interval.
`When a drug is given orally, the fraction (F) of the administered
`dose that reaches the systemic circu lation is an expression of the
`drug's bionvailnbility. A reduction in bioav;1ilability may reflect a
`poorly formu lated dosage form that fails to disintegrate or dissolve
`in the gastrointestinal fluids. Regulatory standards have reduced the
`extent of this problem. Drug interactions also can impair absorption
`after oral dosing. Bioavailability also may be reduced due to drug
`metabolism in the gastrointestinal tract ond(or the liver during the
`absorption process, the.first-pass effect. This is a particular problem for
`drugs that are extensively extracted by these organs, and considerable
`interpalient variability often exists in bioavnilabilit.y. Lidocaine for
`the control of <~rrhythmias is not administered orally because of the
`lirst-puss effect. Drugs that are injected intramuscularly also may
`have low bioavailabiJity. e.g., phenytoin. An unexpected drug
`
`

`

`

`

`398
`
`PART FOUR Ct.INICAL PHAHW\COLOGV
`
`TAULE 66·1 Clearance of drugs
`
`Drug
`
`Renal
`clearance,+
`mUm in
`
`Nonrenal
`clcamnce.
`mUm in
`
`12
`340
`Ampicillin'
`68
`10
`Carbenicillin
`IJO
`36
`Digoxin'
`78
`3
`Gentamicin
`0
`60
`Kanamycin
`Penicillin 0 1
`36
`340
`• The ''nom1nl" renal clcnmnces are those associated with u clcaruncc of creatinine or
`100 n1Umin.
`• Tht fmction of digoxin ab~•rlxd after an ool dose (F) is approxim:uely 0. 75. and F
`for ampicillin is 05.
`t One microgl'l\m uf penicilli!! G ~ 1.6 units.
`
`may be es~imatcd from the information in Table 66-2 and the
`nomogram (Fig. 66-S). Table 66-2 gives the fraction of the usual
`dose of a drug required at n creatinine clearance of z.ero (dose
`fmetion0). Tbe nomogram presents the dose fraction as a linear
`function of creatinine clearance.
`To calculate the dose fraction,~> rhe dose fraction0 is obtained from
`Table 66·2. plotted on the left ordinate of the nomogram, and
`connected by a strulght line to the upper right-hand corner of the
`nomogram. This line describes the dose fraction over a range of
`creatinine clearances from 0 to 100 mUmin. The point of intersection
`between the measured creatinine clearance (on the Jower abscissa)
`and lhis dose-fraction line is a coordinate with the dose fraction (on
`the left ordinate) corresponding to that particular creatinine clearance . ..
`For example. if a p:uient with a creatinine clearance of 20 mUmin
`
`In the patient with renal insufficiency, this computation provides
`an average plasma level during a dosage interval that is the same as
`the average plasma level during the dosage interval with normal renal
`fu nction; the fluctuations between peaks and troughs, however, will
`be Jess pronounced , but the peak value could be below the therapeutic
`level. Alternatively, the normaJ (I .5 mg/kg) dose of gentamicin could
`be administered bur the dosage interval prolonged by the modification
`factor based on the chunge in clearance
`
`8 h x 81 mLimi~ = 52 h
`12.4 mUmm
`
`ln this case, the plasma levels may be subtherapeutic for a delererious
`length of time during the dosage lntervaJ.
`In some instances it may be desirable to calculate a dose that will
`yield a cenain plasma level at steady state. This approach is most
`appropriate for constanl intravenous infusions where 100 percent of
`the dose is delivered to the systemic circulation. When clearance of
`a drug in a patient wirh renal insufficiency is calculated as above,
`then
`
`Dose.;
`(amtlunit time)
`
`Clri
`(voVunit time)
`
`X
`
`Cp
`(amr/vol)
`
`where the time, amount, and volume terms are uniform.
`If a plasma concentration of carbenicillin of 100 tJ.g/mL is the
`therapeutic objective in a patient with a creatinine clearance of 25
`mUm in , the infusion rate is calculated as follows. Carbenicillin
`clearance is
`
`Cl" = ( 68 X ~~) + 10 = 27 mUmin
`
`Therefore. carbenicillin should be infused at a rate of 2700 f.lg/min.
`Should rhc method of calculating dose based on the desired plasma
`level be applied to inrermiuenr-dose therapy, panicular attenrion
`shouJd be given to the fact that the calculation is based on an average
`plasma level and that peak plasma levels will be higher. In addition.
`if an oral drug is not completely absorbed, the computed dose must
`be divided by the fracrion (F) thut reaches the systemic circulation
`(see ··sready State•· above).
`The fractional rate constant (k) approach For many drugs.
`clearance data in renal fai lure are nor available. ln these cases. the
`fraction of lhc normal dose that is required in a patient with renal
`failure can be approximated from the ratio of the fractional rate
`constanr for elimination from the body in renal failure (kJ to that
`with normal renal function (k). This approach requires the assumprion
`!hilt the distribution of the drug (V,1) is not affected by renaJ diseuse.
`The approach is the same us that employed with cleurance data:
`
`k
`Dose = dose x ...!!
`k
`S~nce the ratio_k,1/k i~ the fraction of the usual close employed in a
`g1ven degree of renaJ u1sufficiency, it is tcrrt1cd the dose fractirm and
`
`h
`
`TABLE 66-2 Estimated fraction of usual dose of drug
`required for a patient with a creatinine clearance of zero
`(dose fractlon0 ) and average overall fractional elimination rate
`constant for a patient with normal renal function (k)
`
`Drug
`
`Dose fractiono
`
`k (per hour)
`
`ANTIBIOTICS
`
`Amikacin
`Amoxicillin
`Ampicillin
`Aztreonam
`Carbenicillin
`Cefazolin
`Cc Fotax i me*
`Cefoxltin
`Cefrnudime
`Ccfrri:u:onc•
`Cephalexin
`Cephalothin
`Chlorumphenicol
`Clprotloxndn
`Clindamycin
`Cloxacillin
`Dicloxncillin
`Doxycycline
`Erythromycin
`Gentamicin
`lmipenem
`Isoniazid:
`Fast mactivators
`Slow inactivarors
`Methicillin
`Minocycline
`Nafcillin
`Norfloxacin
`Oxacillin
`Penicillin G
`Piperacillin
`Rifampin
`Streptomycin
`Sulfadiazine
`Sulfnmcthoxazole
`Telracyc:line
`Tlcarclllin
`Tobramycin
`Trimethoprim
`Vancomycin
`
`MISCELLANEOUS DRUGS
`
`Chlorpropamide
`Lidocaine
`Sultinpymzone
`
`CII.RDI ~C GLVCOSlDES
`
`0.05
`0. 15
`01
`0.25
`0. 1
`0.06
`0.3
`0. 1
`0.1
`0.5
`0.04
`0,02
`0.8
`0.33
`0.8
`0.25
`0.5
`O.k
`0.7
`0.05
`0.25
`o.s
`0.5
`0. 12
`0.9
`0.4
`0.5
`0.25
`0.1
`0.33
`L.O
`0.05
`0.45
`0.85
`0.12
`0. 1
`0.05
`0.45
`0,03
`
`0.4
`0.9
`0.55
`
`0.4
`0.7
`0.6
`0.4
`0.6
`0.35
`0.7
`0.8
`0.4
`0.09
`0.7
`1.4
`0.3
`0.2
`0.2
`1.2
`t .2
`0.03
`0.5
`0.3
`0.7
`
`0.5
`0.25
`1.4
`0.06
`1.2
`0.2
`1.4
`
`1.4 o.s
`
`0.25
`0.15
`0.7
`0,07
`0.08
`0.6
`0.35
`0.06
`0.12
`
`0,02
`0.4
`0.3
`
`k (p~r day)
`
`Digitoxin
`Digoxin
`'' Se4! text on dos•1ge in rcnnl disease.
`- EMmn:nol cl~orance nl~o mny be recluc~d in patiems wilh renul fuilure who nre ur~mic
`aud /ilf ill
`
`0. 1
`0.45
`
`07
`0.3
`
`

`

`

`

`400
`
`H\RT FOUR CLINICAL f'Ho\flMAI.!OL.OttV
`
`[n most cases the degree of binding is fairly consLant across !he
`therapeutic concentration range so that significant error is not caused
`by individualizing therapy on the basis of total drug levels in plasma.
`However, states such as hypoalbuminemia, liver disease, and renal
`disease can decrease the extent of drug binding, panicularly of acidic
`and neutral drugs, so that at any total plasma level there is a greater
`concentration of free drug and a risk of inc!eased response and
`toxicity. Other conditions, e.g., myocardial infarction, surgery,
`neoplastic disease, rheumatoid arthritis. and bums. that lead to an
`increased plasma concentrntion of the acute-phase reactant alpha1•
`acid glycoprotein have the opposite effect on the basic drugs that are
`bound to this macromolecule. The drugs for which changes in binding
`are important are those which are normally highly bound in the plasma
`(>90 percent) because a small alteration in the extent of binding
`produces a large change in the amount of drug in the unbound form.
`The consequences of these binding changes, panicularly with
`respect to total drug levels, depend on whether the clearance and
`distribulion are dependent on the unbound or total drug. For many
`drugs, elimi.nntion and distribution are largely restricted to the unbound
`fraction, and therefore, a decrease in binding leads to an increase in
`the clearance and distribution of the drug. The relative magnitudes
`1f these changes are such that the net effec~ is to shorten the half(cid:173)
`lire. The appropriate modification of the dos:lge 1e·~imen in conditions
`with reduced drug binding, as is the case of phenytoir. in renal failure ,
`is simply to administer the usual daily dose of the drug but in divided
`doses at more frequenr intervals. Individualization of therapy can then
`be based on either the clinical response or the plasma concentration
`of unbound drug. It is critical that tl1e patient not be titrated into the
`usual therapeutic range for concentration of total drug in plasma,
`since this will lead to excessive response nod toxicity.
`In the case of drugs bound to alpha 1-acid glycoprotein , the disease(cid:173)
`induced increase in binding has the opposite effects of reducing the
`cleorance and distribution of total drug. Accordingly, constant rate
`infusion of lidocaine to control arrhythmias after myocardial infarction
`leads to an accumulation of total drug. However. the clearance of
`unbound and pharmacologically active drug remains essentially
`unchanged . Again, it is critical that the patient not be dosed on the
`basis of total drug concentrations in the plasma. since this will be
`associated with subtherapeutic levels of unbound drug.
`
`VARIABLE ACTIONS OF DRUGS CAUSED BY
`GENETIC OII=FEAENCES IN THEIR METABOLISM
`
`Isoniazid, hydralazine, procainamide, and a
`ACETYLATION
`number of other drugs are metabolized by acetylation of a hydrazino
`or amino group. This reaction is catalyzed by N-acetyl transferase-2.
`an enzyme in !he liver cytosol that transfers an acetyJ group from
`acetyl coenzyme A to the drug. Individuals differ markedly in the
`rate at which drugs are acetylated, and there is a bimodal distribution
`of the population into •·rapid acetylators" and "slow acetylators. "
`The rate of acetylation is under genetic control: slow acetylation is
`an autosomal recessive trait.
`Responses ro hydralazine therapy are dependent on the acetylation
`phenotype. The hypotensive effect of hydralazine is greater in patients
`who acetyl ate the drug slowly, and the lupus erythematosus-like
`syndrome produced by hydralazine occurs almost cxclusiveJy in those
`with slow acetylation. Thus it may be of value to know the acetylation
`phenotype as a predictor of which patients with hypertension might
`benefit from an increase in the dose of hydralazine above the 200 mg
`daily that can be safely employed in the population at large.
`Acetylation phenotype can be determined by measuring the ratio
`of acetylated to nonacetylated dapsone or sulfamethazine in plasma
`or urine following administration of a test dose of these acetylation
`substrates. The ratio of monoacetyldapsone to dapsone in plasma at
`6 h after dapsone administration is less than 0.30 for slow acetylators
`and greater than 0.35 for rapid acctylator:s. A1 6 h following the
`administration t)f sulfamethazine, less than 25 percent of the drug in
`
`the plasma is in the acetyJated form in slow acetylators (in rapid
`acetylators. more than 25 percent): in the uri1~e coJlected in the S- to
`6-h interval after administration, less than 70 percent of the drug is
`in the acetylated form in slow acetylators (in rapid acetylators, more
`than 70 percent). More recently, the acetylation of a metabolite of
`caffeine, possibly the most widely consumed drug worldwide because
`of its presence in a variety of foodstuffs, has been used as an indicator
`of phenotypic status. In U1is procedure, the urinary molar ratio of 5-
`ucetylamino-6-amino-3-methYiuracil to methylxanthine after ingestion
`of a drink of coffee or cola is detem1ined. Antimodes of about 1.8
`and 6.6 separate slow, intermediate, and rapid acetylators; moreover,
`these three groups appear to correspond to the expected genotypes.
`ln healthy
`METABOLISM BY MIXED-FUNCTION OXIOASES
`individuals taking no other medications, the major determinant of the
`rate of metabolism of drugs by the hepatic mixed-function oxidases
`is genetic. Hepatic endoplasmic reticulum contains a family of
`cytochrome P450 isoenzymes with differeqt substrate specificities.
`Many drugs undergo oxidative metabolism by more than one iso·
`enzyme, and the steady state concentrations of such drugs in the
`plasma is a function of the sum of the activities of these and other
`metabolizing enzymes. When a drug is metabolized by multiple
`pathways, the catalytic acli vi ties of the participating enzymes are
`regulated by a number of genes so that the frequency of clearance
`rates and steady state concentrations of the drug tend to distribute
`unimodally within the population. The range of activity may differ
`markedly (tenfold or more) between different individuals. as is the
`case for chlorpromazine, and there is no way to make a prior prediction
`of the rate.
`For certain metabolic pathways, bimodally distributed activity
`suggests control by a single gene, and several polymorphisms have
`been identified. As a result, two phenotypic populations are usually
`present analogous to the situation with N-acetylation (see above). A
`majority of the population are extensive metaboUzers (EM), and a
`smnller group of individuals of the poor metabolizer (PM) phenotype
`have an impaired , if not an absent, ability to metabolize the drug.
`For example, aboor 8 to I 0 percent of whites are unable to form the
`4-hydroxy metabolite of the test drug debrisoquin. and this trait is
`inherited in an autosomal recessive fashion.lmportanUy. the putatively
`involved cytochrome P450 isoenzyme 206 is also at work in the
`biotransformation of other drugs whose metabolic fate. therefore.
`cosegregutes with the debrisoquin trait. These other drugs include
`antiarrhythmic agents (propafenone, necairude), beta-adrenoceptor
`blockers (alprenolol , metop(oJol, timolol), tricylic antidepressants
`(nortriptyline, desipramine, imipramine, clomipr;unine), neuroleptic
`drugs (perphenaz.ine, thioridazine, and possibly ftuoxetine), and
`certain opiates such as codeine. Thus a much reduced analgesic effect
`of codeine is obtained in PM patients because of impaired production
`of the active metabolite morphine. A similar situation occurs with the
`oxidative polymorphism that involves the metabolism of mephenytoin.
`The situation is further complicate<! by interethnic differences in the
`frequency of the polymorphisms. Por example, impaired hydroxyla(cid:173)
`tion of mcphenytoin is present in only 3 to 5 percent of whites. but
`the incidence is about 20 percent in individuals of Japanese descent;
`likewise, the frequency of the PM phenotype for debrisoquin hydroxyl·
`ation appears to decrease as one moves from western (8 to 10 percent)
`to eastern (0 to 1 percent) popul ~ttion groups .
`Polymorphisms in drug-metabolizing ability may be associated
`with large differences in the disposition of the drug among individuals.
`especially when the involved pathway is a major contribution to the
`overall elimination of the drug. For example, the oral clearance of
`mephenytoin differs 100- to 200-fold between individuals of the EM
`and PM phenorypes. As a result, peak plasma concentrations and
`bioavailnbility after oral administration may be profoundly increased
`and the rate of drug elimination decreased in PM individuals. This
`in turn results in drug accumulation and exaggerated pharmacologic
`responses, including toxicity, when usual dn1g dosages are adminis(cid:173)
`tered to patients with the PM phenotype. Or1.1g interactions between
`compounds that are metabolized by cytochrome P450 206 or which
`
`

`

`CltAI'TEii 68 PRINCIPl!'S OF ORlJG lHERAPV
`
`401
`
`.inhibjt its activirty noncompctitively, e .g., quinidine; may be of
`considerable cHruical importance in patients with the EM phenotype,
`since such concomitant administration often leads to impaired drug
`handling similar t.o that in the PM phenotype. Effective individualiza(cid:173)
`tion of drug therapy is even more critical when using drugs exhibiting
`polymorphic drug metabolism.
`
`DRUG USE IN THE ELDERLY (See also Chap. 8)
`
`The elderly (>65 years) constitute about 12 percent of the U.S.
`population, and will increase to about 10 percent, or 50 to 60
`million inilividuals, over the next 20 years. These patients use a
`disproportionate amount of prescription medicaiions (30 percent), and
`in addition, 70 percent· of tbe elderly regularly use oveHhe-counter
`drugs, compared with only 10 percent in the general adult population.
`Aging results in c;banges in organ function, especially in those organs
`involved in drug disposition, as well as alterations in body size and
`composition. Not surprisingly, therefore, pbarroacokinet:lc differences
`are often present in elderly i ndividuals compared with younger
`cohorts. UnfortUnately. few generalizations appear to exist with
`respect to the tytpe, magrutude, or clinical importance of any age(cid:173)
`related changes or their extent in an individual patient. Multiple
`diseases are also common in geriatric patients, and it is not unusual ,
`therefore, that a l1arge number of drugs are required, wruch may result
`in drug interactions that, along with increased vulnerability to
`morbidity and mortality, contribute to the hjgher incidence of adverse
`drug reactions in elderly patients. fncreased sensitivity oftarget organs
`and impaired phrsiologjc control systems. such as those involved in
`the regulation of the circulation, also may be a factor. Accordingly,
`optimization of d,rug therapy in the elderly, particularly frail patients ,
`is often difficult, since a variety of factors that are frequently
`poorly defined accentuate the usual interindividuaJ variability in drug
`response.
`Although malitY individuals preserve good renal. function into old
`age, as' a group, elderly patients have an increased and predictable
`likelihood of imp•aired renal excretion of drugs. Even in the absence
`of kidney disease , ~enal clearance is generally reduced by about 35
`to 50 percent in elderly patients. Dosage adjustments analogous to
`tl1ose in patients with kidney dysfunction (see above) are therefore
`necessary for drugs that are predominantly eliminated from the body
`by the renal route, e.g., digoxin, aminoglycosides, lithium, and other
`drugs listed in Table 66-2. 1n this regard , it is important !o recognize
`that the reduced muscle mass present in older individuals results in
`a reduction in !"he rate of creatinine production; thus a normal
`serum creatinine concentration can be present even though creatinine
`clearance is impatired.
`Aging also results in a decrease in liver size and blood flow , and
`possibly reduced. activity of hepatic drug-metabolizing enzymes:
`accordingly, the hepatic clearance of some drugs is impaired in the
`elderly. Unfortunately, no consistent pattern of clinical application
`appears to be pn!sent. Moreover, changes that may exist are often
`modest relative to the interindividual variability within the patient
`population. However, even small reductions in hepatic extraction
`"'lllay result in a !;ignificanl increase in oral bioavailability of drugs
`with a high first-Jpass effect, such as propranolol and labetalol.
`. As a consequence of impaired clearance and/or increased distribu(cid:173)
`tion, the elimination half-Jives of drugs may increase with aging.
`Thus, if a dosage modification in an elderly patient is required, it is
`often possible to .accomplish this by decreasing the frequency of drug
`administration. possibly along with a reduction in dose.
`Even if the pharmacokinetics of a drug are not altered , elderly
`patients may reqllire a smaller drug dosage because of an increase in
`Pharmacodynami•c sensitivity. Examples include increased analgesic
`effects of opioids, increased sedation from benzodiatepines and other
`ccn.tral nervous s.ystem depressants. and increased risk of bleeding
`W11l!e receiving arntjcoagulam therapy. even when clotting parameters
`are Well controlled. Exaggerated responses to cardiovascular drugs
`
`are also common because of the impaired responsiveness of normal
`homeostatic mechanisms. Such age-related changes require close
`monitoring of the patient's clinical response and appropriate dosage
`titration .
`fn general, drug therapy for the elderly should be attended by
`increased alertness to the possibility of moderate reductions in the
`clearance of drugs and instances of exaggerated pharmacodynamic
`responsiveness.
`
`INTERACTIONS BETWEEN ORUGS
`
`The effect of some drugs can be altered markedly by the administration
`of other agents. Such interactions can sabotage therapeutic intent by
`producing excessive drug action (with adverse effects) or decreasing
`the action of a drug, rendering rl ineffective. Drug interactions must
`be considered in the differential iliagnosis of unexpected responses
`to drugs, recogruzing that patients often come to the physician with
`a legacy of drugs acqujred during previous medical experiences. A
`meticulous drug history wm minimize the unknown elements in the
`therapeutic milieu; it should include examination of the patient's
`medications and calls to the pharmacist to identify prescriptions. if
`necessary.
`There are two plincipal types of interactions between drugs.
`Pharmacokinetic interactions result from alteration in the delivery of
`drugs to their sites of action. Phannacodynamic interactions are tbose
`in which the responsiveness of the target organ or system is modified
`by other agents.
`An index of the drug inter