`
`www.elsevier.nl/locate/ ejps
`
`Formulation-dependent food effects demonstrated for nifedipine
`modified-release preparations marketed in the European Union
`*
`b
`a
`a
`c
`¨
`Barbara S. Schug
`, Erich Brendel , Dorte Wolf , Meinolf Wonnemann , Manfred Wargenau ,
`a
`Henning H. Blume
`aSocraTec R&D, Feldbergstr.59, D-61440 Oberursel, Germany
`bBayer AG, Institute of Clinical Pharmacology, Aprather Weg, D-42096 Wuppertal, Germany
`¨
`M.A.R.C.O., Institute for Biomedical Statistics, Markenstr.5–13, D-40227 Dusseldorf, Germany
`
`a ,
`
`c
`
`Received 25 September 2001; received in revised form 18 January 2002; accepted 23 January 2002
`
`Abstract
`
`The objective of this study was a comparative investigation of the influence of concomitant food intake on the bioavailability of two
`(cid:210) (cid:210)
`nifedipine-containing controlled-release formulations. Adalat OROS and CORAL were compared in a randomised, non-blind, four-way
`crossover design in 24 healthy, male subjects after single dose administration following a high fat American breakfast or an overnight fast
`of 12 h, respectively. Plasma samples were withdrawn until 48 h post-dose. In the fasted state, the bioavailability (AUC and C
`values)
`max
`(cid:210) (cid:210)
`was lower for CORAL than for Adalat OROS. Under fed conditions, differences in bioavailability between both products were
`markedly increased. With respect to the therapeutic use of both products, the most important finding was the significant dose-dumping
`effect observed after fed administration of CORAL , resulting in nifedipine plasma concentrations of nearly three- to four-fold in 11 of
`(cid:210) (cid:210)
`24 volunteers. The mean ratio of C
`was 235% comparing CORAL with Adalat OROS under these conditions. The formulation-
`max
`dependent food interaction observed in this study may be therapeutically relevant, especially in the case of changing administration
`conditions or switching from one product to the other.
`2002 Published by Elsevier Science B.V.
`
`Keywords: Nifedipine; Monolithic erosive tablet; Osmotic push–pull system; Food effect; Dose-dumping; Bioavailability
`
`1. Introduction
`
`Modified release dosage forms are normally developed
`in order to reduce the dosing frequency for better therapeu-
`tic compliance in chronic treatment and/or to reduce
`maximum peak plasma levels in the case of concentration-
`related side-effects. On the other hand, such products bear
`the risk of
`formulation-related interactions during the
`absorption process, especially in the case of concomitant
`food intake (Blume et al., 1996). Furthermore, food effects
`are normally not predictable from the in-vitro characterisa-
`tion of dosage forms, although a higher probability is
`observed for
`formulations with pH-dependent
`release
`properties. The general conclusion from numerous previ-
`ous investigations is that food effects not only depend on
`the physicochemical properties of the drug, but often result
`
`*Corresponding author. Tel.: 149-6171-5857-11;
`5857-25.
`E-mail address: barbara.schug@socratec-pharma.de (B.S. Schug).
`
`fax: 149-6171-
`
`from formulation characteristics (Karim et al., 1985a,b;
`Waldman and Morganroth, 1995). This is why current
`international regulatory guidelines request food interaction
`studies for the approval of newly developed modified
`release products as well as for generic developments
`(CPMP Note for Guidance, 1999; FDA, 1997).
`Osmotically driven gastrointestinal therapeutic systems
`(GITS) were identified as being very robust
`towards
`potential food interactions (Modi et al., 2000). Such a
`system was developed for nifedipine (Adalat OROS) in
`order to allow once-daily administration instead of the
`twice-daily dosage regimen required with conventional
`modified release tablets.
`Nifedipine is a dihydropyridine with a molecular weight
`of 346.3 and a pK value of .13, and is practically
`a
`insoluble in water. Such physicochemical properties of a
`drug may complicate the development of modified release
`dosage forms. However, other modified release nifedipine
`formulations for once-daily administration based on di-
`verging galenic principles have been approved as generic
`
`0928-0987/02/$ – see front matter
`PII: S0928-0987( 02 )00008-8
`
`2002 Published by Elsevier Science B.V.
`
`Page 1
`
`SHIRE EX. 2036
`KVK v. SHIRE
`IPR2018-00290
`
`(cid:210)
`(cid:211)
`(cid:210)
`(cid:211)
`
`
`280
`
`B.S. Schug et al. / European Journal of Pharmaceutical Sciences 15(2002)279–285
`
`forms by health authorities in the European Union. For one
`of these products, Slofedipine XL, a pronounced food
`interaction was detected, which resulted in significant lag-
`times with absorption-free time spans of more than 15 h in
`the majority of volunteers after fed administration (Schug
`et al., 2001). This investigation showed that, despite the
`harmonised registration requirements in the European
`Union, quality differences with a potential
`impact on
`efficacy and safety still exist.
`The aim of
`this study was to compare the bio-
`availabilities of two 60 mg oral nifedipine modified release
`formulations, CORAL
`(D.R. Drug Research S.R.L.,
`Milano, Italy), a generic erosive tablet for once-daily
`administration, and Adalat OROS (Bayer AG, Lever-
`kusen, Germany), both marketed in member states of the
`European Union, and to investigate the impact of concomi-
`tant
`food intake on the in-vivo performance of both
`products.
`
`2. Methods
`
`2.1. Clinical study
`
`The study was performed in accordance with ICH-GCP
`requirements and the current version of the Declaration of
`Helsinki.
`The investigations followed a randomised, non-blind,
`four-period changeover design in 24 healthy, male subjects
`with washout periods of at
`least 1 week between the
`treatment periods.
`Pre-examination of the subjects included assessment of
`general health status by anamnesis and a physical examina-
`tion, blood pressure and pulse rate measurements, a 12-
`lead ECG, haematological and clinical chemical parame-
`ters as well as urinalysis. Inclusion and exclusion criteria
`were chosen to ensure the safety of the volunteers and to
`exclude pathological factors which might have an influence
`on the bioavailability of the products. Alcohol and drug
`tests were performed prior to each dosing.
`Volunteers were hospitalised for 12 h prior to and for 48
`h after dosing. Every subject received single oral doses of
`(cid:210) (cid:210)
`CORAL (test) or of Adalat OROS (reference) given
`under standardised conditions together with 150 ml non-
`carbonated water either after an overnight fast of at least
`12 h and immediately after eating a high-fat breakfast.
`The subjects remained in the supine position for another
`4 h. Standardised meals were served 4, 7 and 11 h
`post-dose. Conditions were chosen in accordance with
`international
`requirements for
`food interaction studies
`(Draft Guidance for Industry, FDA, 1997). Blood samples
`for the analysis of nifedipine concentrations were collected
`at 0, 0.5, 1, 1.5, 2, 3, 4, 5, 6, 8, 10, 12, 15, 24, 30, 36 and
`48 h after administration. Plasma was prepared under
`protection from daylight due to the photo-instability of
`nifedipine, deep frozen and stored below 220 8C until
`
`analysis. For safety evaluation, vital signs, ECG and
`laboratory parameters were repeatedly determined during
`the hospitalisation phase. Subjective well-being was sur-
`veyed by actively requesting for adverse events in a non-
`leading manner and by documentation of spontaneous
`reporting. Adverse events as reported by the volunteers
`were classified according to severity and potential relation
`to the study medication. Any concomitant medication
`within the course of the study was documented.
`Only healthy male, Caucasian subjects who had given
`their written consent were enrolled in the study. A total of
`27 subjects entered the study. All 27 subjects included in
`the study were investigated for safety analysis. The mean
`age of the 27 subjects was 28.0 years (range 19–39 years),
`mean weight 78.3 kg (range 61–100 kg) and mean height
`181.9 cm (range 167–202 cm). Average BMI was calcu-
`2
`2
`lated as 23.61 kg /m (range 19.5–26.9 kg/m ). Three of
`the subjects dropped out and were replaced: one subject
`dropped out from the study during the first period and
`refused any post-study examination for personal reasons;
`another volunteer withdrew due to severe headache and in
`the third case the volunteer was withdrawn after an adverse
`event had occurred which was classified as not related to
`the study medication. Thus, a total of 24 subjects com-
`pleted all four treatment periods of the study and were
`used for pharmacokinetic analysis.
`
`2.2. In vitro dissolution
`
`The investigational products, each containing 60 mg
`nifedipine, were tested prior to the clinical study with
`identical dissolution conditions in order to allow com-
`parability. After method optimisation, dissolution was
`performed with a standardised compendial Paddle ap-
`paratus with a rotation speed of 50 rpm (n56 for each
`value) using different aqueous buffers containing 1%
`sodium dodecyl sulfate (SDS) in order to achieve sink
`conditions (0.1 N HCl, acetate buffer pH 4.5, phosphate
`buffer pH 6.8 and phosphate buffer pH 8.0, 900 ml each).
`All investigations were performed under complete protec-
`tion from daylight.
`
`2.3. Bioanalytical procedure
`
`Plasma samples were assayed using a LC–MS/ MS
`method operating in the ESI
`(1) mode with MRM
`validated according to international requirements (Shah et
`al., 1992). Amlodipine was used as internal standard. The
`calibration curve obtained after linear regression ranged
`from 0.1 to 100.16 mg/l.
`Quality control (QC) samples were analysed together
`with the study samples. Mean day-to-day precision values
`of the assay procedure as calculated from QC results were
`7.69% (0.13 mg/l), 5.51% (6.43 mg /l) and 4.80% (79.09
`mg/l). Accuracy was determined as a mean deviation of
`0.27% (0.13 mg/l), 21.21% (6.43 mg/l), and 21.51%
`
`Page 2
`
`(cid:210)
`(cid:210)
`
`
`B.S. Schug et al. / European Journal of Pharmaceutical Sciences 15(2002)279–285
`
`281
`
`(79.09 mg/l), respectively. All analyses were performed
`under total protection from daylight.
`
`2.4. Pharmacokinetic evaluation
`
`The pharmacokinetic parameters of nifedipine were
`determined model-independently for each subject by use of
`the WinNonlin software (version 3.0) considering the
`actual sampling times. All data analyses were performed
`with SAS
`for Windows 95/NT (version 6.12, SAS
`Institute, Cary, NC, USA). C
`and t
`were read
`max
`max
`directly from the observed concentration–time points.
`Areas under the plasma concentration vs.
`time curves,
`AUC(0–t ), were calculated according to the linear
`n
`trapezoidal rule during the absorption phase and according
`to the logarithmic trapezoidal rule in the terminal phase
`until
`the time of
`the last quantifiable concentration.
`AUC(0–‘) was calculated as the sum of AUC(0–t ) and
`n
`AUC extrapolated from the last measured value to infinity
`considering the terminal elimination rate constant. Further-
`more, AUC for the intended dosing interval of 24 h
`[AUC(0–24)] was also calculated. Lag-time,
`, was
`t
`lag
`determined directly from the observed concentrations as
`the actual blood sampling time corresponding to the last
`sample with nifedipine concentrations below the LOQ after
`dosing and prior to the first quantifiable sample. Half-value
`duration (HVD) was calculated as the time that the plasma
`nifedipine concentration levels remained above 50% of the
`observed maximum concentration. Mean residence time
`(MRT), the average time a molecule remains in the body,
`was calculated as the ratio of AUMC-to-AUC(0–‘),
`where AUMC is the total area under the first moment
`curve from time zero to infinity.
`were
`The pharmacokinetic parameters AUC and C
`max
`assumed to be log-normally distributed. Log-transformed
`values of these pharmacokinetic characteristics were sub-
`mitted to separate analyses of variance (ANOVA) consider-
`ing sequence, subject (sequence), period, food, formula-
`tion, and food*formulation effects. Based on these analy-
`ses, point estimates for the ratio ‘test/ reference’ following
`both fed and fasting conditions were calculated by re-
`transformation of the logarithmic data. The corresponding
`95% confidence intervals were derived for further explorat-
`ory statistical assessment of differences using the within-
`subject variability from the ANOVA; occasionally, 90%
`confidence intervals were used for further investigation of
`the presence of a food interaction.
`
`3. Results
`
`time
`Mean and individual plasma concentration vs.
`profiles measured in this study are shown in Figs. 1–5,
`pharmacokinetic results are presented in Table 1 and the
`statistical evaluation is summarised in Table 2.
`The plasma concentration vs.
`time profiles of both
`
`Fig. 1. Mean plasma concentration (6S.D.) vs. time curves of nifedipine
`(cid:210) (cid:210)
`determined after oral administration of Adalat OROS and CORAL
`under fasting conditions and after a high-fat breakfast in 24 healthy young
`volunteers in a four-period changeover design.
`
`after
`determined
`formulations
`administration
`fasted
`showed a difference throughout the investigated 48 h blood
`sampling range (Fig. 1). After a lag time of 0.5 to 1.5 h
`(median 1 h), mean plasma concentrations of Adalat
`OROS increased to a level of about 20 mg/l within the first
`5 h, resulting in a plateau until nearly 24 h p.a., followed
`by a slow and continuous decrease until 48 h p.a. After
`administration of CORAL , the mean nifedipine profile
`increased without any lag time to a concentration of nearly
`15 mg /l after 5 h, also followed by a plateau until about 24
`h and a subsequent constant decrease of the curve.
`The differences between the products after fasted ad-
`ministration as shown by the graphs were reflected in the
`major pharmacokinetic parameters and the statistical analy-
`sis. Mean AUC(0–‘) values of 395.9 mg h/l for CORAL
`and 487.7 mg h/l for Adalat OROS confirm a greater
`
`Fig. 2. Individual plasma concentration vs. time curves of nifedipine
`determined after oral administration of Adalat OROS under fasting
`conditions in 24 healthy young volunteers in a four-period changeover
`design.
`
`Page 3
`
`(cid:210)
`(cid:210)
`(cid:210)
`(cid:210)
`(cid:210)
`(cid:210)
`
`
`282
`
`B.S. Schug et al. / European Journal of Pharmaceutical Sciences 15(2002)279–285
`
`Fig. 5. Individual plasma concentration vs. time curves of nifedipine
`Fig. 3. Individual plasma concentration vs. time curves of nifedipine
`(cid:210) (cid:210)
`determined after oral administration of Adalat OROS after a high-fat
`determined after oral administration of CORAL under fed conditions in
`breakfast in 24 healthy young volunteers in a four-period changeover
`24 healthy young volunteers in a four-period changeover design.
`design.
`
`nifedipine bioavailability for Adalat OROS. AUC(0–t )n
`values were of a comparable magnitude to AUC(0–‘)
`since, in most cases, the extrapolated part of AUC(0–‘)
`was relatively small. For the intended dosage interval,
`AUC(0–24) was calculated as only 272.8 mg h/l for
`(cid:210) (cid:210)
`CORAL , whereas the corresponding value for Adalat
`OROS was 321.4 mg h/l. C
`values (geometric mean)
`max
`(cid:210) (cid:210)
`were 20.3 mg/l for CORAL and 23.2 mg/l for Adalat
`OROS. Since the shape of both profiles is similar, these
`differences clearly represent deviations in the extent of
`bioavailability and not in rate.
`Statistical evaluation indicated lower values for Cmax
`when comparing CORAL with Adalat after fasted ad-
`ministration (point estimate: 87%, 95% CI: 74–103%), as
`well as for AUC(0–‘) (point estimate: 81%, 95% CI:
`67–99%). Evaluation of AUC(0–24) showed a similar
`result (point estimate: 85%, 95% CI: 72–100%).
`
`Striking differences between the formulations were
`observed after administration under fed conditions. The
`mean plasma concentration vs.
`time curve of Adalat
`OROS increased (after a lag time of 0.5 to 1.5 h) to
`slightly higher levels (about 24 mg /l) compared with
`fasting conditions, resulting in a plateau until nearly 24 h
`p.a. followed by a slow and continuous decrease until 48 h
`p.a. In contrast, the mean curve of CORAL showed an
`enormous increase up to a maximum of about 55 mg/l
`after 5 h. This maximum was followed by a steep decrease
`until 15 h post-application. Mean plasma levels were
`clearly below those of Adalat OROS beyond 15 h p.a.
`Again, these differences in the shape of the curve were
`in accordance with the data obtained from pharmacokinetic
`evaluation. Total AUC(0–‘) (567.6 mgh/l for CORAL ,
`502.4 mgh/l for Adalat OROS) differed by approximate-
`ly 10%, indicating a slightly higher extent of bioavail-
`ability after fed administration of CORAL . Thus,
`the
`relation of the extent of bioavailability between the prod-
`ucts is reversed when changing from fasted to fed adminis-
`tration. Extrapolated parts of AUC(0–‘) are small, thus no
`relevant differences are observed when comparing total
`AUC(0–‘) with AUC(0–t ). On the other hand,
`the
`n
`difference between C
`values becomes striking under fed
`max
`conditions: geometric means were calculated as 64.2 mg/l
`(cid:210) (cid:210)
`for CORAL and 27.4 mg /l for Adalat OROS. Under
`these conditions,
`the differences in C
`clearly reflect
`max
`product-related discrepancies in drug release. Obviously,
`the release controlling system of CORAL is switched off
`when co-administered with food.
`In order to describe the specific modified release charac-
`teristics, MRT and HVD were calculated. MRT values
`were found to be comparable for both dosage forms under
`fasting conditions. Differences in the mean values of HVD
`reflect the course of the different profiles after administra-
`Fig. 4. Individual plasma concentration vs. time curves of nifedipine
`(cid:210) (cid:210)(cid:210)
`tion of CORAL compared with Adalat OROS. The
`determined after oral administration of CORAL under fasting conditions
`MRT and HVD values underline the assumption of a lack
`in 24 healthy young volunteers in a four-period changeover design.
`
`Page 4
`
`(cid:210)
`(cid:210)
`(cid:210)
`(cid:210)
`(cid:210)
`(cid:210)
`(cid:210)
`(cid:210)
`(cid:210)
`
`
`B.S. Schug et al. / European Journal of Pharmaceutical Sciences 15(2002)279–285
`
`283
`
`Table 1
`Geometric mean values/ geometric S.D. (range) of nifedipine pharmacokinetic parameters in plasma following a single oral dose of 60 mg Adalat OROS
`or 60 mg CORAL in the fasted and fed state (all subjects valid for PK and safety, N 5 24)
`
`Parameter
`
`Adalat OROS
`a
`AUC(0–‘)
`AUC(0–t )
`n
`AUC(0–24)
`a
`AUC(t –‘)
`n
`C
`t
`1 / 2
`aMRT
`HVD
`b
`t
`t
`
`max
`b
`
`lag
`
`max
`a
`
`Unit
`
`mg h/ l
`mg h/ l
`mg h/ l
`%
`mg/ l
`h
`h
`h
`h
`h
`
`(cid:210)
`
`a
`
`max
`a
`
`max
`b
`
`CORAL
`AUC(0–‘)
`AUC(0–t )
`n
`AUC(0–24)
`a
`AUC(t –‘)
`n
`C
`t
`1 / 2
`aMRT
`HVD
`b
`t
`t
`
`mg h/ l
`mg h/ l
`mg h/ l
`%
`mg/ l
`h
`h
`h
`h
`h
`
`lag
`a Only N 5 22 observations available for the fasted state.
`b Median (range).
`
`Fasted state
`
`Fed state
`
`487.7/1.88 (148.4–1172.1)
`475.7/1.83 (147.7–1155.7)
`321.3/1.68 (128.47–776.6)
`1.22/2.94 (0.22–11.05)
`23.2/1.63 (10.3–59.9)
`4.90/1.44 (2.58–9.69)
`19.1/1.25 (12.3–29.1)
`21.2/1.49 (9.08–36.5)
`9 (5–36)
`1 (0.5–1.5)
`
`395.9/2.05 (43.0–1063.6)
`376.8/1.95 (40.9–998.4)
`272.7/1.82 (40.9–678.2)
`1.90/3.70 (0.24–23.12)
`20.3/1.76 (5.58–54.5)
`6.36/1.63 (2.67–18.4)
`18.0/1.37 (8.09–35.6)
`14.5/2.16 (0.96–31.3)
`8 (1–24)
`0 (0–0)
`
`502.4/ 1.84 (135.4–1269.8)
`495.4/ 1.81 (134.7–1155.74)
`357.3/ 1.75 (99.6–915.2)
`0.93/ 4.13 (0.14–14.2)
`27.4/ 1.61 (10.3–61.8)
`4.94/ 1.52 (2.32–12.9)
`17.9/ 1.30 (9.83–27.7)
`17.3/ 1.58 (6.27–34.0)
`9 (5–30)
`1.5 (0.5–3)
`
`567.6/ 1.65 (235.8–1359.2)
`562.8/ 1.65 (232.8–1339.2)
`506.9/ 1.66 (202.3–1192.5)
`0.59/ 2.35 (0.13–3.56)
`64.2/ 1.79 (21.0–165.0)
`5.63/ 1.31 (3.08–8.77)
`11.8/ 1.22 (7.61–17.27)
`5.18/ 1.55 (2.20–17.6)
`5 (3–10)
`0 (0–0.5)
`
`max
`
`of robustness of the generic formulation under fed con-
`increased with food; the mean fed vs. fasted ratio was
`118% (90% CI: 103–135%). The mean ratio for AUC(0–
`ditions. MRT values of CORAL showed a decrease from
`24) for this within-product comparison was calculated as
`18.0 h under fasted to 11.8 h under fed conditions, while
`111% (90% CI: 97–128%).
`HVD values were reduced from 14.5 to 5.18 h.
`In contrast, pronounced differences in terms of the
`The within-product comparison of bioavailabilities after
`AUC(0–‘) and C
`values were determined in the case
`fed vs. fasted administration did not indicate a food effect
`max
`(cid:210) (cid:210)
`of CORAL under fed vs. fasting conditions. The mean
`in the case of Adalat OROS. The mean fed vs. fasted
`AUC ratio was calculated as 106% (90% CI: 90–124%),
`fed vs. fasted ratio for AUC(0–‘) was calculated as 144%
`which is within the generally used acceptance criteria for
`(95% CI: 120–174%), and 317% (95% CI: 270–372%) for
`bioequivalence (80–125%). C
`was found to be slightly
`C . Accordingly, AUC(0–24) underlined the tremendous
`max
`difference in the in vivo performance of the dosage form
`when switching from fasted to fed conditions, resulting in
`a mean ratio of 186% (95% CI: 157–220%).
`The safety evaluation showed that, in total, 21 out of 27
`subjects experienced 100 adverse events, and 84% of the
`events were reported to be at least possibly related to the
`study medication. The events comprised headache, back
`pain, substernal chest pain, asthenia, fever, tachycardia,
`phlebitis, syncope, thrombophlebitis, tooth pain, dyspepsia,
`gastro-enteritis,
`rhinitis, nausea, vomiting,
`increase in
`creatine phosphokinase, agitation,
`insomnia, pharyngitis,
`eczema, herpes simplex and ear pain. The type of adverse
`events observed here generally meet the expectations for a
`study under hospitalised conditions with a vasoactive drug
`compound, i.e. the predominant type of adverse events can
`be attributed either to the study conditions or can alter-
`natively be explained by the vasodilatating effect of
`nifedipine. Nineteen out of 27 subjects (70%) suffered
`from headache,
`i.e. 72% of all adverse events were
`
`Table 2
`Statistical evaluation of the pharmacokinetic parameters for the com-
`(cid:210) (cid:210)
`parison of Adalat OROS and CORAL under fasting conditions and
`after a high-fat breakfast with calculation of point estimates and affiliated
`confidence intervals
`
`Parameter
`
`Comparison
`
`Mean
`ratio
`
`95% CI
`
`AUC(0–‘)
`
`AUC(0–24)
`
`C
`
`max
`
`(cid:210) (cid:210)
`Fasted: CORAL /Adalat OROS
`(cid:210) (cid:210)
`Fed: CORAL /Adalat OROS
`Adalat OROS: fed /fasted
`CORAL : fed/ fasted
`(cid:210) (cid:210)
`Fasted: CORAL /Adalat OROS
`(cid:210) (cid:210)
`Fed: CORAL /Adalat OROS
`Adalat OROS: fed /fasted
`CORAL : fed/ fasted
`(cid:210) (cid:210)
`Fasted: CORAL /Adalat OROS
`(cid:210) (cid:210)
`Fed: CORAL /Adalat OROS
`Adalat OROS: fed /fasted
`CORAL : fed/ fasted
`
`0.81
`1.11
`1.06
`1.44
`
`0.85
`1.42
`1.11
`1.86
`
`0.87
`2.35
`1.18
`3.17
`
`(0.67, 0.99)
`(0.93, 1.33)
`(0.87, 1.28)
`(1.20, 1.74)
`
`(0.72, 1.00)
`(1.20, 1.68)
`(0.94, 1.32)
`(1.57, 2.20)
`
`(0.74, 1.03)
`(2.00, 2.76)
`(1.00, 1.38)
`(2.70, 3.72)
`
`Page 5
`
`(cid:210)
`(cid:210)
`(cid:210)
`(cid:210)
`(cid:210)
`(cid:210)
`(cid:210)
`(cid:210)
`(cid:210)
`(cid:210)
`
`
`284
`
`B.S. Schug et al. / European Journal of Pharmaceutical Sciences 15(2002)279–285
`
`Table 3
`Frequency and intensity of headache observed after administration of
`(cid:210) (cid:210)
`Adalat OROS and CORAL
`under fasting conditions and after a
`high-fat breakfast (in total there were 72 episodes of headache in 19
`subjects; in the case of several episodes per subject within one study
`period the worst intensity was primarily chosen; counts in parentheses
`refer to all episodes)
`
`(cid:210) (cid:210)
`Adalat OROS
`CORAL
`
`Fasting
`(n525)
`
`6 (8)
`3 (4)
`1 (1)
`
`10 (13)
`
`Fed
`(n524)
`
`3 (9)
`3 (4)
`2 (3)
`
`8 (16)
`
`Fasting
`(n525)
`
`3 (8)
`(5)
`3 (3)
`
`Fed
`(n525)
`
`6 (14)
`6 (9)
`4 (4)
`
`10 (16)
`
`16 (27)
`
`Mild
`Moderate
`Severe
`
`Total
`
`episodes of headache. The frequency and intensity of
`headache after fed administration of CORAL was higher
`compared with the other
`treatments, which probably
`reflects the higher plasma concentrations of the vasoactive
`study drug nifedipine in this study situation. An overview
`of the frequency and intensity of headache under the
`different
`treatments is given in Table 3. However, as
`expected from the small sample size and the high vari-
`ability of such a safety parameter,
`the correlation of
`intensity of headache with the observed maximum plasma
`concentrations gave no unambiguous relationship.
`There were no serious or unexpected adverse events
`observed during the study. No relevant effects of the
`administered drug products on blood pressure and heart
`rate were found during the hospitalisation phases. More-
`over, none of the changes in clinical laboratory parameters
`determined within the course of the study were considered
`clinically relevant.
`The mean profiles obtained from the in vitro dissolution
`investigations performed with both dosage forms are
`presented in Fig. 6. Adalat OROS proved to be robust
`towards the different pH values of the media under these
`
`conditions: the product displayed comparable dissolution
`profiles with a 50% release after about 13 h and 90 to
`100% after 24 h over the entire pH range investigated. In
`contrast, results for CORAL displayed the slowest release
`in 0.1 N HCl with 50% of the labelled dose released after
`13 h and about 70% after 24 h. A slightly faster release
`was observed in buffer pH 4.6 and 8.0, whereas the fastest
`release was detected in buffer pH 6.8 with 50% of the
`labelled dose dissolved after 1.5 h and about 80% after 4 h,
`demonstrating a lack of robustness of the release mecha-
`nism towards changes in the pH of
`the dissolution
`medium. After 24 h,
`the relative dissolution showed a
`trend towards lower values for CORAL compared with
`Adalat OROS under identical conditions. In addition to
`the data presented here the effect of agitation, osmotic
`pressure and different concentrations and types of surface
`active ingredients (ionic and non-ionic) on in vitro dissolu-
`tion was investigated and suggested that both dosage forms
`are robust towards these parameters (data not shown).
`
`4. Discussion
`
`The objective of this study was to test for product-
`related differences in the oral bioavailability of two once-
`daily modified-release formulations of nifedipine adminis-
`tered in both the fasted and fed state.
`The findings indicate that the in vivo performance of the
`push–pull system (Adalat OROS) is very robust towards
`concomitant
`intake of food, whereas the erosive tablet
`system (CORAL ) is highly sensitive to food intake. As a
`consequence, AUC(0–‘) and, in particular, C
`increased
`max
`significantly when CORAL was administered to fed
`subjects compared with the fasted state. Obviously, under
`fed conditions, drug release from this product occurs much
`faster than under fasting conditions and,
`thus, release
`control by the dosage form is lost if food is co-adminis-
`
`(cid:210) (cid:210)
`Fig. 6. Mean in vitro dissolution profiles of Adalat OROS and CORAL in different dissolution media: 0.1 M HCl, acetate buffer pH 4.5, phosphate
`buffer pH 6.8 and phosphate buffer pH 8.0, 1% SDS, 900 ml each, in a standardised compendial Paddle apparatus with a rotation speed of 50 rpm (n56 for
`each value).
`
`Page 6
`
`(cid:210)
`(cid:210)
`(cid:210)
`(cid:210)
`(cid:210)
`(cid:210)
`(cid:210)
`(cid:210)
`
`
`B.S. Schug et al. / European Journal of Pharmaceutical Sciences 15(2002)279–285
`
`285
`
`tered with the product. Such a strikingly high food effect is
`known as the ‘dose dumping’ phenomenon and has been
`described in the literature especially for modified release
`preparations
`containing theophylline
`(Karim et
`al.,
`1985a,b).
`The high nifedipine plasma concentrations (up to 165
`mg/l) observed after fed administration of CORAL (for
`comparison:
`the highest maximum concentration after
`administration of Adalat OROS was 48 mg/l) are ex-
`pected to be of relevance for the tolerability of the drug
`product. Since the sample size of the study was not
`planned for a safety comparison, a product-related evalua-
`tion of safety data can only be performed on a descriptive
`level. There is no clear correlation between the number and
`severity of drug-related adverse events; the total number of
`subjects suffering from headache after CORAL adminis-
`tered together with food is higher than after fed administra-
`tion of Adalat OROS.
`In order to increase our understanding of the underlying
`biopharmaceutical mechanism responsible for such in vivo
`results, the in vitro dissolution properties of both dosage
`forms were characterised and compared in buffers of
`different pH using various experimental conditions (e.g.
`agitation, osmolality and surfactants). In vitro dissolution
`of nifedipine from Adalat OROS appears to be robust
`against all chosen modifications, whereas CORAL was
`susceptible to pH changes in such a way that release
`occurred faster with increasing pH values of the medium.
`According to current knowledge, a single unit dosage form
`is expected to pass through the empty stomach much faster
`than in the fed state (Davis, 1989). Thus, under fasting
`conditions the product may reach the duodenum within a
`few minutes, where almost neutral pH values can be
`expected. In contrast, food intake results in a longer
`residence of the dosage form in the stomach (up to 12 h)
`and at elevated gastric pH values of between 4 and 6
`(Blume et al., 1996). Thus,
`the pH-dependent
`in vitro
`release may partly contribute to the observed dose-dump-
`ing phenomenon, but does not explain it completely.
`Additional investigations showed no relevant influence of
`agitation, osmotic pressure or type and concentration of
`surfactant on the in vitro dissolution characteristics of
`CORAL . Thus,
`the in vitro dissolution behaviour of
`CORAL cannot completely explain the observed in vivo
`phenomenon, however it gives a pronounced indication for
`a lack of robustness of the dosage form.
`According to current European regulations (CPMP Note
`for Guidance, 1999), assessment of bioequivalence after
`single dose fed and fasted administration is requested for
`an Abbreviated New Drug Application to ensure the
`appropriate efficacy and safety of a generic modified
`release product. Consequently, both products used in this
`investigation, the generic preparation CORAL and the
`reference product Adalat OROS, should exhibit compar-
`able in vivo performance under all experimental conditions
`
`tested in this study. The results presented in this paper do
`not confirm bioequivalence of both formulations, neither
`after fasted nor after fed administration. The pronounced
`differences observed, particularly in the fed state, do not
`support therapeutic equivalence and, thus, interchangeabili-
`ty of both products. Moreover, the observed dose-dumping
`effect might be of therapeutic relevance for the treatment
`of patients with CORAL , especially in cases when
`patients switch from fed to fasting administration or vice
`versa.
`
`Acknowledgements
`
`The authors would like to thank W. Martin, Pharmakin
`GmbH, for bioanalytical support. Furthermore, the authors
`would like to thank S. Krautwald and G. Zenker for
`excellent technical assistance during the conduct of the
`clinical study.
`
`References
`
`Blume, H.H., Schug, B.S., Potthast, H., 1996. Influence of food on the
`bioavailability of controlled/modified release products. In: Kuhlmann,
`J., Weihrauch, T.R. (Eds.), Food–Drug Interactions. Clinical Pharma-
`¨
`cology, Vol. 12. Zuckschwerdt, Munchen, pp. 25–32.
`CPMP Note for Guidance, 1999. CPMP Note for Guidance on Quality of
`Modified Release Oral and Transdermal Dosage Forms (Section II:
`Pharmacokinetic and Clinical Evaluation), CPMP/EWP/280/96.
`Davis, S.S., 1989. Behaviour of single unit and multiple unit dosage
`forms in the gastrointestinal tract. In: Capsugel Symposium Papers:
`London Symposium, pp. 7–10.
`Food and Drug Administration, 1997. Draft Guidance for Industry:
`food-effect bioavailability and bioequivalence studies. US Department
`of Health and Human Services, CDER.
`Karim, A., Burns, T., Janky, D., Hurwitz, A., 1985a. Food-induced
`changes in theophylline absorption from controlled-release formula-
`tions. Part II: Importance of meal composition and dosing time relative
`to meal intake in assessing changes in absorption. Clin. Pharmacol.
`Ther. 38, 642–647.
`Karim, A., Burns, T., Wearly, L., Streicher, J., Palmer, M., 1985b.
`Food-induced changes in theophylline absorption from controlled-
`release formulations. Part
`I: Substantial
`increased and decreased
`absorption with Uniphyl tablets and Theo-Dur Sprinkle. Clin. Phar-
`macol. Ther. 38, 77–83.
`Modi, N.B., Wang, B., Hu, W.T., Gupta, S.K., 2000. Effect of food on the
`pharmacokinetics of osmotic controlled-release methylphenidate HCl
`in healthy subjects. Biopharm. Drug Dispos. 21, 23–31.
`Schug, B.S., Brendel, E., Chantraine, E., Wolf, D., Martin, W., Schall, R.,
`Blume, H.H., 2001. Dosage form related effect of food on the
`pharmacokinetics of nifedipine: pronounced lag-time after a high-fat
`breakfast. Br. J. Clin. Pharmacol. (in press).
`Shah, V.P., Midha, K.K., Dighe, S., McGilveray, I.J., Skelly, J.P., Yacobi,
`A., Layloff, T., Viswanathan, C.T., Cook, C.E., McDowall, R.D.,
`Pitt