Eur J Clin Pharmacol (2002) 58: 453–458
`DOI 10.1007/s00228-002-0502-1
`
`P H A R M A C O K I N E T I C S A N D D I S P O S I T I O N
`
`Ola Junghard Æ Mohammed Hassan-Alin
`Go¨ ran Hasselgren
`The effect of the area under the plasma concentration vs time
`curve and the maximum plasma concentration of esomeprazole
`on intragastric pH
`
`Received: 7 November 2001 / Accepted: 2 July 2002 / Published online: 3 September 2002
`Ó Springer-Verlag 2002
`
`Objective: The aim of this study was to create a useful
`model of the effect of the area under the plasma con-
`centration vs time curve (AUC) and the maximum
`plasma concentration (Cmax) of esomeprazole on intra-
`gastric pH, measured as the percentage of total time with
`intragastric pH above 4 (%pH>4) during a 24-h period.
`Methods: The evaluation is based on esomeprazole data
`from two crossover studies. In the first study (n=36),
`intragastric pH and plasma concentrations were mea-
`sured on day 5 of repeated once-daily 20-mg and 40-mg
`doses of esomeprazole during fasting conditions. In the
`second study (n=24), measurements were made on days
`1 and 5 of repeated once-daily dosing with 40 mg of
`esomeprazole under fasting and fed conditions. A model
`was applied in which the logistic function of %pH>4
`was assumed to be linearly dependent on log-trans-
`formed AUC and Cmax. The effects of repeated dosing
`and of fed relative to fasting conditions were included in
`the model, and the interindividual variation in %pH>4
`was accounted for.
`Results: The effect of the pharmacokinetic variables
`AUC and Cmax of esomeprazole on %pH>4 can be
`adequately described by a model using a logistic func-
`tion for %pH>4 and a normally distributed error. In
`this model, log-transformed AUC and Cmax were both
`statistically significant. The model showed that
`for
`a fixed AUC, a decrease in Cmax gives an increase in
`%pH>4. A decrease in AUC, keeping Cmax fixed, gives
`a decrease in %pH>4, but a simultaneous decrease in
`Cmax and AUC will result in a less pronounced decrease
`in %pH>4. The model may be used for predicting
`differences in %pH>4 between two formulations, based
`on assessments of AUC and Cmax. Repeated dosing gave
`an increased %pH>4, where approximately half of the
`increase stemmed from increased AUC and Cmax, and
`the rest could be attributable to the persistent blockade
`
`O. Junghard (&) Æ M. Hassan-Alin Æ G. Hasselgren
`AstraZeneca R&D Mo¨ lndal, 431 83 Mo¨ lndal, Sweden
`E-mail: ola.junghard@astrazeneca.com
`Tel.: +46-31-7761881
`
`of the proton pumps. Food intake reduced AUC and
`Cmax but had no obvious effect on %pH>4, which is
`explained by a prolonged time period with quantifiable
`plasma concentrations.
`Conclusion: The effect of the pharmacokinetic variables
`AUC and Cmax of esomeprazole on %pH>4 can be
`adequately described by a model using a logistic func-
`tion for %pH>4 and a normally distributed error.
`
`Keywords Esomeprazole Æ Intragastric pH Æ
`Pharmacokinetics
`
`Introduction
`
`Esomeprazole is the first proton pump inhibitor (PPI)
`developed as an optical isomer (S-omeprazole) for the
`for
`treatment of acid-related diseases. The target
`esomeprazole is the proton pump in parietal cells. The
`compound accumulates in the acidic compartment of the
`parietal cells, where the molecule is transformed to its
`active form, the sulphenamide, which produces a pro-
`found suppression of gastric acid secretion by irrevers-
`ible binding and blocking of the H+/K+ adenosine
`triphosphatase (ATPase), the proton pump.
`The mode of action of PPIs is rather complex, since
`the proton pumps in parietal cells can not be blocked
`unless they are in an active, secreting stage [1]. The
`longer the plasma concentration remains above a certain
`level, the larger the number of proton pumps that can be
`effectively inhibited as they become continually activated.
`Thus, in theory, a prolongation of the absorption phase
`could result in better efficacy due to a more protracted
`plasma concentration curve. The shape of the plasma
`concentration vs time curve is largely determined by the
`relation between the area under the plasma concentra-
`tion vs time curve (AUC) and the maximum plasma
`concentration (Cmax). When developing different drug
`formulations or evaluating other factors influencing
`absorption, a model
`characterising the
`interplay
`between AUC, Cmax, and pharmacodynamic effects is
`
`MYLAN PHARMS. INC. EXHIBIT 1068 PAGE 1
`
`MYLAN PHARMS. INC. EXHIBIT 1068 PAGE 1
`
`

`

`454
`
`a very helpful tool for estimating and explaining the
`resulting efficacy changes, in this context measured as
`intragastric pH.
`is a
`Gastro-oesophageal reflux disease (GORD)
`common disorder usually diagnosed by the presence of
`heartburn and acid regurgitation and/or the presence of
`oesophageal lesions. The development of symptoms and
`mucosal injury in patients with GORD is highly de-
`pendent on intragastric pH [2], and those with a re-
`fluxate of pH<4 experience more intense symptoms [3,
`4]. There is a clear, positive correlation between the
`fraction of time with intragastric pH>4 and healing of
`oesophageal lesions after 8 weeks of antireflux treatment
`[5]. The percentage of time with intragastric pH>4 has
`consequently been accepted as a benchmark variable in
`the large number of recent comparisons between various
`proton pump inhibitors [6]. Consequently, we used
`fraction of time with intragastric pH greater than 4
`(%pH>4) as the efficacy measure in this study.
`Repeated dosing with esomeprazole increases AUC
`and Cmax as well as intragastric pH [7, 8]. The increased
`plasma concentrations after repeated administration of
`esomeprazole can be due to inhibition of CYP2C19. The
`increase in intragastric pH stems partly from the in-
`creased AUC and Cmax and partly from the persistent
`blockade of the H+/K+ ATPase during the remaining
`life cycle of the proton pumps, which is 3 days on av-
`erage [9]. The relation between these two contributors
`and the increase in %pH>4 has not been quantified.
`Food intake influences AUC and Cmax of esomep-
`razole, but it does not seem to affect %pH>4. The
`reason for this has not been studied.
`The aim of the present investigation was to:
`
`1. Develop a model of the relationship between AUC,
`Cmax, and intragastric pH in order to be able better to
`predict and explain the outcome based on changes of
`the two pharmacokinetic variables
`2. Quantify the contributions of AUC, Cmax, and the
`irreversible inhibition of
`the proton pump on
`%pH>4 after repeated dosing
`3. Examine reasons for the lack of influence of food
`intake on %pH>4
`
`Methods
`
`Subjects
`
`Esomeprazole data from two crossover studies were evaluated. The
`studies were conducted in accordance with the Declaration of
`Helsinki and approved by the ethics committees of the University
`of Gothenburg, Sweden, and the Commission for Ethical Questions
`at the University of Bern, Switzerland. Signed informed consent
`was obtained from all subjects before enrollment in the studies.
`In the first study [10] (study 1, n=36), intragastric pH and
`plasma concentrations were measured on day 5 of repeated once-
`daily doses of 20 mg and 40 mg of esomeprazole and 20 mg of
`omeprazole during fasting conditions. Only esomeprazole data
`were used in this evaluation. In the second study (study 2, n=24),
`measurements were made after a single dose on day 1 and after
`repeated once-daily dosing on day 5 with 40 mg of esomeprazole
`under fasting and fed conditions of drug administration (with drug
`intake after a standardised breakfast). Description of the two
`studies and summarised demographic data are shown in Table 1.
`
`Measurements
`
`Intragastric pH and plasma concentrations of esomeprazole were
`measured on day 5 of repeated once-daily dosing in study 1 and on
`days 1 and 5 in study 2. A microelectrode (Ingold bipolar glass)
`attached to a battery-powered solid-state data logger was used for
`the pH recording. A 2-point calibration of the electrode was per-
`formed before and after each 24-h recording. The electrode was
`inserted transnasally and placed 10 cm below the lower oesopha-
`geal sphincter during ongoing pH recording. For each 24-hour pH
`recording, %pH>4 was calculated.
`Blood samples for the assay of esomeprazole in plasma were
`taken just before drug administration and 0.5, 1, 1.5, 2, 2.5, 3, 4, 6,
`and 8 h after drug administration in study 1. In study 2, they were
`taken before and 0.5, 0.75, 1, 1.25, 1.5, 2, 2.5, 3, 3.5, 4, 4.5, 5, 5.5, 6,
`7, 8, 10, 12, and 14 h after drug administration. The samples were
`collected in heparinised tubes and centrifuged, and the plasma was
`frozen until analysis.
`Plasma samples were analysed for esomeprazole using normal
`phase liquid chromatography with UV detection at AstraZeneca
`R&D Mo¨ lndal [11]. The limit of quantification for this method is
`25 nmol/L, with a coefficient of variation under 20%.
`The AUC is the total area under the plasma concentration vs
`time curve and was calculated by the log/linear trapezoidal method
`from time zero to the last measurable plasma concentration, Clast,
`and by extrapolation to infinity by Clast/k. The elimination rate
`constant (k) was estimated from individual linear regression on the
`terminal part of the log plasma concentration vs time curve and
`included at least the last three data points. Cmax is the observed
`maximum plasma concentration.
`
`Table 1. Description of the
`studies and summary of demo-
`graphic data. GORD gastro-
`oesophageal reflux disease
`
`Study 1 (n=36)
`
`Study 2 (n=24)
`
`Study design
`Treatments
`Periods
`Administration
`Assessments
`Population
`Gender (n) males/females
`Mean age in years (range)
`Mean weight in kg (range)
`Mean height in cm (range)
`
`Crossover
`E40, E20, (O20)a
`3
`Capsules
`Day 5c
`GORD patients
`15/21
`45 (29–58)
`80 (46–108)
`171 (154–190)
`
`Crossover
`E40 fed, E40 fastingb
`2
`Capsules
`Days 1, 5c
`Healthy subjects
`12/12
`30 (23–43)
`67 (50–89)
`173 (155–190)
`
`aE40, 40 mg esomeprazole; E20, 20 mg esomeprazole; O20, 20 mg omeprazole (O20 data are not used
`in the evaluation)
`bFed and fasting refer to conditions of drug administration
`cDay 5 of repeated once-daily dosing
`
`MYLAN PHARMS. INC. EXHIBIT 1068 PAGE 2
`
`MYLAN PHARMS. INC. EXHIBIT 1068 PAGE 2
`
`

`

`Fig. 1. Percentage of time with intragastric pH>4 (%pH>4) vs
`AUC, data from study 1 (n=36) with repeated dosing of 20 mg and
`40 mg of esomeprazole. The lines connect values from the same
`patient
`
`455
`
`is a parameter to be estimated together with EC50 and Emax. In this
`application, E denotes the expected value of %pH>4, and Emax is
`set to 100%.
`Equation 1 may be rewritten as
`logðE=ðEmax EÞÞ ¼ n logðEC50Þ þ n log C
`
`ð2Þ
`
`ð3Þ
`
`Equation 2 relates the effect, %pH>4, on a single concentra-
`tion C. In order to evaluate the combined effect of AUC and Cmax
`is replaced by (b1log
`term (n log C)
`on %pH>4,
`the last
`AUC+b2log Cmax). By inserting Emax=1 and –nÆlog(EC50)=b0,
`equation 2 then becomes
`Þ ¼ b0 þ b1 log AUC þ b2 log Cmax
`
`log E=ð1 EÞð
`The expression b1log AUC+b2log Cmax in equation 3 equals
`(b1+b2)log AUC – b2log(AUC/Cmax). The last expression reveals
`that the logistic function of %pH>4 is a linear combination of log-
`transformed AUC and the log-transformed ratio AUC/Cmax.
`Viewing the plasma concentration vs time curve as a triangle, with
`AUC being the area and Cmax the height, the ratio AUC/Cmax is
`proportional to the base of the triangle, which may be interpreted as a
`measure of the time period with quantifiable plasma concentrations.
`Equation 3 describes the model for an individual subject. We
`now assume that the models for individual subjects differ only by
`different intercepts, b0. Thus, for each subject included in the
`analysis, a term ciSi is added to the right-hand side of the equation,
`where ci is a parameter and Si is a dummy variable which takes the
`value 1 for subject i and 0 for all other subjects. Other effects, for
`example the effect of the irreversible inhibition of the proton pump,
`are similarly assumed to affect the model only by a term added to
`the right-hand side of the equation.
`In study 1, two doses of esomeprazole were given. The 40-mg
`dose is associated with higher AUC and Cmax and also a higher
`%pH>4 than the 20-mg dose (Table 2). In order to evaluate
`whether the higher %pH>4 for the 40-mg dose could be explained
`by some dose-related effect in addition to the effect of increased
`AUC and Cmax, a term bDoseDose40, where Dose40 is a dummy
`variable and bDose a parameter, was added, giving
`Þ¼ b0þb1 log AUCþb2logCmaxþc1S1þ :::þcmSm
`
`log Eð1EÞð
`þbDoseDose40
`
`ð4Þ
`
`Fig. 2. Percentage of time with intragastric pH>4 (%pH>4) vs
`log-transformed AUC, data from study 1 (n=36), with repeated
`dosing of 20 mg and 40 mg of esomeprazole. The AUC unit is
`lmol · h/L. The lines connect values from the same patient
`
`Model building
`
`Figure 1 and Fig. 2 show data from study 1. The %pH>4 is
`plotted in Fig. 1 vs untransformed AUC and in Fig. 2 vs log-
`transformed AUC. Some subjects did reach 100% or close to it,
`and we will assume that with sufficiently high plasma concentra-
`tions during the 24-h period, this is valid for all subjects. Most
`subjects will have a %pH>4 above zero even without drug intake,
`but we will assume that %pH>4 is zero when no plasma con-
`centrations are detected. Thus, for an individual subject, one as-
`sumption is that %pH>4 increases from 0 and approaches 100%
`when the plasma concentrations are sufficiently high during the
`24-h period. Figure 1 and Fig. 2 indicate that a reasonable starting
`point would be a model where, for an individual subject, %pH>4
`describes a logistic type of curve when plotted against log-trans-
`formed AUC. Such a relationship is the basis for the sigmoid Emax
`model, and we will modify this model to include effects of AUC,
`Cmax, repeated dosing, and food intake.
`The sigmoid Emax model is usually expressed as [12]
`ð1Þ
`E ¼ EmaxCn=ððEC50Þn þ CnÞ
`where E is the effect, Emax is the maximum effect, C is the con-
`centration, EC50 is the concentration needed for 50% effect, and n
`
`This is the model used for separate analysis of study 1. In study
`2, esomeprazole was given both during fasting conditions and after
`a standardised breakfast (fed condition), and measurements were
`taken on day 1 and day 5. The AUC, Cmax, and %pH>4 are
`summarised in Table 2, and mean plasma concentration vs time
`curves are presented in Fig. 3. Repeated dosing increases all three
`variables, whereas fed conditions give decreased AUC and Cmax
`but affect %pH>4 very little compared to fasting conditions. In
`order to quantify the effect of the irreversible inhibition of the
`proton pump, a term bDay5D5 was added, where bDay5 is a pa-
`rameter and D5 a dummy variable that takes the value 1 for
`measurements made on day 5 and the value 0 for measurements on
`day 1. Similarly, in order to evaluate whether there is an additional
`effect of food intake, additional to the effect of changed AUC and
`Cmax, a term bfedCfed was added, giving the model for study 2.
`
`
`
`log E=ð1EÞð
`
`Þ¼b0þb1 logAUCþb2 logCmaxþc1S1þ:::þcmSm1
`þbDay5D5þbfedCfed
`
`ð5Þ
`
`Statistical methods
`
`The two studies were analysed first separately, and statistically
`nonsignificant study-specific effects were identified. Using pooled
`data from all subjects, analysis of the combined data was then
`performed with a model in which effects found to be not statisti-
`cally significant in the first analyses were removed.
`
`MYLAN PHARMS. INC. EXHIBIT 1068 PAGE 3
`
`MYLAN PHARMS. INC. EXHIBIT 1068 PAGE 3
`
`

`

`456
`
`Table 2. Summary of
`assessments of AUC, Cmax, and
`%pH>4
`
`Treatment
`
`AUC (lmol · h/L)
`Geometric
`mean (range)
`
`Cmax (lmol/L)
`Geometric
`mean (range)
`
`Study 1
`Study 1
`Study 2
`Study 2
`Study 2
`Study 2
`
`E20a
`E40a
`Fasting day 1
`Fed day 1
`Fasting day 5
`Fed day 5
`
`4.2 (0.9–14.9)
`12.8 (4.0–25.6)
`4.4 (1.6–12.1)
`1.9 (0.7–5.1)
`10.4 (5.4–17.9)
`5.8 (2.4–13.8)
`
`2.1 (0.5–4.8)
`4.7 (1.6–9.6)
`2.9 (1.3–6.2)
`0.6 (0.1–2.5)
`4.7 (3.1–7.9)
`1.8 (0.7–4.3)
`
`aE20, 20 mg esomeprazole; E40, 40 mg esomeprazole
`
`%pH>4
`Mean (range)
`
`52 (14–99)
`70 (19–100)
`27 ( 1–58)
`23 ( 10–61)
`59 (19–88)
`56 (25–76)
`
`log-transformed AUC and log-transformed Cmax were
`estimated to be 1.02 and –0.47, respectively.
`In the analysis of the two studies combined, using the
`model described by equation 5 but without the term for
`fed conditions, AUC, Cmax, and the effects of subjects
`and the irreversible inhibition of the proton pump were
`all statistically significant (P<0.0001). The parameter
`estimates are given in Table 3. The residuals (observed
`%pH>4 minus predicted %pH>4) show a constant
`variance, with negative and positive residuals evenly
`spread across the range of predicted values (Fig. 4),
`which indicates an adequately fitted model.
`In study 2, during fasting conditions, mean %pH>4
`increased from 27% to 59% (Table 2) from day 1 to day
`5. From the model and the estimated parameters, it
`could be calculated that the raised AUC and Cmax would
`increase %pH>4 from 27% (day 1) to only 42% (day
`5). Thus, around half of the total increase in %pH>4 is
`attributable to raised AUC and Cmax, and the rest is
`attributable to the irreversible inhibition of the proton
`pump. This relation holds also during fed conditions.
`The influence of Cmax on %pH>4 was negative and
`that of AUC was positive (Table 3). Figure 5 illustrates
`
`Table 3. Parameter estimates in the final model
`
`Parameter
`
`Variable/factor
`
`Estimate (95% CI)
`
`b1
`b2
`bDay5
`
`Log (AUC)
`Log (Cmax)
`Day 5
`
`1.11 (0.84 to 1.38)
`–0.53 (–0.77 to –0.30)
`0.75 (0.52 to 0.97)
`
`Fig. 4. Residuals (observed minus predicted values of %pH>4) vs
`%pH>4, predicted by a model including log-transformed AUC,
`log-transformed Cmax, and factors for subjects and repeated dosing
`
`Fig. 3. Mean plasma concentrations of esomeprazole following
`single (day 1) and repeated (day 5) oral administration of 40-mg
`esomeprazole capsules after food intake and during fasting
`conditions
`
`To support the use of log-transformed AUC and Cmax, the data
`were reanalysed using untransformed AUC and/or Cmax. Log-
`transformed AUC and Cmax were also analysed one by one in order
`to evaluate which variable is the best predictor of %pH>4.
`The observed %pH>4 values are assumed to be generated
`from the respective model by a normally distributed error term
`added to the expected %pH>4. The parameters b0, b1, and b2
`together with the parameters for the dummy variables are esti-
`mated by fitting the model to the data. The deviance is used as a
`measure of goodness-of-fit. In this analysis, with normally dis-
`tributed error, the deviance divided by the degrees of freedom
`(deviance/df) is an estimate of the error variance. The smaller the
`error variance, the better the model fits the data.
`The SAS procedure GENMOD, with a logit link function and
`normal distribution, was used for the analysis.
`
`Results
`
`Analysis of study 1 showed that the dose effect, mea-
`sured by the parameter bDose in equation 4, was not
`statistically significant (P=0.41). Removing the dose
`effect from the model, the parameters for log-trans-
`formed AUC and log-transformed Cmax, b1, and b2 in
`equation 4 were estimated to be 1.20 and –0.66, re-
`spectively.
`Analysis of study 2 showed that the additional effect
`of food, measured by the parameter bfed in equation 5,
`was not statistically significant (P=0.38), whereas the
`parameter bDay5 was statistically significant (P<0.0001)
`and estimated at 0.75. After removing the term for fed
`conditions
`from the model,
`the parameters
`for
`
`MYLAN PHARMS. INC. EXHIBIT 1068 PAGE 4
`
`MYLAN PHARMS. INC. EXHIBIT 1068 PAGE 4
`
`

`

`457
`
`obvious effect on %pH>4. However, the relative de-
`crease in Cmax is more pronounced than the relative
`decrease in AUC. This means that the form of the
`plasma concentration vs time curve became lower on
`average but wider with a longer duration of quantifiable
`plasma concentrations, as illustrated by Fig. 3.
`In the same way, the plasma concentration vs time
`curve may look different for different formulations of
`the drug. For example, a tablet that releases the active
`compound more slowly than a standard formulation
`may give a lower AUC but at the same time an even
`more pronounced decrease in Cmax. The model predicts
`that two different formulations of esomeprazole will
`have an equal effect on %pH>4 whenever the ratio of
`Cmax for formulation 1 to that for formulation 2 is re-
`lated to the corresponding ratio in AUC values by
`Cmax 1=Cmax 2 ¼ ðAUC1=AUC2ÞB
`where B=–b1/b2. Inserting the parameter estimates
`from Table 3 gives B an estimated value of 2. Thus, the
`two formulations could be expected to have an equal
`effect on %pH>4 when the ratio in Cmax equals the
`squared ratio in AUC. This relation between AUC and
`Cmax holds approximately when comparing fed and
`fasting conditions in study 2 (Table 2).
`This model is based on a number of assumptions and
`approximations. However, the parameter estimates in
`the current model seem rather robust. For example, the
`lack of effect of food intake on %pH>4, seen in study 2,
`could have been predicted by the parameter estimates
`from the separate analysis of study 1 together with the
`AUC and Cmax values for study 2.
`In conclusion,
`the effect of AUC and Cmax of
`esomeprazole on %pH>4 can be described by a useful
`model with an adequate fit. After repeated dosing,
`around half of the increase in %pH>4 stemmed from
`increased AUC and Cmax, and half of the increase can be
`attributable to the persistent blockade of the proton
`pumps. The reduced AUC after food intake is com-
`pensated by a prolonged time period with quantifiable
`plasma concentrations, which explains the lack of in-
`fluence of food intake on %pH>4.
`
`References
`
`1. Hirschowitz BI, Keeling D, Lewin M, Okabe S, Parsons M,
`Sewing K, Wallmark B, Sachs G (1995) Pharmacological as-
`pects of acid secretion. Dig Dis Sci 40 [Suppl 2]:3S–23S
`2. Joelsson B, Johnsson F (1989) Heartburn – the acid test. Gut
`30:1523–1525
`3. Hunt RH (1999) Importance of pH control in the management
`of GERD. Arch Intern Med 159:649–657
`4. Smith JL, Opekun AR, Larkai E et al (1989) Sensitivity of the
`esophageal mucosa to ph in gastroesophageal reflux disease.
`Gastroenterol 96:683–689
`5. Bell NJ, Burget D, Howden CW, Wilkinson J, Hunt RH (1992)
`Appropriate acid suppression for the management of gastro-
`oesophageal reflux disease. Digestion 51 [Suppl 1]:59–67
`6. Ro¨ hss K, Wilder-Smith C, Claar Nilsson C, Lundin C, Has-
`selgren G (2001) Esomeprazole 40 mg provides more effective
`
`Fig. 5. Relationship between AUC (lmol · h/L), Cmax (lmol/L),
`and expected %pH>4 for a subject with %pH>4=60% when
`AUC=10 and Cmax=4. The upper curve corresponds to Cmax=1
`and the lower to Cmax=8
`
`how %pH>4 depends on AUC and Cmax for a given
`subject. Thus, for a fixed AUC, an increase in Cmax is
`expected to reduce %pH>4. For a fixed Cmax, on the
`other hand, %pH>4 increases rapidly with increased
`AUC.
`The model with log-transformed AUC and Cmax
`values gives a better fit
`(deviance/df=0.0077)
`than
`models where AUC and/or Cmax are untransformed
`(deviance/df ranges from 0.0082 to 0.0096).
`Reducing the model by excluding Cmax or AUC, one
`at a time, shows that %pH>4 is somewhat better pre-
`dicted by a model including AUC (deviance/df=0.0091)
`than by a model including Cmax (deviance/df=0.0126).
`Thus, AUC seems to be a better predictor of %pH>4
`than Cmax.
`
`Discussion
`
`The results show that the effect of the pharmacokinetic
`variables AUC and Cmax of esomeprazole on %pH>4
`can be adequately described by a model using a logistic
`function for %pH>4 and a normally distributed error.
`The use of log-transformed AUC and log-transformed
`Cmax in the model is supported by a worse fit (larger
`deviance) for models with untransformed AUC and/or
`untransformed Cmax.
`According to this analysis, the parameter bDay5 was
`statistically significant. This means that there is an effect
`on %pH>4 of the irreversible blockade of the H+/K+
`ATPase during the remaining life cycle of the proton
`pump, additionally to the effect of raised AUC and Cmax
`from day 1 to day 5.
`In contrast, the parameter bfed was not statistically
`significant in the separate analysis of study 2, and thus
`there is no evidence of an additional effect of food intake
`on %pH>4. It may be surprising that the lower AUC
`and Cmax during fed conditions compared to the AUC
`and Cmax during fasting conditions (Table 2) had no
`
`MYLAN PHARMS. INC. EXHIBIT 1068 PAGE 5
`
`MYLAN PHARMS. INC. EXHIBIT 1068 PAGE 5
`
`

`

`458
`
`acid control than standard doses of all other proton pump
`inhibitors. Gastroenterol 120 [Suppl 1]:2140
`7. Andersson T, Hassan-Alin M, Hasselgren G, Ro¨ hss K, Wei-
`dolf L (2001) Pharmacokinetic studies with esomeprazole, the
`S-isomer of omeprazole. Clin Pharmacokinet 40:411–426
`8. Wilder-Smith C, Ro¨ hss K, Lundin C, Rydholm H (2000)
`Esomeprazole 40 mg provides more effective acid control than
`pantoprazole 40 mg. Gastroenterol 118 [Suppl 2]:A22–A23
`9. A¨ belo A, Eriksson UG, Karlsson MO, Larsson H, Gabrielsson
`J (2000) A turnover model of irreversible inhibition of gastric
`acid secretion by omeprazole in the dog. J Pharmacol Exp Ther
`295:662–629
`
`10. Lind T, Rydberg L, Kyleba¨ ck A et al (2000) Esomeprazole
`provides improved acid control vs. omeprazole in patients with
`symptoms of gastro-oesophageal
`reflux disease. Aliment
`Pharmacol Ther 14:861–867
`11. Lagerstro¨ m PO, Persson BA (1984) Determination of omep-
`razole and metabolites in plasma and urine by liquid chro-
`matography. J Chromatogr 309:347–356
`12. Gabrielsson J, Weiner D (1997) Pharmacokinetic and phar-
`macodynamic data analysis: concepts and applications, second
`edn. Swedish Pharmaceutical Press, Stockholm, pp 187–190
`
`MYLAN PHARMS. INC. EXHIBIT 1068 PAGE 6
`
`MYLAN PHARMS. INC. EXHIBIT 1068 PAGE 6
`
`

`

`Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
`
`MYLAN PHARMS. INC. EXHIBIT 1068 PAGE 7
`
`MYLAN PHARMS. INC. EXHIBIT 1068 PAGE 7
`
`