This article is protected by copyright and is provided by the University of Wisconsin-
`
`Madison under license from John Wiley & Sons. All rights reserved. Br1 Clin Pharmacol 1998; 46: 479—487
`
`The influence of drug input rate on the development of tolerance
`to frusemide
`
`Monique Wakelkamp,l Gunnar Alvz'tn,l Harry Scheinin "2 & johan Gabrielsson3
`1Division chlinimI Pharmacology, Department ofMedical Laboratory Sciemes 8 Technology, Karolinska Institute, Huddinge University Hospital,
`Huddinge, Sweden, 2Departirient ofAnaesthesiology, Turku University Central Hospital, Turku, Finland and 3Department ofDrug Metabolism and
`Phan’nacoleineticc, Astra Arms/15, Sodma’lje, Sweden
`
`rate on its pharmacokinetic—
`Aims Understanding the impact of drug input
`pharmacodynarnic relationship may lead to a more optimal drug therapy. The aim
`of the present study was to investigate the influence of the rate of administration on
`tolerance development to frusemide, by giving the drug at four different infusion rates.
`Methods Eight healthy volunteers were given 10 mg of frusemide on four different
`occasions, as a constant—rate intravenous infusion during 10, 30, 100 and 300 min,
`respectively. Urinary volume and contents of frusemide and sodium were measured
`in samples collected over 8 h.
`Results The four different infusion rates systematically influenced the frusernide
`excretion rate versus diuretic and natriuretic response relationship. Counter—clockwise
`hysteresis occurred for the most rapid infusion rate, whereas a progressive clockwise
`hysteresis was observed for the slower infusions, indicating development of tolerance.
`For each subject, diuresis and natriuresis were modeled for all four treatments
`simultaneously, using a feedback tolerance model. It was not possible to describe
`the data using a model without tolerance. The time course of tolerance development
`showed remarkable differences between the infusion rates. The intensity ofmaximum
`tolerance development was significantly less for the slowest infusion rate compared
`with the more rapid infusions and it appeared significantly later. However, no
`differences in diuretic or natriuretic response were found between the treatments.
`Conclusions The direction of the hysteresis loop is dependent on the input rate of
`fi'usemide. After the administration of a single low dose of frusemide,
`the time
`course of tolerance, rather than the integrated time course of tolerance, is influenced
`by the drug input rate.
`
`Keywords: frusemide, drug input rate, pharmacodynamics, tolerance
`
`Introduction
`
`In the past 20 years pharmaceutical dosage forms have
`become more complex in their design. Novel drug delivery
`systems have been developed that achieve pharmacological
`efleca not solely based on chemical structure, but also on
`the basis of controlling the rate of administration of the
`drug [1]. Clinically important changes in drug effects may
`result from changes in drug delivery [2]. For the loop
`diuretic fiusemide, it has been shown that if identical total
`doses are given, a slow and constant input of the drug into
`the body induces a higher total effect compared with a rapid
`administration. This has been explained by investigating the
`efficiency of the drug [3—9]. However, the pharrnacokinetic—
`pharrnacodynamic relationship of frusemide may be altered
`by development of tolerance [6, 9, 10]. It has been suggested
`that for drugs that elicit compensatory homeostatic mechan—
`
`Correspondence‘ Dr Monique Wakelkamp, Division of Clinical Pharmacology,
`Department of Medical Laboratory Sciences and Technology, Huddinge University
`Hospiml, S—l4l 86 Huddinge, Sweden.
`
`isms, a slow input of the drug will trigger fewer homeostatic
`reactions,
`leading to a higher cumulated pharmacological
`response. On the other hand, for drugs that elicit ‘true _
`tolerance’, a slow input over a prolonged period of time
`will enhance the development of tolerance, reducing the
`total response [11].
`The rapid intravenous administration of multiple frusem—
`ide doses activates counter—regulatory mechanisms leading
`to a profound decrease in diuretic and natriuretic effects
`[12]. However, clockwise hysteresis,
`(a sign of tolerance
`development) has been observed during slow input of the
`drug, for example after the administration of oral doses of
`finsemide in combination with food or of controlled release
`formulations [5, 7, 10]. Understanding the consequences of
`changing
`drug
`input
`rate
`for
`its
`pharmacokinetice
`pharmacodynamic profile may lead to more adequate drug
`therapy. In the present study, frusemide was administered
`four times as a single dose using different infusion rates,
`with the aim to investigate and model the influence of the
`rate of administration on tolerance development
`to the
`drug.
`
`© 1998 Blackwell Science Ltd
`
`479
`
`IPR2015—00410
`
`Petitioners' EX. 1032
`
`Page 1
`
`IPR2015-00410
`Petitioners' Ex. 1032
`Page 1
`
`

`

`M. Wakelkamp et al.
`
`Methods
`
`Subjects
`
`Eight white male subjects participated in the study after
`giving written informed consent and after approval had been
`obtained from the Huddinge University Hospital Ethics
`Committee. Their ages ranged from 26 to 41 years and their
`body weights from 66 to 84 kg. All subjects were considered
`healthy according to detailed medical history and physical
`examination, which included an ECG and laboratory
`investigations. Smokers or heavy consumers of cafi‘eine
`(more than four cups of colfee daily) were excluded from
`participating in the study. None of the subjects regularly
`used any medications.
`
`Study design
`
`The study had a randomized cross—over design using two
`replicate 4 X 4 latin squares balanced for possible carry—over
`eflfects. The four frusemide infusions were administered on
`separate study days with intervals of at least
`1 week. To
`standardize experimental circumstances, no medications were
`allowed within 3 days before each study day. Also,
`the
`subjects were asked to refrain from alcohol and extreme
`physical activity within this period. Standardized meals were
`provided the day before and during each study day with
`a
`total
`content of 159 mmol
`sodium and 81 mmol
`
`potassium day_1 and caffeinated drinks were not allowed
`during this time. Urine was collected for 24 h on the day
`before each study day to assess adherence to the diet and to
`estimate basal diuresis. The subjects fasted overnight and the
`study started in the morning at the closure of the 24 h urine
`collection. Before the start of the experiment, a cannula was
`inserted into an antecubital vein of each arm and the subjects
`assumed the supine position for 30 min. Blood samples were
`taken to measure plasma active renin, angiotensin II,
`aldosterone, antidiuretic hormone (ADH), atrial natriuretic
`peptide (ANP) and catecholamines (adrenaline, noradrenaline
`and dopamine). After
`the blood samples, each subject
`emptied his bladder and was weighed, after which the
`administration of frusemide was started (at O h). Frusemide
`10 mg ml_1
`(FuriX®, Benzon
`Pharma, Copenhagen,
`Denmark) was diluted with saline solution to a concentration
`of 1 mg mlil. A 10 mg dose of frusemide was given
`intravenously of four different
`infiision rates, 60mlhT],
`20 m1 h_1, 6 ml th and 2 ml h.1 using a Syringe—Minder
`90 infiision pump (CRITIKON, UK). The total infiision
`times were 10, 30, 100 and 300 min, respectively. The
`infusion fluid was protected fi'om light. The subjects
`provided urine samples by voiding at 15 min intervals for
`the first 5 h and at 30 min intervals for the last 3 h after
`
`every dose. Blood samples were taken to measure plasma
`
`active renin, angiotensin II, aldosterone, ADH, ANP and
`catecholamines at 15, 30, 105 and 300 min and at 6 and 8 h
`after dosing. Blood samples were always taken before voiding
`and the subjects remained in the supine position, except
`when voiding. Every 30 min,
`the subjects drank 50ml
`water. Lunch was served 5 h after dosing. Plasma samples
`were stored at ~70° C. The urine volumes were weighed
`
`480
`
`and aliquots were carefully protected from light and stored
`at —70° C until analyzed for frusemide and sodium.
`
`Analytical methods
`
`Frusemicle concentrations in urine were determined in
`duplicate by h.p.1.c.
`[13]. The lower limit of quantitation
`(LLQ) was 0.125 ug 1111—1. The intra—assay
`coefficient
`of variation at
`lugml—1 was 6.4% and the inter—assay
`coefficient of variation was 7.6%. Sodium, plasma active
`renin, angiotensin II, aldosterone, ADI-I, ANP and cat—
`echolamines were analyzed elsewhere according to standard
`methods.
`
`Data analysis
`
`response may be
`the pharmacological
`frusemide,
`For
`monitored by measuring changes in volume diuresis or in
`the excretion rate of sodium. As loop diuretics exert their
`diuretic effects mainly from the luminal surface of the renal
`tubule [14],
`the pharmacological effects of firusemide are
`adequately described as a function of the urinary excretion
`rate of the drug [15, 16]. An apparent delay in the onset of
`diuretic response in relation to the urinary excretion rate of
`fi'usemide is commonly observed. Indirect—response models
`may be applied when a time lag exists between the
`concentration of a drug in plasma or biophase and the
`pharmacodynarnic response [17, 18]. In the present study, a
`modified indirect—response model was used for analysis of
`the pharmacokinetic—pharmacodynamic relationship [12,
`19]. The fiusemide excretion rate (pharmacoleinetics) and
`diuresis and natriuresis (pharmacodynamics) were modeled
`in separate steps and all subjects were analyzed individually.
`A multiexponential model was applied to describe the time
`course of the frusernide excretion rate (ER in ugmin_1)
`from 0 to 8 h:
`
`ER: 2 Aim—“i“
`i=1
`
`(1)
`
`The most appropriate pharmacokinetic model was selected
`by residual analysis. Each infusion rate was modelled separately
`in order to describe the observed fi'usemide excretion rates as
`closely as possible. The pharmacokinetic parameters were
`then fixed and ER served as input to the pharmacodynarnic
`model. For each subject,
`the pharmacodynamic model was
`regressed to the diuresis and natriuresis data for all four
`frusemide infusion rates simultaneously (PCNONLIN version
`4.2, Scientific Consulting Inc., Cary, N.C., USA). Uniform
`weights (a constant variance) were applied.
`The basic premise of an indirect—response model is that a
`measured response (R) to a drug is produced by indirect
`mechanisms. Factors controlling the production (Iain) or the
`loss
`(lewd of a response variable may be stimulated or
`inhibited by the drug [17, 18]. In our study, the response
`variables measured are diuresis (in ml minTl) and natriuresis
`(in minolmin_1). The rate of change of the diuretic
`(natriuretic)
`response (R) over time with no frusemide
`present can then be described as follows:
`dR
`—= kin ~ k0“, x R
`dt
`
`(2)
`
`© I998 Blackwell Science Ltd Brj Cfin Pharmacol, 46, 479—487
`
`IPR2015—OO410
`
`Petitioners' EX. 1032
`
`Page 2
`
`IPR2015-00410
`Petitioners' Ex. 1032
`Page 2
`
`

`

`where kin represents the zeroiorder constant for production
`of the diuretic response and k0“, defines the first—order rate
`constant for loss of response.
`Because frusemide inhibits the reabsorption of chloride,
`sodium and water [14], an indirect—response model was used
`assuming frusemide to increase the diuretic (natriuretic)
`response by inhibiting koul: according to an inhibition
`function KER):
`
`Imax >< ERY
`1(ER)=1 ‘—— .
`105'0 + ERY
`
`(3)
`
`where 1m represents the maximum effect attributed to the
`drug, ICSO represents the frusemide excretion rate producing
`50% of the maximum drug—induced inhibition potency and
`y is the sigmoidicity factor. The rate of change of diuretic
`(natriuretic) response over time with fi'usemide [12, 13, 18]
`can then be described as follows:
`
`dR
`— = kin —— k0“, >< 1(ER) x R
`dt
`
`(4)
`
`Maximum inhibition is obtained when ER >>IC50 and HER)
`then approaches 1—13,”. W'hen ER approaches zero, the
`net effect approaches the baseline eEect, i_e. I(ER)—>1.
`To account for tolerance, an endogenous regulator (or
`modifier) M was implemented into the model as a feedback
`mechanism [12, 19]. An increase in response R causes an
`increase in modifier M. For simplicity, M is assumed to be
`solely governed by R and a single rate constant kwl as
`follows:
`
`Drug input rate and tolerance to frusemlde
`
`Calculations
`
`The rate and extent of tolerance development was further
`investigated by calculating the time course of the modifier
`M and the time and size ofmaximum tolerance development
`(1mm and Mmax) from the estimated parameters for each
`frusemide treatment. Also,
`the AUC of M was calculated
`from the predicted M values using the linear trapezoidal
`rule, extrapolated to infinity. Differences in Mmax and AUC
`of M between the four treatments for diuresis and natriuresis
`were analyzed by repeated measures ANOVA. For pairwise
`comparisons, P values were adjusted using the Bonferroni
`multiple comparisons test. Differences in tmaXM between the
`treatments were analyzed by Friedman’s ANOVA for
`repeated measures, followed by Dunn’s test for multiple
`comparisons. Differences in total recovery of frusemide in
`urine,
`total diuresis and total natriuresis between the four
`treatments were analyzed by repeated measures ANOVA.
`For pairwise comparisons, P values were adjusted using the
`Bonferroni multiple comparisons test.
`
`Results
`
`The administration of fiusernide at different infiision rates
`had
`a
`profound
`effect
`on
`the
`pharmacokinetic—
`pharrnacodynarnic relationship of the drug. Figures
`1—4
`display the individual data of subject 2, which are representa—
`tive of all subjects. The most rapid (10 min) infiision gave
`rise to distinct counter—clockwise hysteresis of the frusemide
`excretion rate 125 response curve, indicating a delay between
`the excretion rate and the diuretic effect. In this case, the
`
`dM
`—:ktolXR—'I€tolXM
`dt
`
`(5)
`
`a
`
`25
`
`Al\)01O
`
`_r O
`
`
`
`_I.
`
`Dluresls(mlmin“)
`
`Natriuresis(mmolmin-‘)0—Lin01
`
`o
`
`
`20
`4o
`60
`so
`160
`120140166
`Frusemide excretion rate (pg min 1)
`
`
`
`When M increases, it counter-balances R, as M is assumed
`to stimulate Jewt according to a stimulation function S(M):
`M
`
`S(M)=(1+—)
`
`M50
`
`(6)
`
`M50 was assumed to be equal to unity. The full model can
`now be described as follows:
`
`dR
`— = kin _ kout X
`dt
`
`1mm x ERY
`1—_
`[CEO-FER?
`
`X R x (1 +M)
`
`(7)
`
`M is defined according to equation 5. A value for maximum
`response
`including
`tolerance
`development
`‘Rm’
`(basal+drug—induced) can be calculated from the parameter
`values obtained [12] according to
`
`1
`Rm=——+
`2
`
`kin
`l
`— + —— '
`4
`kout X (1—1max)
`
`<8)
`
`At equilibrium, the steady—state values of R and M are equal.
`For an overall evaluation, the same tolerance model was
`also applied to pooled data of all eight subjects
`(784
`observations), for both diuresis and natriuresis. Here,
`the
`pharrnacokinetic parameters obtained fiorn simultaneous
`modeling of all subjects for each infusion rate were used
`as input.
`
`o
`
`20
`
`120140 160
`100
`80
`60
`4o
`Frusemide excretion rate (pg min”)
`
`Figure 1 Diuresis (a) and natriuresis (b) 115 frusemide excretion
`rate following the 10min (0), 30 min (A), 100 min (0) and
`300 min (A) infusion in subject 2.
`
`© P398 Blackwell Science Ltd Br1 Clin Pharmacol, 46, 479—487
`
`481
`
`IPR2015—00410
`
`Petitioners' Ex. 1032
`
`Page 3
`
`IPR2015-00410
`Petitioners' Ex. 1032
`Page 3
`
`

`

`81a)
`7.
`5.
`
`E4
`2 3‘
`
`2 1
`
`
`
`-,.
`
`OF— ‘
`.
`'
`0‘1 2
`3
`4
`5
`6
`7 89101112131415
`12,63)
`'l'1me(h)
`
`
`
`o
`
`1
`
`2 ’3
`
`'4 5’" 6'7 8 9101112131415
`T1me(h)
`
`Figure 4 Simulation of tolerance development (M) For diuresis
`(a) and natriuresis (b) vs time following the 10 min (solid line),
`30 min (broken line), 100 min (line with large dots) and 300 min
`(line with small dots) infusion in subject 2. The horizontal bars
`indicate the respective infusion limes.
`
`total natriuresis between the four treatments. There was also
`no difference in total urinary recovery of finisernide between
`the treatments. The mean isd, values and the 95%
`confidence intervals for the recovery of fi-usemide (mg)
`were: 6.7:09; (5.9—7.5) for the 10 rnin infusion, 6.4i05;
`(6.0—6.8)
`for
`the 30min, 6.5i12;
`(5.54.5)
`for
`the
`100 min and 6.2i0.4; (5.8—6.5) for the 300 min infiision.
`The total diuretic and nalIiuretic response from 0 to 8 h
`obtained from all subjects is shown in Table 1.
`A good fit for the frusemide excretion rate was obtained
`for all subjects. A tri—exponential model with firsteorder
`input and firstiorder output, or a model with constant
`intravenous input and first—order output gave the best
`description of the data, depending on the infusion rate and
`the shape of the frusernide excretion rate 115
`time curve
`(Figure 2). The pharmacokinetic parameters obtained were
`introduced into the pharmacodynamic tolerance model.
`Tables 2 and 3 present
`the pharrnacodynamic parameter
`estimates obtained from the fit of the tolerance model for
`diuresis and natriuresis, respectively. Interestingly,
`the rate
`constant for tolerance development ktol was consistently
`higher for natriuresis than for diuresis. The mean half—life
`for tolerance development was 42 min for natriuresis 125
`139 min for diuresis. Figure3 shows the observed and
`calculated values of diuretic and natriuretic response of
`subject 2. High consistency was obtained between the
`predicted and observed data for all subjects, except for the
`diuresis dam of subject 8. The subset of the diuresis and
`natriuresis data after the 100 min infusion was left out from
`
`© I998 Blackwell Science Ltd Br] Clin Phun‘nacul, 46. 479—487
`
`IPR2015—OO410
`
`Petitioners' EX. 1032
`
`Page 4
`
`M. Wakelkamp et aI.
`
`
`
`Figure 2 Observed (symbols) and predicted (solid lines)
`fiusemide excretion rate 115 time following the 10 min (0),
`30 min (A), 100 min (0) and 300 min (A) infusion in subject 2.
`
`a
`
`35»
`
`30-
`
`25l
`MO
`
`mm
`
`Dluresls(ml a
`
`10
`
`
`
`
`
`om
`
`bim'tn
`
`_l
`
`
`
`Natriuresis(mmolmin") A
`
`.0
`
`Figure 3 Observed (symbols) and predicted (solid lines) diuresis
`(a) and natriuresis (b) as time following the 10 min (0), 30 min
`(A), 100 min (0) and 300 min (A) infiision in subject 2.
`
`than the rate of
`rate of the drug was greater
`input
`equilibrium between the pharmacokinetics (ER) and the
`pharrnacodynarnics (R), as well as the rate of equilibrium
`between the pharmacodynamics (R) and the development
`of tolerance (M) However, a clockwise hysteresis loop was
`observed after the 30 min infusion, which increased in size
`after the slower infusion times of 100 and 300 min (Figure 1).
`With these three infusion rates, the rate of input was slower
`than the PK—PD equilibrium rate, as well as the rate of
`tolerance development It was not possible to describe the
`data using a model without
`tolerance. The clockwise
`hysteresis indicated acute within—dose tolerance development
`to the diuretic effect of frusemide for the slower infusion
`rates. However, there was no difference in total diuresis or
`
`482
`
`IPR2015-00410
`Petitioners' Ex. 1032
`Page 4
`
`

`

`coolnooooo
`o
`‘3
`szfirmowo
`Q
`"‘1 H—lHHHF!
`v1
`
`126
`
`Q mothmHm
`Q mmmd-omsroo
`N
`H—i
`Fit—«Haw
`
`125
`
`107—145
`
`87—162
`
`4523
`
`:3?SE
`5%
`_
`53
`.55
`w
`
`t
`I»;
`.A-e
`“3
`Ch:
`2:“.
`
`D
`a
`
`t._t/
`
`e:
`W
`
`Ffi'OVPO‘ODO
`gCOw—‘mNO
`FHHHHVflF‘v—l
`
`10
`
`114
`
`106—122
`
`mOfi-flhhhch
`006300010000
`we HHHH
`
`96—156
`
`36
`
`126
`
`«Neva-mono)
`
`Mean
`
`s.cl.
`
`95%C1
`
`
`
`Ncfiwwoostofl
`C:
`vwmmmmmv—r
`<2 mean—hwon
`V—«HHHHHHH
`in
`
`1240—1601
`
`216
`
`1420
`
`D
`Q
`N
`
`
`
`r:
`*5
`e
`:3
`EE
`Va
`
` e
`a
`a:
`"‘30
`QNM
`
`Q
`w
`
`a
`£3
`2
`3
`V:
`
`stooxoonwoocs
`oxixommNooN
`oevrcomtrve‘
`H HHHfiN
`
`528
`
`1346
`
`090:)?th
`mnmommvn
`cupcake-own
`H
`HV—iV—i—i
`
`237
`
`1097
`
`NHWNv—CMN
`(“mews—CV?!)
`oommmmwo
`a
`w—HF‘N
`
`539
`
`1142
`
`
`
`
`
`692—159389941296904—1788
`
`w—INMVKVOBOO
`
`Mean
`
`s.d.
`
`95%Cl
`
`
`
`Drug input rate and tolerance to frusemide
`
`the analysis of subject 3. This was needed because of the
`extreme deviation of this subset from the remaining data of
`subject 3 and also in comparison to similar subsets of the
`other subjects. For subject 6 imprecise parameter values
`were obtained for kin and Item, but not
`for the other
`parameters. This may be due to the fact
`that
`the first
`datapoint of this subject already was the highest
`in the
`datasct, causing the estimates of [em and kw: to be uncertain.
`Otherwise, a good parameter precision was obtained for
`all subjecis.
`The different infusion rates caused striking differences in
`the time course of tolerance development for both diuresis
`and natriuresis. This is illustrated by Figure 4-, showing the
`calculated values of the modifier M as a function of time. It
`
`can be observed that the time course of tolerance develop—
`ment was very similar for the 10 min and 30 min infusions
`as shown by similar shapes of M. However,
`the time of
`maximum tolerance development occurred later for the
`slower
`infusions of 100 min and 300 min,
`respectively.
`Interestingly,
`the time course of M clearly lagged behind
`with respect to the infusion rate for the 10, 30 and 100 min
`infusions, but not
`for
`the 300 min infusion. Here,
`the
`development of tolerance closely followed the pattern of
`induced diuresis and natriuresis. returning back to baseline
`after the infusion was stopped.
`The time 0mm”) and size (Mmax) of maximum tolerance
`development are shown in Tables 4 and 5, respectively. The
`baseline value for M was subtracted from the peak height
`to yield the net increase caused by the fi'usemide treatment.
`The time of maximum tolerance development appeared
`significantly later for the 300 min infusion compared to the
`10 min and 30 min infusions for both diuresis and natriuresis.
`The mean difference between the tmmM of the 300 min
`infiision and the 10 min infusion was 3.9 h for diuresis and
`4.1 h for natriuresis,
`respectively. The mean difference
`between the 300 min and the 30 min infusion was 3.8 h for
`diuresis and 4.0 h for natriuresis, respectively (Table 4). Also,
`the peak value of M (Mmax) was significantly lower for the
`slowest
`infusion compared to all other infusion rates for
`both diuresis and natriuresis (Table 5).
`In spite of these
`notable differences in the appearance of tolerance develop—
`ment as a function of the infiJsion rate, the AUC values
`of M were rather similar. There was only a significant
`difference between the total AUC of the most rapid and
`the slowest infiasion for diuresis (P< 0.05), but not for the
`other infiision rates. There was no difference in total AUC
`of M for natriuresis. The meanisd. values of total AUC
`for diuresis (ml min—1 h) were: 144;”; for the 10 min,
`15i1.1 for
`the 30 min, 17i1.8 for
`the 100 min and
`18 i3.9 for the 300 min infusion. The meanis.d values
`of total AUC for natriuresis (mmol min_1 h) were 2.1 i 0.5,
`2.2 i 0.3, 2.7i 1.1 and 2.1 i0], respectively.
`Similar results were obtained when diuresis and natriuresis
`
`data were modeled for all subjects pooled together (the
`diuresis data of subject 8 were included in this analysis).
`Also in this case, there was no difference in AUC of M for
`natriuresis. For diuresis,
`the AUC showed a tendency
`towards slight increase with infusion rate.
`In spite of the acute development of tolerance for all four
`infiision rates, no hormonal counter—regulatory actions could
`be observed. Levels of plasma active renin, angiotensin II,
`
`l?
`:4
`E
`"C1i::
`5
`o\°LnCh
`'3L1
`-
`'21
`A
`:75|
`
`ED
`1;
`u
`u
`u
`y:
`H
`~
`5
`v-‘
`W
`ru
`E»
`B
`.2m
`P..i
`.2L4asn:
`E
`"‘2E
`.0
`'«7‘:.0a.
`.E-c.—<17md
`
`m 2
`
`'....
`
`e
`o
`om
`"aEa
`oo_i
`0'to
`o”
`""
`s.
`a)
`:>
`
`oS(
`
`U>
`
`.H
`Cl.)
`
`0EEG
`
`)v;
`.51.:
`er.
`.-
`:1
`O
`H
`'u.‘o
`G)a,
`o
`'5‘
`=
`O
`1'7:
`<5
`=
`"‘
`2‘
`a
`d)
`1..
`di~a
`.2m
`9’
`E
`'1:rs
`L."
`’USa
`5;:
`v)
`<1)
`4—4
`.2"Ca‘
`goo
`fillw-i
`a):
`Eu
`r—1
`NB
`I":
`
`© I998 Blackwell Science Ltd Br j Clin Pharmacol, 46, 479—487
`
`483
`
`IPR2015—OO410
`
`Petitioners' EX. 1032
`
`Page 5
`
`IPR2015-00410
`Petitioners' Ex. 1032
`Page 5
`
`

`

`M. Wakelkamp et al.
`
`Subjea‘
`
`km
`(ml (min 1.)“)
`
`Table 2 PD modeling of diuresis. kin is the zero—order rate constant for production of diuretic response, [com is the first—order rate
`constant for loss of diuretic response, Imax is a parameter representing the maximum effect of the drug on the inhibition function, ICSO
`represents the fiusemide excretion rate producing 50% of Imx, Y is the sigmoidicity {actor and [em] is the first—order rate constant for
`production and loss of M. R0 and Rmax are secondary parameters representing basal diuresis and maximum diuresis including tolerance
`development (basal+drug—induced). Mean (s.d.) n=7.
`Parameter
`ICSU
`(,ug min
`
`kcu:
`(WV
`41
`177
`53
`42
`69
`451
`96
`
`Imzvr
`
`kml
`R0
`Rm.
`
`1)
`y
`(II—1)
`(ml minAI)
`(ml min—1}
`
`0.94
`0.97
`0.96
`0.97
`0.94
`0.97
`0.95
`
`\IAQG\DOOO
`
`11
`
`2.0
`2.5
`1.7
`1.6
`1.1
`1.6
`1.9
`
`0.3
`0.5
`0.6
`0.4
`0.2
`0.3
`0.2
`
`1.9
`1.4
`1.7
`1.6
`1.8
`2.0
`2.3
`
`9
`10
`10
`12
`9
`13
`12
`
`N
`
`00\l0\U1-P<43N>—‘
`
`A)d
`
`.) J l
`
`\nNo:
`
`133
`0.96
`11
`1.8
`0.3
`1.8
`m :15
`148
`0.01
`0.4
`0.1
`0.3
`2
`
`
`Table 3 PD modeling of natriuresis. kin is the zero—order rate constant for production ofnatriuretic response, k0“. is the first—order rate
`constant for loss ofnatriuretic response, Imax is a parameter representing the maximum effect of the drug on the inhibition fiincfion,
`ICSO represents the fi-usemide excretion rate producing 50% of [mm 7 is the sigmoidicity factor and kwl is the first—order rate constant for
`production and loss of M. R0 and Rmx are secondary parameters representing basal natriuresis and maximum nattiuresis including
`tolerance development [basal—l—drug—induced). Mean (s.d.) n=8.
`Parameter
`
`kin
`kour
`IC50
`km:
`R0
`Rmax
`
`Subject
`(mmol (min 111/Hi)
`(I’l— 1)
`In.”
`flag min_ 1)
`y
`(ll—1)
`(mmol mini )
`(mmol min— 1)
`
`1.5
`0.12
`0.8
`2.5
`10
`0.96
`134
`18
`1
`1.7
`0.11
`0.8
`2.3
`9
`0.97
`910
`111
`2
`1.7
`0.09
`0.9
`1.7
`5
`0.98
`467
`44
`3
`2.2
`0.08
`0.8
`1.7
`5
`0.99
`369
`33
`4
`1.5
`0.11
`0.5
`1.6
`5
`0.97
`365
`45
`5
`2.9
`0.14
`3.7
`2.6
`13
`0.99
`1589
`263
`6
`1.8
`0.13
`0.4
`2.3
`1 1
`0.97
`447
`66
`7
`1.9
`0.09
`0.4
`1.3
`3
`0.98
`456
`47
`8
`1.9
`0.11
`1.0
`2.0
`8
`0.98
`592
`7
`Mean
`
`
`
`
`
`
`
`
`80 457 0.01 3 0.5 1.1 0.02s.d. 0.5
`
`Table 4 The time of maximum tolerance development (tmaw) after an infusion dose of 10 mg fi'usemide given over 10, 301 100 and
`300 min.
`
`
`thfoT
`tmaxMfi"
`Namtirzsis (h)
`Dimesis {h}
`Iryiasion time (min)
`Iryfitsian time (min)
`Subject
`10
`30
`100
`500
`Subject
`10
`30
`100
`300
`
`
`5.06
`1.91
`1.19
`0.90
`1
`5.17
`2.03
`1.55
`1.15
`1
`5.03
`1.87
`0.71
`0.78
`2
`5.05
`1.95
`0.97
`0.90
`2
`5.03
`—
`1.01
`1.01
`3
`5.05
`—
`1.15
`1.24
`3
`5.04
`1.85
`0.98
`0.85
`4
`5.07
`1.91
`1.18
`1.00
`4
`5.06
`2.11
`1.46
`1.26
`5
`5.28
`2.66
`2.16
`2.17
`5
`5.18
`1.77
`0.65
`0.43
`6
`5.26
`2.12
`1.27
`1.15
`6
`5.07
`1.90
`1.07
`0.90
`7
`5.13
`1 .99
`1.23
`1 .05
`7
`5.10
`2.07
`1.23
`1.47
`8
`—
`—
`—
`—
`8
`5.07
`1.93
`1.04**
`095m
`Mean
`5.14
`2.11
`136*
`1.24***
`Mean
`
`
`
`
`
`
`
`
`
`0.43 0.39 0.28 0.10 s.d. 0.31 0.27 0.12s.d. 0.05M“
`
`Differences in treatments were analyzed by Friedman’s ANOVA for repeated measures, followed by Dunn’s test for pairwise comparisons. *:P<0.05,
`**:P<0.01, ***:P<0.001 when compared with the 300 min infusion.
`
`484
`
`© I998 Blackwell Science Ltd Brj Clin Pharmacol, 46, 479—487
`
`IPR2015—00410
`
`Petitioners' EX. 1032
`
`Page 6
`
`IPR2015-00410
`Petitioners' Ex. 1032
`Page 6
`
`

`

`Drug input rate and tolerance to frusemide
`
`Table 5 The size of maximum tolerance development (me) after an infiision dose of 10 mg frusemide given over 10, 30, 100 and
`300 min.
`
`
`Subjett
`
`10
`
`30
`
`i "”fo natriuresi:
`Mmmfln diuresis
`(mmol mirf 1)
`(ml min—1)
`Infisian time (min)
`Inflation time (min)
`100
`300
`Subject
`10
`30
`100
`300
`
`
`0.41
`0.82
`0.87
`0.82
`1
`2.80
`3.63
`3.66
`3.32
`l
`0.45
`0.82
`0.81
`0.85
`2
`3.89'
`5.57
`4.79
`5.01
`2
`0.42
`—
`0.83
`0.94
`3 ,
`3.29
`—
`4.87
`5.36
`3
`0.58
`1.04
`1.03
`0.85
`4
`3.74
`5.27
`4.88
`4.15
`4
`0.36
`0.62
`0.59
`0.54
`5
`2.02
`2.25
`2.07
`1.93
`5
`0.37
`1.48
`2.29
`2.35
`6
`2.82
`3.92
`3.95
`3.84
`6
`0.25
`0.76
`0.71
`0.65
`7
`1.90
`3.89
`3.52
`3.25
`7
`0.51
`0.86
`0.67
`0.75
`8
`7
`—
`—
`—
`8
`0,42
`0.91"
`097*
`097*
`Mean
`2.92
`4.09***
`3.96***
`3.84**
`Mean
`
`
`
`
`
`
`
`
`
`0.57 0.55 0.28S.Cl.0.781.201.021.16s.d. 0.10
`
`Differences in treatments were analyzed by repeated measures ANOVA, followed by the Bonferroni multiple comparisons test for pairwise comparisons.
`*;P<0.05, **:P<0.01. ***:P<0.001 when compared with the 300 min infilsion.
`
`aldosterone and ANP did not Show any systematic changes
`during the day. The catecholamine and ADH values could
`not be evaluated, due to technical problems with the analyses.
`
`Discussion
`
`Much is unknown about the factors that cause variability in
`pharmacodynamics and the mechanisms involved [20]. In
`the present study, there was a striking difference in the time
`course and appearance of tolerance development to fruseni—
`ide, depending on die rate of input of the drug. Distinct
`clockwise hysteresis loops indicating tolerance development
`were observed for the slower infiisions, in contradistinction
`to the most rapid infusion rate of frusemide.
`Instead, a
`counter—clockwise hysteresis indicated the delay between
`the frusemide excretion rate and the diuretic eflect. In this
`
`case, the drug input rate exceeded the rate of loss of response
`and tolerance development
`throughout
`the duration of
`the infusion.
`
`the AUC of the modifier M, used to
`In this study,
`quantify the development of tolerance after each infusion,
`was rather similar for all treatments. However, the similarity
`in AUC values may be a consequence of the characteristics
`of the present feedback model rather than a demonstration
`that
`total
`tolerance development was the same.
`In this
`model, the AUC value of M is not independent from the
`AUC of R (total response), which was the same for all
`infusions. Nevertheless, the similarity in total diuretic and
`natriuretic response and total
`fi'usernide recovery may
`indicate that there was no or very little difference in total
`tolerance development
`to the same dose. The onset of
`tolerance development was much more rapid for the 10 and
`30 min infusions, compared to the slower input rates. For
`the slowest infiision,
`tolerance had reached its maximum
`only after more than 5 h for both diuresis and natriuresis.
`Also, the peak value of tolerance development was much
`lower for the 300 min infusion compared to the other
`treatments.
`
`In our study, the kinetics of the response variable R was
`found to be very fast in comparison to the kinetics of the
`
`the value of Icon: was high in
`is,
`regulator M. That
`comparison to km]. For diuresis,
`the mean half—life for
`tolerance development was 139 min, whereas the mean
`half—life for the first—order loss of response (leout) was only
`0.3 min. For natriuresis,
`the mean half—lives were 42 and
`0.1 min, respectively. The relatively low value of km] rate-
`limited the development of tolerance, causing this to clearly
`lag behind the infusion rate and the diuretic and natriuretic
`response. For
`the most
`rapid infusion,
`the clockwise
`hysteresis was even entirely obscured by the counter—
`clockwise hysteresis loop caused by the indirect response of
`the drug [12, 17]. However, the level of response after the
`slowest 300 min infusion was low enough to enable M to
`exert a more direct feedback, leading to a large clockwise
`hysteresis loop.
`It has been suggested that the influence of drug input rate
`on tolerance development is not only dependent on the rate
`of administration itself, but also on the mechanism respon—
`sible for tolerance [11]. Castaneda et al. approached this
`issue by considering tolerance as either being caused by
`compensatory homeostatic (counter—regulatory) mechanisms
`or by more intrinsic processes (‘true tolerance’). For drugs
`that elicit compensatory homeostatic mechanisms, such as
`nifedipine, a slow input of the drug would trigger fewer
`homeostatic reactions, leading to an increased pharmacologi—
`cal response. For drugs that elicit ‘true tolerance’, such as
`nitrates, a slow input would enhance the development of
`tolerance, reducing the response [11].
`With nifedipine [21],
`the concentration-effect relation—
`ships for two infusion rates were profoundly different. This
`was suggested to be caused by stronger counter—regulation
`in baroreceptor reflex activation of the higher infusion rate.
`Similar observations have been reported for other calcium
`antagonists such as nisoldipine [22] and other vasodilator
`agents such as prazosin [23]. On the other hand, a slow and
`constant drug input rate has been found disadvantageous for
`the anti—anginal effect of nitrates, because tolerance occurs
`after several doses [24]. Psychomotor and subjective adverse
`effects of diazepam were found to be less after
`the
`administration of controlled release capsules compared to
`
`© I998 Blackwell Science Ltd Br 1 Clin Pharmacol, 46, 479—487
`
`485
`
`IPR2015—00410
`
`Petitioners' EX. 1032
`
`Page 7
`
`IPR2015-00410
`Petitioners' Ex. 1032
`Page 7
`
`

`

`M. Wakelkamp et al.
`
`plain tablets in healthy volunteers [25] despite similar plasma
`concentrations. This was attributed to stronger tolerance
`development
`for
`the
`controlled release
`formulation.
`Administranon of an intravenous bolus dose of morphine to
`rats causes less tolerance than two different constant rate
`infusions [26]. However, tolerance development was stronger
`for the higher than for the lower infusion rate. For many
`drugs, it may prove diflicult to separate counter—regulation
`from ‘true tolerance’. In many cases,
`the mechanisms for
`the development of