`
`Research Article
`
`Investigation of the Gastrointestinal Transit and In Vivo Drug
`Release of Isosorbide-5-Nitrate Pellets
`
`Wilfried Fischer, I,2 Anna Boertz,' Stanley S. Davis,1,3 Raj Khosla,',3 Willi Cawello,1
`Klaus Sandrock,' and Giinther Cordes'
`
`Received November 4, 1986; accepted July 22, 1987
`
`An oral formulation of controlled-release isosorbide-5-nitrate pellets has been used to investigate the
`location of pellets in the gastrointestinal (GI) tract and, in parallel, to measure the drug absorption
`from these locations. Using the method of gamma scintigraphy the transit times and spreading of
`pellets in the GI tract have been determined. The method of numeric deconvolution was applied to
`calculate the drug input into the systemic circulation. The results indicate that a well-absorbed sub-
`stance such as isosorbide-5-nitrate is absorbed from the stomach and small intestinal in a manner that
`is controlled by the properties of the pellets. Drug absorption is reduced in the colon. The average
`transit time from mouth to colon is 6 to 8 hr, which represents the maximum acceptable time for drug
`release for this oral controlled-release preparation. Taking into account these relations an isosorbide-
`5-nitrate pellet formulation with a bioavailability of 84% has been developed that maintained the min-
`imal therapeutic plasma level for more than 16 hr after application.
`KEY WORDS: controlled-release pellets formulation; gamma scintigraphy; gastrointestinal transit
`time; isosorbide-5-nitrate; absorption rate; numerical deconvolution.
`
`INTRODUCTION
`
`One aim in developing a controlled-release drug formu-
`lation is to reduce the frequency of drug administration. This
`can be achieved by maintaining the therapeutic plasma con-
`centration for an extended period of time. The once-daily
`administration of a drug formulation resulting in a constant
`plasma level is often thought to be optimal. However,
`problems can occur for the individual case where tolerance
`effects preclude a constant plasma level over a period of 24
`hr. Furthermore, a circadian rhythm of the plasma levels is
`required with some active substances (organic nitrates, cor-
`ticosteroids), which should be achieved by the release pro-
`file of the drug.
`A drug that requires defined plasma-level fluctuation in
`order to avoid tolerance phenomena is isosorbide-5-nitrate
`(IS-5-N). On the other hand, a controlled-release prepara-
`tion for a once-daily application is useful, and therefore a
`prolonged drug release must combine an optimal time of
`drug release that assures a good bioavailability with a re-
`lease profile that provides a tolerance-avoiding plasma-level
`profile. This study defines such a time course of drug release
`from IS-5-N formulations without a significant loss of bio-
`availability.
`Sustained-release IS-5-N shows varying bioavailabil-
`
`ities according to the extent of retardation of release. For
`example, in a study conducted by Simbec (1), a depot for-
`mulation providing sustained release in vitro over 14 hr
`showed a relative bioavailability of approximately 60%. If
`the in vitro period of retardation is reduced to 8 hr, the bio-
`availability rises to 84% (2). IS-5-N is reported to be well
`absorbed (up to 100%) from the upper gastrointestinal tract
`and is not affected by first-pass metabolism (3). Therefore,
`reduction in bioavailability of a controlled-release form
`could result from a decrease in absorption or metabolism in
`the lower intestinal regions (colon).
`In order to investigate how long the release of IS-5-N
`can be retarded without a loss of bioavailability (i.e., the
`area of gastrointestinal tract in which IS-5-N is absorbed),
`the gastrointestinal transit and in vivo dissolution/absorption
`of IS-5-N from Elantan long pellets (Pharma Schwarz,
`GMBH) were studied. One capsule of Elantan long contains
`50 mg of IS-5-N as a sustained-release formulation. The dis-
`tribution and residence time of radiolabeled Elanton long
`pellets in the gastrointestinal tract were determined by
`gamma scintigraphy. The analysis of plasma samples drawn
`at the same times as the scintigraphic images allowed calcu-
`lation of the time-dependent input of the IS-5-N in correla-
`tion with the locations of the pellets.
`
`MATERIALS AND METHODS
`
`' Schwarz GMBH, Monheim, West Germany.
`2 To whom correspondence should be addressed at Pharmaceutical
`Development Department, Schwarz GMBH, Mittelstr. 11-13,
`D-4019 Mannheim, West Germany.
`3 Department of Pharmacy, University of Nottingham, England.
`
`Manufacture of 111In-Labeled Controlled-Release Pellets
`
`Sugar beads (0.6-0.71 mm in diameter) were coated
`with a solution of "'In-DTPA complex in a specially de-
`signed small-scale Wurster apparatus. One milliliter of
`
`0724-8741/87/1200-0480805.00/0 CO 1987 Plenum Publishing Corporation
`
`480
`
`Page 1
`
`SHIRE EX. 2033
`KVK v. SHIRE
`IPR2018-00290
`
`
`
`Controlled-Release Isosorbide-5-Nitrate Pellets
`
`481
`
`"lInC1 solution (0.05 M HC1; radioactive concentration, 15.0
`mCi/m1; radioactive purity, 99.98%; New England Nuclear,
`D-6072 Dreiech) was mixed with 5µl of DTPA solution (100
`DTPA; Merck, D-6100 Darmstadt). This solution was di-
`luted with 6 ml ethanol, containing 3.125 mg ethylcellulose,
`and was sealed onto the beads with a water-impermeable
`lacquer consisting of ethylcellulose, polyvinylacetate, and
`talc (leakage rate, 1.1% at 1 hr and 25.5% at 4 hr; water,
`37°C; paddle apparatus). Onto this layer was added 70% of
`the drug substance IS-5-N, the release controlling mem-
`brane, and the loading dose of 30% IS-5-N using the Wurster
`equipment.
`
`Methods
`
`The assay of IS-5-N was by a reversed-phase high-per-
`formance liquid chromatography method (eluant, 30/70
`methanol/water; column, Zorbax C18 (DuPont); flow rate,
`1.5 ml/min; temperature, 40°C; UV detection, 220 nm; RT,
`==3 min).
`Two dissolution methods were used: the USP XXI
`paddle method and the NF XIV rotating bottle method,
`modified.
`The dissolution media were (i) artificial gastric juice,
`pH 1.2; (ii) phosphate buffer, pH 5.5; (iii) phosphate buffer,
`pH 7.5; (iv) demineralized water; (v) a solution of 0.05%
`polysorbate 80 in water (Atlas Chemie, D-4300 Essen; sur-
`face tension, 40 mNm-1); and (vi) an aqueous solution of 2%
`methylcellulose (Dow Chemicals, D-6000 Frankfurt; vis-
`cosity, 42 mPa • sec). The substances were used as supplied
`by the manufactures, without further purification.
`
`Description of Volunteers
`
`The study was conducted in healthy male volunteers of
`different age, weight, and height; the mean age was 25.5 ±
`7.4 years (range, 18-41 years). The mean weight was 69.3 ±
`9.3 kg (range, 55-88 kg) and the mean height was 177.1 ±
`6.9 cm (range, 165-184.5 cm). The dietary intake was con-
`trolled before and during the investigation. The study was
`approved by the Ethical Committee of Nottingham Univer-
`sity and was conducted in accordance with the Helsinki
`Guidelines for Ethics in Research.
`
`Administration of the Formulation
`
`The composition of the pellets was qualitatively and
`quantitatively in accordance with the commercial product
`Elantan-long with the exception of the radioactive label.
`One capsule, containing radioactive labeled pellets (2 MBq/
`dose) and 50 mg IS-5-N, was administered, together with
`100 ml water, to six male, healthy subjects after a standard
`breakfast comprising 1.5 bread rolls, a piece of salami sau-
`sage, 1 slice of cheese, 1 egg, butter, marmalade, and 2 cups
`of coffee (caloric value, 2300 kJ).
`In the second part of the study (after a washout period
`of 1 week), each of the subjects received an Elantan 20
`tablet (Pharma Schwarz GMBH, D-4019 Monheim).
`
`Gamma Scintigraphy
`
`In the first part of the study the subjects stood in front
`of a gamma camera equipped with a parallel collimator with
`
`a 40-cm field of view. Anterior and posterior views of 60-sec
`duration were taken at suitable times. A piece of adhesive
`tape containing 111InC1 was used as an external marker, fixed
`above the liver. The radioactivity was measured in the
`stomach and the colon regions ( = regions of interest; ROI)
`and was corrected for the decay of radioactivity and back-
`ground counts. The geometric mean of the anterior and pos-
`terior views was calculated as discussed previously (4).
`Plasma samples were drawn simultaneously.
`
`Assay of Plasma Samples
`
`Plasma samples were taken either by repeated veni-
`puncture or with an intravenous cannula. At least 5 ml of
`blood was taken at the following times after the dose: 0 (pre-
`dose), 0.5, 1, 2, 4, 6, 8, 10, 12, and 24 hr. Each blood sample
`was centrifuged and the supernatant plasma was pipetted off
`into glass test tubes. Plasma samples were stored at -20°C
`until assayed. Details of the gas chromatographic assay have
`been described elsewhere (5). The reproducability was
`checked on five identical plasma samples, analyzed as de-
`scribed above, each spiked with known amounts of IS-5-N.
`According to these results the precision and accuracy of the
`method were judged as good (see Table II).
`
`In Vivo Dissolution/Absorption of IS-5-N
`
`The in vivo release profile for each subject was deter-
`mined with the use of the point-area numerical deconvolu-
`tion method (6).
`The kinetic parameters, volume of distribution, and
`half-life of IS-5-N for each subject were determined from the
`second part of the study using Elantan 20. These param-
`eters, after computer fitting of the data using the Nonlin pro-
`gram (7), were used as basal data for the deconvolution
`method.
`
`RESULTS
`
`In Vitro Dissolution of IS-5-N from Elantan long Pellets
`
`Scanning electron micrographs of the controlling mem-
`brane of the small batch and those of a large-scale produc-
`tion batch prepared using a coating pan showed excellent
`uniformity. The profiles of the in vitro drug release for the
`
`Table I. Influence of Testing Conditions on the in Vitro Drug Re-
`lease from IS-5-N Pellets°
`
`Conditions of drug release
`
`Model
`
`Medium
`
`Ka (%/hr) ± SE
`
`Rotating bottle
`
`USP XXI
`paddle apparatus
`
`pH 1.2
`pH 5.5
`Water
`cr = 72 mN/m
`= 42 mPa • sec
`
`Water
`13 rpm
`90 rpm
`
`8.11 ± 0.06
`8.6 ± 0.08
`
`9.6 ± 0.10
`9.67 -± 0.04
`
`9.78 ± 0.04
`9.68 ± 0.05
`
`Constant of zero-order release; if, surface tension;
`
`viscosity.
`
`Page 2
`
`
`
`482
`
`Fischer, Boertz, Davis, Khosla, Cawello, Sandrock, and Cordes
`
`Table II. Precision and Accuracy of the IS-5-N Plasma Assay by
`Gas Chromatography
`
`10
`
`10
`
`20
`
`00
`10
`a (mNim
`
`BO
`
`rpm
`
`pH units
`
`c)
`d)
`Fig. 2. Comparison of the in vitro drug release of IS-5-N de-
`pending on different dissolution models. k°, zero-order release
`constant. (a) Rotating bottle, influence of surface tension. (b)
`Rotating bottle, influence of viscosity. (c) Rotating bottle, influ-
`ence of pH variation. (d) Paddle apparatus, influence of hydro-
`dynamics.
`
`and arrival of the pellets in the colon are given for each sub-
`ject in Table III.
`Emptying of the pellets from the stomach and their ap-
`pearance in the colon are expressed by the time at which
`50% (T50%) of the radioactivity is measurable in the ROI. The
`time for small intestine transit is calculated from the differ-
`ence in these times as discussed previously (4).
`The data for the gastric emptying of pellets obtained in
`the present study can be discussed in relation to literature
`data for the emptying of pellets from a fasted stomach (8,9)
`or following a light breakfast taken 1 hr before application.
`Here it was observed that pellets left a fed stomach more
`slowly and steadily than an empty stomach and that the gra-
`dient of the emptying curve gave an indication of the
`spreading of the pellets (6). After intake of the pellets on an
`empty stomach typical emptying times of T50% from 50 to 80
`min were found. Furthermore, the pellets could leave the
`fasted stomach as a bolus, with little or no spreading in the
`small intestine (8). In contrast, pellets administered to a fed
`stomach had gastric emptying times of 119 and 285 min, de-
`pending on the size of the breakfast administered (light
`versus heavy) (4,10). The mean value of 188 min found in the
`present work is between these two values.
`The small intestine transit time of 222 min in the present
`study corresponds well with the mean value of 204-225
`minutes reported by Davis et al. (4,11,12). The leakage of
`the label from the pellets at longer time periods may con-
`tribute to an overestimation of the spreading of the pellets
`system in the terminal ileum and colon. However, such
`leakage should not have a pronounced effect on the calcu-
`lated small intestine transit time and colon arrival time since
`Davis et al. (12) have shown that the intestinal transits of
`pellets and solutions are very similar.
`
`Amount of
`IS-5-N added
`(ng/ml)
`
`Mean
`amount found
`
`No. of
`determinations
`
`50
`100
`400
`
`51.4
`96.0
`395.0
`
`5
`5
`5
`
`Coefficient
`of variation
`(%)
`
`8.6
`4.0
`2.9
`
`small-scale and commercial products were not significantly
`different as measured by the USP XXI paddle method
`(Fig. 1).
`It had been shown previously with the commercial
`product that the drug release was nearly independent of the
`pH, viscosity surface tension, and hydrodynamics of the
`dissolution media. The radiolabeled product was evaluated
`similarly. The release of the retarded IS-S-N dose can be
`described as a zero-order process up to 80% dissolution. It
`can be concluded from Table I that neither pH, viscosity,
`surface tension variation, nor change in the dissolution
`model affected the drug release by more than 10% (Fig. 2).
`The drug release was diffusion controlled and nearly inde-
`pendent of environmental changes. From this it could be
`concluded that the in vivo release profile should be similar to
`that obtained in vitro if the dosage form was rate controlling
`in vivo.
`
`Gamma Scintigraphy
`
`Representative scintiscans showing the gastrointestinal
`transit of labeled pellets are shown in Fig. 3. The pellets
`emptied in a uniform manner from the stomach and spread
`within the small intestine. Evidence of regrouping at the
`ileocecal junction was observed with some subjects. The
`pellets were seen to spread well within the large intestine.
`The mean transit profile is given in Fig. 4. The time for the
`emptying from the stomach, transit in the small intestine,
`
`a—o trade pellets
`- 37 In —labeled pellets
`
`100
`
`90-
`
`80-
`
`70-
`
`60-
`
`50-
`
`40-
`
`30-
`
`20-
`
`10—
`
`% Drug released
`
`6
`
`7
`
`6
`
`0
`
`0
`
`1
`
`2
`
`4
`3
`5
`Time (h)
`Fig. 1. Comparative in vitro drug re-
`lease profile of "'In-labeled pellets and
`Elantan long pellets (USP paddle
`method; 90 rpm; water, 37°C).
`
`Page 3
`
`
`
`Controlled-Release Isosorbide-5-Nitrate Pellets
`
`483
`
`100 c%
`
`80-
`
`60-
`
`40-
`
`20—
`
`% Radioactivity in ROI
`
`o—o GASTRIC EMPTYING
`o—o COLON ARRIVAL
`
`0
`0
`
`100
`
`400
`200
`300
`Time (mins)
`
`500
`
`600
`
`Fig. 4. Gastrointestinal transit of Elantan long
`pellets following a standard breakfast (mean + SE).
`ROI, region of interest. Left curve, stomach; right
`curve, colon.
`
`Absorption Profile
`
`Plasma-level versus time profiles are given in Fig. 5,
`and the in vivo dissolution characteristics calculated there-
`from are given in Table IV. All subjects showed a rapid rise
`in IS-5-N level in the plasma. After 0.5 hr effective levels of
`over 100 ng/ml were achieved which were maintained for
`periods of 16-20 hr.
`The results for Cinax are in the range of 362 to 595 ng/mg,
`and those for Tillax between 4 and 6 hr. The accumulated in
`vivo release/absorption amounts to 75-95% (average,
`84.6%; coefficient of variation, 11.1%), which is also the rel-
`ative bioavailability of Elantan long in comparison to
`Elantan 20 tablets (Fig. 6). The mean residence time (Table
`V) in comparison to Elantan 20 tablets is 2.5 hr longer.
`The differential in vivo IS-5-N dissolution/absorption
`profile indicates a three-step mechanism. During the first
`hour approximately 25% of the applied dose was absorbed.
`This corresponds to the loading dose. After this initial phase
`a nearly constant absorption for up to 6 hr was observed.
`During this time approximately 40% of the dose was ab-
`sorbed. In the final period, which lasts for another 12 hr,
`19% of the dose was absorbed. The combined absorption of
`84% of the applied dose is equivalent to the bioavailability of
`IS-5-N for this preparation.
`It is interesting to examine the correlation between the
`gastrointestinal transit of the pellets and the drug input func-
`tion (Fig. 7). From these data it is clear that the loading dose
`
`Table III. Gastrointestinal Transit Times (T500 (Minutes) of "ItIn-
`Labeled IS-5-N Pellets
`
`Subject
`No.
`
`Emptying
`of stomach
`
`Small intestine
`transit
`
`Arrival in
`the colon
`
`2
`3
`4
`5
`6
`
`SE
`
`195
`205
`120
`90
`210
`305
`188
`31
`
`235
`260
`270
`210
`125
`230
`222
`21
`
`430
`465
`390
`300
`335
`535
`409
`35
`
`Fig. 3. Gamma scintigraphs of isosorbide-
`5-nitrate pellets after application following
`a standard breakfast. (a) Fifteen minutes
`after intake; the hard gelatin capsule is
`broken up and the pellets are dispersed in
`the stomach. (b) Nine hours after intake;
`some of the pellets are dispersed in the as-
`cending colon, and the rest are concen-
`trated before the valva ileocaecalis.
`
`Page 4
`
`
`
`Fischer, Boertz, Davis, Khosla, Cawello, Sandrock, and Cordes
`
`— Differential Input
`o—e Cumulative Input
`
`1.0
`
`0.8-
`
`0.6-
`
`0.4-
`
`0.2-
`
`Fraction released
`
`484
`
`Plasmaconcentration (ng/m1)
`
`0.0
`
`12
`Time (h)
`Fig. 6. In vivo IS-5-N release/absorption (mean + SD) as
`calculated by numerical deconvolution from the plasma-level
`profiles.
`
`16
`
`20
`
`24
`
`Spreading of the pellets in the different regions has been re-
`ported previously.
`The results of the present investigation show that there
`is an optimum period of time of 6 to 7 hr after intake for
`absorption of a well-absorbed substance such as IS-5-N.
`This optimal time corresponds to the residence time of the
`pellets in the stomach and the small intestine, where the ac-
`tive substance is absorbed (rate limited) according to the re-
`lease profile of the pellets.
`As the IS-5-N release from the controlled-release
`pellets is nearly independent of environmental changes, it
`can be concluded that the difference between in vitro release
`and in vivo absorption reflects the reduced absorption of
`IS-5-N in the lower parts of the small intestine and regions
`of the colon. The absorption decreases when the pellets ac-
`cumulate at the ileocecal sphincter and then pass into the
`colon. In the subsequent 12-hr period in the colon, where
`distribution of the pellets occurs, only 10-20% of the ap-
`plied dose was absorbed.
`The results of this study show that a release time of 8 hr
`for IS-5-N from controlled-release preparations is the max-
`imum time to assure nearly complete absorption of the drug.
`With the administration of 50 mg IS-5-N once a day, the
`minimal therapeutic plasma level of 100 ng/ml can be main-
`tained for a period of approximately 18 hr. After this time
`the plasma level falls below 100 ng/ml, thereby avoiding tol-
`erance phenomena (13).
`
`Table V. Mean Residence Times (MRT) of Is-5-N After Application
`of Elantan long Pellets and Tablets
`
`Single application
`Fig. 5. IS-5-N plasma-level profiles of the volunteers after a
`single application of 50 mg IS-5-N as controlled-release
`pellets following a standard breakfast.
`
`and a fraction of the controlled-release dose are released in
`the stomach as expected. The reduction of drug absorption
`that occurs 6 hr after administration can be explained by the
`fact that at this time about 40% of the pellets are located in
`the colon. The remaining 60% are spread within the stomach
`and small intestine. However, it is believed that the majority
`of these pellets is in the lower parts of the small intestine
`near the ileocecal sphincter. Consequently we find a re-
`duced absorption capacity and the measured absorption rate
`falls. During the subsequent 12 hr the pellets spread in the
`colon and a continuous process of absorption takes place.
`
`DISCUSSION
`
`The transit of dosage forms in the small intestine is
`known to be reasonably constant as discussed above; there-
`fore, the variable factor in the passage of pellets from mouth
`to colon is the time required for gastric emptying. This time
`can be specifically influenced by the intake of food. In the
`present study the combined residence time of the pellets in
`the stomach and small intestine region was 6 to 7 hr and the
`pellets were well spread in the jejunum and the ileum. The
`data obtained for the pellets in the colon region indicate that
`the pellets can accumulate at the junction between the small
`intestine and the ascending colon, i.e., at the ileocecal
`sphincter, and then enter the colon as one or a few boluses.
`
`Table IV. Correlation of in Vitro Drug Release with in Vivo Drug
`Release/Absorption of IS-5-N from "In-Labeled Pellets"
`
`Time
`(hr)
`
`In
`vitro
`
`Accumulated release/absorption
`of IS-5-N (%) in vivo in subject
`
`RC
`
`RW SA NR WM X
`
`Coefficient
`of variation
`(%)
`
`Subject
`No.
`
`2
`4
`6
`8
`24
`
`42.8
`61.8
`
`78.8
`100
`
`29
`41
`50
`53
`77
`
`46
`64
`77
`82
`90
`
`27
`46
`58
`63
`85
`
`32
`56
`68
`74
`97
`
`41
`58
`67
`66
`74
`
`35
`43
`64
`67.6
`84.6
`
`3.6
`4.2
`4.6
`4.9
`4.2
`
`" Subject SW was not considered because a value for the distribu-
`tion volume was not available.
`
`1
`2
`3
`4
`5
`6
`
`± SE
`
`MRT (hr)
`
`Elantan long
`
`10.59
`9.54
`9.99
`8.24
`8.29
`10.92
`9.56 ± 1.14
`
`Elantan 20
`
`7.39
`6.87
`7.25
`6.81
`
`7.08 ± 0.28
`
`Page 5
`
`
`
`Controlled-Release Isosorbide-5-Nitrate Pellets
`
`485
`
`2n
`
`z
`
`absorbed drug such as IS-5-N is poorly absorbed when it
`enters the colon. A CR formulation that releases the drug
`over the whole length of the gastrointestinal tract, i.e.,
`stomach to colon, is sensible only if the active substance is
`absorbed from the colon at a rate comparable to that in the
`small intestine. If this requirement is not met, a loss of bio-
`availability can be expected. As a rule this reduced absorp-
`tion will lead to a decreased plasma level, so that a twice-
`daily application of the controlled-release form will be re-
`quired in order to provide a constant plasma level.
`
`100
`
`a
`
`80
`
`60
`
`40
`
`20
`
`% Radioactivity in Rol
`
`100
`
`400
`300
`200
`Time (mins)
`
`500
`
`600
`
`REFERENCES
`
`1. D. E. Davies, E. Evans, and B. Hagerty. Final Report, Simbec
`Research Laboratories (1983).
`2. R. Bonn. Miinch. Med. Wschr. 129:221-222 (1987).
`3. S. G. Wood, B. A. John, L. F Chasseaud, R. M. Major, M. E.
`Forrest, R. Bonn, A. Darragh, and R. F Lambe. Arzneim. For-
`esch.1Drug Res. 34:1031-1035 (1984).
`4. S. S. Davis, J. G. Hardy, M. J. Taylor, D. R. Whalley, and
`C. G. Wilson. Int. J. Pharm. 21:331-340 (1984).
`5. J. Evers, R. Bonn, A. Boerts, W. Cawello, V. Luckow, M. Fey,
`E Abudan, and H.-A. Dickmans. Eur. J. Clin. Pharmacol. 32
`(1987) (in press).
`6. F Langenbucher. Pharm. Ind. 44:1166-1171 (1982).
`7. C. Metzler, G. L. Elfring, and A. E McEven. Biometrics 30:562
`(1974).
`8. E. Hunter, J. T. Fell, and H. Sharma. Drug Dev. Ind. Pharm.
`8:751-757 (1982).
`9. F N. Christensen, S. S. Davis, J. G. Hardy, M. J. Taylor, D. R.
`Whalley, and C. G. Wilson. J. Pharm. Pharmacol. 37:91-95
`(1985).
`10. S. S. Davis, J. G. Hardy, M. J. Taylor, D. R. Whalley, and
`C. G. Wilson, Int. J. Pharm. 21:167-177 (1984).
`11. S. S. Davis. In D. D. Breimer and P. Speiser (eds.), Topics in
`Pharmaceutical Sciences, Elsevier, Amsterdam, 1983, pp.
`205-215.
`12. S. S. Davis, J. G. Hardy, and J. W. Fara. Gut 27:886-892
`(1986).
`13. E. Noack and R. Bonn. In B. W. Milner (ed.), Controlled Drug
`Delivery, Wissenschaftliche Verlagsgesellschaft mbH, Stutt-
`gart, 1987, pp. 29-48.
`
`3
`
`2j
`z
`
`1 w
`
`100
`
`b
`
`80
`
`60
`
`40
`
`20
`
`% Radioactivity in Rol
`
`0
`0
`
`100
`
`200
`400
`300
`Time (mins)
`
`500
`
`600
`
`Fig. 7. (a) Differential drug input function and gastric emp-
`tying of Elantan long pellets (mean + SE). The input fraction
`corresponds to the right ordinate. ROI, region of interest.
`Upper curve, gastric emptying. (b) Differential drug input
`function and colon arrival of Elantan long pellets (mean +
`SE). Right curve, colon arrival.
`
`For the future design of controlled-release (CR) dosage
`forms it can be concluded from this study that even a well-
`
`Page 6
`
`