`Associate Professor of Clinical Pediatrics
`New York University School of Medicine
`Assistant Director of Pediatrics
`Director, Pediatric Intensive Care Unit
`University Hospital
`New York University Medical Center
`New York, New York
`
`ASSOCIATE EDITOR
`Joan Holter Gildea, R.N., B.S., MA.
`Clinical Assistant Director of Nursing
`NewYork University Medical Center
`New York, New York
`
`
`
`CRITICAL CARE
`PEDIATRICS
`
`
`
`Philadelphia London Toronto Mexico City Rio de Janeiro Sydney Tokyo Hong Kong
`
`|eeoonneeeeriepetSinece=eR
`
`
`annneaerere-=
`ian
`
`1985
`W. B. SAUNDERS COMPANY
`
`TEVA EX. 1021
`Page 1
`
`TEVA EX. 1021
`Page 1
`
`
`
`
`
`TEVA EX. 1021
`Page 2
`
`TEVA EX. 1021
`Page 2
`
`
`
`
`
`
`concentration; in 3.3 half-lives, it will be 90 RXty,
`C,= (14.4)
`percent; and in five half-lives, it will be 97
`
`0.7% Va x Wt.
`percent of the plateau level.
`Dosage recommendations for continuousin-
`
`fusions are designed to produce appropriate
`
`plasma concentrations at equilibrium. The phe-
`
`nomenonjust described, which is often termed
`drug accumulation, entails a delay in achieving this
`
`concentration. The magnitude of the delay is
`
`related to the half-life of the drug, whereas the
`ultimate concentration is determined by the V4,
`
`the half-life, and the rate of administration.
`
`
`
`CLINICAL PHARMACOLOGYIN THE CRITICALLY ILL CHILD / 91
`
`where R is expressed in mg/hr and C,, is ex-
`pressed in mg/L.
`Equation 14.4 indicates that equilibrium drug
`concentrationis related to three variables: half-
`life (t,,), volume of distribution (V,), and rate
`of administration (R, or D/T). Thus, doubling
`V4 has the same effect upon steady state con-
`centration as halving t,,. Either alteration will
`lead to a 50 percent reduction in drug concen-
`tration that can be exactly offset by doubling
`the dosage.
`
`
`
`es
` YtnandS5
`
`
`
` poaesoe2menepeehemeenatne)—o
`
`
`
`AHAOD
`
`oT1
`AOM1
`
`Plasma Drug Clearance
`
`The plasma clearance (Cl) of a drug is of
`primary importance in appreciating the rela-
`tionship between rate of drug administration
`and consequent drug concentration. Drug
`clearance, like creatinine or inulin clearances,
`is determinedbyrelating therate of elimination
`(E) to the plasma concentration at equilibrium
`(C,.):
`
`Cl = D|n
`
`At equilibrium, E = R. Thus,
`
`R
`cl= =
`°
`Cn.
`
`or, with rearrangement,
`
`R
`==
`C= 5
`
`14.3
`(14.3)
`
`R=caxc,
`
`°
`
`(14.3A)
`
`where R = rate of administration.
`Equation 14.3A emphasizesthatit is drug clear-
`ance, rather than half-life, which determines
`the rate of administration (R) or dosage per
`interval (D/T) necessary to achieve a specific
`concentration. Ultimately, clearance is related
`to half-life and volume of distribution:
`
`Cl
`
`07 x Va XWe.
`ty,
`
`when V, is expressed in L/kg, half-life (t,,) in
`hours, and clearance (Cl) in L/hr.
`With substitution into equation 14.3, this be-
`comes:
`
`Multiple Dose Kinetics
`
`The reader has been introduced to the phe-
`nomenon of drug accumulation as it occurs
`during continuous infusion. Drug accumula-
`tion also occurs with intermittent dosage
`schedules.
`Consider a drug that is given by intermittent
`IVinjection. When the first dose is given, the
`concentration of drug is zero. Immediately after
`the dose, a peak concentration is recorded. The.
`concentration then declines at a rate deter-
`mined by the drug’s half-life. If the next dose
`is given before the concentration has once again
`reached zero, the second peak will be higher
`than the first. As this process continues, the
`peak (C,,,.) and trough (C,,,,) levels rise toward
`plateau values, as will the average concentra-
`tion, C,y.. This process is illustrated in Figure
`14-4. Drug accumulation occurs during inter-
`mittent administration when a second, or nth,
`dose is administered before all the previous
`dose has been eliminated. For mostclinical pur-
`poses, this conditionis satisfied when the dos-
`ing interval is less than twice the half-life of
`the drug. As with continuousinfusions, 50 per-
`cent of a plateau concentration is achieved in
`one half-life; 97 percent is achieved after five
`half-lives.
`Cave is analogous to the equilibrium concen-
`tration (C,,) that develops during continuous
`infusion. Thus, its value is determined only by
`the relationship between clearance (Cl) and rate
`of administration (R; see Equation 14.3). The
`peak (C,,,.) and trough (C,,,) concentrations
`fluctuate around the C,,, in a manner that is
`determined by the size of the dose and the
`length of the dosing interval. For example,
`theophylline may be administered by inter-
`mittent IV injection. In an adult, a standard
`regimen calls for 300 mg every 6 hours (1200
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
` PSea
` on:pacesier
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`TEVA EX. 1021
`Page 3
`
`
`
`TEVA EX. 1021
`Page 3
`
`
`
`15
`20TerrTT?rrrrtrd |
`
` pytt
`
`
`
`10 12 14 16 18 20 22 24 26 28 30 Hours
`5
`6
`7 Number of
`3
`Half-lives
`
`
`
`TheophyllineConcentration
`
`(mg/L)
`
`—
`
`Ld
`246 8
`1
`2
`
`Figure 14-4,
`Concentration of theophylline during intermittent IV administration. Note that Cae Cain and C,,. increase
`to equilibrium values. C,,. = C,, that obtains during continuous administrationif the total daily dosages are identical.
`
`remain within the respective therapeutic ranges
`throughout the dosing interval.
`
`Nonlinear Kinetics
`
`
`
` 92 / GENERAL PROBLEMS AND TECHNIQUES OF CRITICAL CARE
`
`mg/day). Alternatively, one may administer
`150 mg every 3 hours (1200 mg/day). Finally,
`some physicians administer a continuous in-
`fusion of 50 mg/hr (1200 mg/day). Both in-
`termittent regimens produce the same(C,y.),
`which is equal to the C,, during the continuous
`infusion, This does not mean that the regimens
`are equivalent. The 6-hour schedule produces
`much greater fluctuation around C,,. than the
`3-hour schedule. With the 3-hour regimen, the
`peaks and troughslie closer to C,,.. This in-
`creases
`the
`likelihood of remaining within the therapeutic
`range throughout the dosing interval (Fig. 14—
`5).
`In this regard, the half-life of a drug is a
`watershed. When a drugis given at an interval
`that is equal to its half-life (T = ty), Crax/Coin
`is approximately 2. During more frequent
`administration (T < ty), Cra/Crmin 8 less than
`2, and during less frequent administration
`(T > ty), Cra/Cmin if greater than 2.
`Thus, drugs with a long half-life, such as
`digoxin or phenobarbital, are often given once
`daily, because even with this schedule, T is less
`than t,, and the plasma concentration remains
`within the relatively narrow therapeutic range
`of these agents. Conversely, theophylline and
`quinidine have relatively short half-lives (3 to
`6 hours in children). When conventional for-
`mulations of these agents are administered,
`they requirerelatively frequent dosing (every
`3 to 6 hours) if the plasma concentration is to
`
`To this point, the discussion has concerned
`first-order kinetic behavior in which a fixed pro-
`portion of drug is eliminated per unit time. Zero-
`order kinetics occurs under some conditions, no-
`tably when plasma drug concentrationsarerel-
`
`
`
`
`
`Concentration(arbitraryunits)
`
`Dosage = D
`Interval = T = ty,
`
`Dosage = D/2
`
`Cave Css
`|
`Interval = Ti2 = ty,/2
`
`Time
`
`Effect of varying both dosage and dose
`Figure 14-5.
`interval upon peak (C,,,,) and trough (C,,,,) concentrations
`during steady state. The solid saw tooth line indicates the
`time concentration curve that results with intermittent IV
`administration of dosage D at an interval T equal to the
`drug tv. Note that C,../Gmin = 2. The interrupted fine
`indicates the curve that results when dosage (D/2) and
`interval (T/2) are halved. Cya/CGmin = 1.5. The straight
`solid line indicates C,,,. which is {he.same during beth
`conditions, and is equal to C,,, which results when the
`sametotal daily dosage is administered by continuous IV
`infusion.
`i
`
`TEVA EX. 1021
`Page 4
`
`TEVA EX. 1021
`Page 4
`
`
`
` i
`
`TEVA EX. 1021
`Page 5
`
`
`
`CLINICAL PHARMACOLOGYIN THE CRITICALLY ILL CHILD / 93
`
`atively large. With zero order or nonlinear ki-
`netics, a fixed amount of drugis eliminated per
`unit time. Ethanol is an extreme example, be-
`cause the usual dosage is large (gm amounts)
`relative to other drugs (mg amounts). Within
`the usual range of blood ethanol concentra-
`tions, humans eliminate about 120 mg per kg
`per hr of the substance. Because the volume of
`distribution of ethanol is about 0.5 L per kg,
`blood ethanol levels decline at a fixed rate of
`20 to 25 mg/dl per hr. Therate of elimination
`does not change with increases in concentra-
`tion. Consequently, increments in dosage pro-
`duce much greater changes in concentration
`than would be the case for a drug eliminated
`in accordance with first-order kinetics.
`Many substances follow a first-order model
`at low plasma concentrations but a zero-order
`modelat higher concentrations. Whenthe tran-
`sition from First- to zero-order elimination oc-
`curs at concentrations appreciably higher than
`the usual therapeutic range, the pharmacoki-
`netic treatment of the drug is uncomplicated
`and a first-order kinetic model will be suffi-
`ciently accurate for most clinical purposes.
`Unfortunately, a few commonly used drugs,
`such as phenytoin andsalicylate, exhibit this
`transition at concentrations within the thera-
`peutic range.
`A change from first- to zero-order kinetics
`as concentration increases is typical of an en-
`zyme-mediated process. This change is due to
`saturation of the enzyme system that is re-
`sponsible for metabolic transformation of the
`drug. There is a limited amount of enzyme at
`the metabolic site; therefore, there is a maxi-
`mum rate at which transformation can occur
`(Vinax): At concentrations that are low relative
`to Ving, first-order behavior predominates. As
`concentration increases, V,,,.
`is approached.
`After V,,., has been achieved, further increases
`in concentration cannot augmentthe metabolic
`rate. Thus, a fixed amount of drug is metab-
`olized per unit time. This amount, of course,is
`equal to Vina. Mathematically, this process is
`described by the Michaelis-Menton equation:
`
`14.5
`
`_ Vinax x Cc
`(14.5)
`K,4+C
`E=
`where E is the rate of elimination or metabo-
`lism;V,,,, is the maximum rate of metabolism;
`K,, is the Michaelis-Menton constant, which
`defines the affinity of the enzymefor the drug;
`and C is plasma drug concentration.
`Note that when C is much less than K,,, E
`varies directly with C. This resemblesa first-
`order process. When C is greater than K,, E
`
`approaches V,,x, and zero-order behavior oc-
`curs.
`There are two important consequences of
`this kinetic behavior. The first is that an in-
`crease in dosage produces an exponential, rather
`than a linear rise in concentration. This occurs
`very often when treating patients with phen-
`ytoin (Fig. 14-6); on occasion, this phenomenon
`is recognized during treatment with theophyl-
`line. It requires that dosage adjustments must
`be made cautiously and in small amounts. The
`second major consequence of Michaelis-Men-
`ton kinetics is that the apparent plasmahalf-life
`increases with the plasma concentration. The
`greater the plasma concentration, the sloweris
`the relative rate of elimination. Using repre-
`sentative values of K,, and V,,.. for phenytoin,
`one can estimate that at a concentration of 10
`mg per L, the apparentt,, of phenytoin is 24
`hours; at a concentration of 25 mg per L, the
`apparent t,,
`is 42 hours. This means: 1) in-
`creases in dosage cause lengthening of the ap-
`parent ty, (thus, Michaelis-Menton kinetics is
`sometimes referred to as dose dependent kinetics),
`2) small increments in dosage can produce huge
`increases in drug concentration, and 3) intox-
`ication with phenytoin will be prolonged, be-
`cause, at high concentrations, elimination is ex-
`tremely slow relative to the amountof drug in
`the body.
`
`MAINTENANCE DOSE
`
`The maintenance dose (MD) is the amount
`of drug (R for continuousinfusion, D/Tfor an
`intermittent schedule) that is administered dur-
`ing equilibrium. Thus, from Equation 14.3A,
`maintenance dose, MD,is equal to the product
`of clearance, Cl, and desired steady state plasma
`concentration, C,,
`(MD = Cl x C,,). The
`maintenance dose is often determined by con-
`sulting standard reference material. In patients
`
`Phenytoin
`(Michaelis-Menton)
`
`Gentamicin
`(First-order)
`
`Concentration
`
`
`
`Daily Dosage
`Effect of dosage upon plasma concentra-
`Figure 14-6.
`tion for drugs following first-order vs. Michaelis-Menton
`kinetics.
`
`—
`
`TEVA EX. 1021
`Page 5
`
`
`
` 94 / GENERAL PROBLEMS AND TECHNIQUES OF CRITICAL CARE
`
`Cc
`
`2 se
`
`coQO<=
`
`8O
`
`owEn
`
`O&a
`
`o
`
`with abnormal drug disposition, better indi-
`vidualization of therapy is achieved if one ap-
`preciates the relationship between changes in
`Cl and consequent changes in C,,. Knowledge
`of a patient’s Cl can be used to calculate dosage
`requirements. This procedure is described in
`the case study at the conclusion of this chapter.
`It is not always appropriate to imifate therapy
`with the maintenance dose. Both continuous
`and intermittent schedules produce a gradual
`rise from theinitial (usually zero) to the equi-
`librium concentration. Recall that 75 percent of
`the plateau is reached in twohalf-lives, and 97
`percent is reached in five half-lives. Thus, for
`drugs that have a long half-life, there will be
`a substantial delay in acquisition of the plateau
`concentration. Because the plateau concentra-
`tion may be close to the minimum effective
`concentration, this delay may be unacceptable
`in acutely ill patients. For example, an asth-
`matic whois simply placed on a theophylline
`infusion will not begin to experiencerelief for
`about 8 hours.
`
`LOADING DOSE
`
`The solution to the problem of delay in
`achieving adequate levels is to administer a
`loading dose (LD). The loading dose is the
`amount of drug that will rapidly produce a
`therapeutic plasma concentration.
`If one is emphasizing a target concentration
`strategy, calculating the loading dose is simple,
`because the loading dose and the desired con-
`centration (C,,) are related through the volume
`of distribution (see Equation 14.2). With ap-
`propriate modifications,
`this expression be-
`comes:
`
`LD = V,(L/kg) x Wt(kg) x C,,(mg/L)
`
`(14.6)
`is the desired equilibrium concen-
`where C,,
`tration. Thus, for a child weighing 10 kg in
`whom one wishes to achieve a plasma theo-
`phylline (V, = 0.5 L/kg) concentration of 12
`mg per L, the correct loading dose is 60 mg.
`This amount of drug should be administered
`slowly, over about a 15-minute period. Im-
`mediately thereafter, the appropriate mainte-
`nance doseis initiated. The effect of a loading
`dose is shown in Figure 14-7.
`If one is using an empirically derived main-
`tenance dose and is not attempting to achieve
`a specific drug concentration (target effect
`strategy), the problem is less straightforward.
`In such cases,
`it is probably best to consult
`
`
`
`:T
`
`1
`
`SSa
`Peeaaee
`
`ae
`
`Se
`eeee
`
`ty,
`
`4 ty
`
`Time
`Figure 14-7. Administration of an appropriate loading
`dose eliminates the delay (4t,,) in achieving equilibrium
`concentration, C,,. Solid line-infusion alone, beginning at
`T = 0. Interrupted line—loading dose at T = 0, followed
`by continuousinfusion.
`
`individual product information whendesigning
`the loading dose. Readers interested in a the-
`oretical approach to this issue should consult
`the suggested reading by Rowland and Tozer.
`Whenintravenous therapy is indicated, the
`loading dose is often given as a single rela-
`tively brief infusion. In the case of a drug with
`a narrow therapeutic range or a prolonged
`phaseofdistribution, the physician may choose
`to divide the loading dose, as is commonly done
`with digoxin. In general, a loading dose is not
`indicated whenthe half-life is much less than
`the dosing interval
`(i.e., drug accumulation
`does not occur) or when the therapeutic range
`is wide. Thus, penicillin therapy does not begin
`with a loading dose. Of course, a loading dose
`is not indicated when there is no urgency in
`achieving the equilibrium drug concentration.
`There is also no point in administering a load-
`ing dose when the half-life of a drug is very
`short, as with most pressor agents, because
`equilibrium conditions are reached in a matter
`of minutes during continuous maintenance in-
`fusion.
`The foregoing kinetic description applies to
`intravenous administration. Following intra-
`muscular or oral administration one must ex-
`tend the analysis by taking into account the
`rate and extent of absorption. Drugs that are
`completely and efficiently absorbed after in-
`tramuscular injection should maintain a similar
`C,ye although peak and trough levels may lie
`closer to C,,. (lower peak, higher trough). After
`oral administration, many drugs either are not
`completely absorbed from the gastrointestinal
`tract or once absorbedare efficiently extracted
`and then biotransformed on the first pass
`through the liver. This process effectively re-
`duces the dosage of drug that reaches the sys-
`temic circulation. Equations 14.2 through 14.4
`are modified by multiplying the dosage ad-
`
`TEVA EX. 1021
`Page 6
`
`TEVA EX. 1021
`Page 6
`
`
`
`CLINICAL PHARMACOLOGY IN THE CRITICALLY ILL CHILD / 95
`
`Table 14-2. Selected Drugs With Altered Volume
`
`of Distribution (V,) in Neonates and Children
`Drug
`Effect onVa
`
`Diazepam
`aa
`Digoxin
`Furosemide
`Gentamicin
`Lidocaine
`Phenobarbital
`Phenytoin
`
`Theophylline
`t = V, larger; | = Va smaller; N = neonate; and C =
`children older than 1 month. The effect of changes in V4
`may be modified by alterations in protein binding and
`elimination rate.
`
`ZZAZzZZZZZ
`
`saa
`
`
`
`shock), the rate of distribution is likely to be
`
`ance. Clearance (Cl) is related to both V, and
`half-life. In infants, drug half-life may be pro-
`longed (vide infra), and this may offset the
`increase in Vy (recall that Cl = 0.7 X V4q/ty).
`The net effect is frequently a reduced dosage
`requirement or an increased loading dose, fol-
`lowed by a reduced maintenance dose.
`
`DISEASE-RELATED CHANGES
`
`Several processes affect V4 by altering body
`water content, body fat content, or the degree
`of protein binding. Notable examples are ure-
`mia, chronic liver disease, and congestive heart
`failure.
`In uremia, the water content of the body is
`frequently greater than normal, Thisfactor, to-
`gether with disturbed protein binding, causes
`the volumeofdistribution (V4) of several drugs
`to increase (e.g., gentamicin) or decrease(e.g.,
`digoxin). As in newborns, the larger V, is fre-
`quently accompanied by prolongation of drug
`half-life.
`Chronic liver disease is associated with de-
`creased levels of plasmaproteins and with fluid
`accumulation. Thus, it is not surprising that the
`V,, of several drugs increases in the presence of
`cirrhosis. Other conditions associated with ex-
`tracellular fluid expansion also increase thedis-
`tribution of certain drugs; for example, ami-
`noglycoside antibiotics distribute into ascitic
`fluid. Thus, in the presence of ascites, the V4
`of these agents may be substantially increased.
`In cystic fibrosis, the V, of several of the ami-
`noglycosides seems to be higher than average.
`Unless daily dosage is increased, this may lead
`to subtherapeutic drug concentrations.
`Abnormalities of regional or global blood
`flow may reduce distribution by limiting per-
`fusion to sites of uptake. In patients with low-
`output states (congestive heart failure, CHF;
`
`TEVA EX. 1021
`Page 7
`
`
`
`ministered (R, or D/T)byf,,., the fraction of
`orally administered drug that reaches the sys-
`temic circulation. Several of the suggested read-
`ings contain a table of representative f,. Values.
`
`KINETIC VARIATION—CLINICAL
`APPLICATION
`
`In the foregoing discussion, three variables
`that affect the plasma drug concentration have
`been identified: 1) the rate of administration
`(R, or D/T), 2) the volumeofdistribution (V4),
`and 3) the half-life (t,,). These three variables
`are related to one another through Equation
`14.4, Of these, only the rate of administration
`is under the physician’s control. Thus, the art
`of pharmacokinetics consists of altering this
`parameter in order to compensate for individual
`differences in elimination or distribution.
`
`Altered Distribution
`
`Critical illness and the age of the patient af-
`fect the distribution of many drugs. Several
`mechanisms may be
`involved,
`including
`changes in the content or, distribution of body
`water, alterations in plasma protein binding,
`perturbations in regional blood flow, and dif-
`ferences in body fat content.
`
`AGe-RELATED CHANGES
`
`The water content of the body changes dra-
`matically with age. At 28 weeks gestation, the
`water content of the body is 85 percent. This
`figure decreases to 70 percent at term and to
`60 percent in adults. There is a concurrent in-
`crease in the amountof body fat from 1 percent
`of body weight at 28 weeks to 15 percent at
`term, as well as altered binding to protein. Dis-
`ease (tachypnea, dehydration),
`the environ-
`ment in which the infant is nursed, and the
`volume and composition of administered fluids
`produce fluctuations in body water. Thus,it is
`anticipated that drugs that are distributed
`Mainly in the body water have a different, usu-
`ally greater, volume of distribution (V,) in in-
`fants and young children than in adults. Table
`14-2 lists several drugs for which the V4 is
`knownto differ between newborn infants and
`adults. This information provides the rationale
`for many empirically determined dosage mod-
`ifications. A larger V; does not necessarily in-
`volve a larger dosage. The actual determinant
`of the maintenance dosage requirementis clear-
`
`a
`
`TEVA EX. 1021
`Page 7
`
`