Aliment Pharmacol Ther 1997 ; 11 : 205–213
`
`Colonic delivery of dexamethasone : a pharmacoscintigraphic
`evaluation
`
`C. J. K E N Y O N, R. V. N A R D I*, D. W O N G*, G. H O O P E R, I. R. W I L D I N G & D. R. F R I E N D*
`Pharmaceutical Profiles Ltd, Highfields Science Park, Nottingham, UK ; and * Cibus Pharmaceutical, Burlingame, California,
`USA
`
`Accepted for publication 29 July 1996
`
`S U M M A R Y
`
`Background : Colonic delivery of corticosteroids may
`reduce the side-effects commonly associated with their
`use. Therefore, we tested the ability of the naturally
`occurring polysaccharide guar gum to deliver a
`corticosteroid, dexamethasone, to the colon using
`pharmacoscintigraphy. Guar gum is metabolized in the
`colon by resident bacterial enzymes to trigger drug
`release.
`Materials : Each subject (eight per group, parallel study
`design) was administered one of four dexamethasone
`(9 mg) tablet formulations, radiolabelled with "&$Sm
`using neutron activation, under fasted conditions. One
`formulation was designed to release drug rapidly
`following ingestion while the other three formulations
`were designed to delay release of dexamethasone to
`varying degrees. Progression of the formulations down
`the gastrointestinal tract was followed by gamma
`scintigraphy. Serum concentrations were measured
`over time to relate disintegration profiles of the tablets
`with pharmacokinetic observations.
`
`$
`
`Results : The immediate release formulation
`disintegrated in the stomach, on average, within 20
`min of dosing. One of the three delayed release
`preparations (CD
`) began to disintegrate in the small
`"
`intestine 1n7p1n0 h after dosing. The second and third
`delayed release preparations (CD
`and CD
`) did not
`#
`begin to disintegrate until 5n8p2n3 and 3n6p1n6 h
`after dosing, respectively. All three colonic delivery
`preparations completely disintegrated in the colon
`ranging from 7n8p2n7 h (CD
`) to 12n4p3n2 h (CD
`)
`"
`#
`following oral administration. Pharmacoscintigraphic
`data indicated that 72–82 % of the dexamethasone was
`delivered into the colon although not all the
`dexamethasone delivered into the colon was absorbed.
`Conclusions : Simple guar gum formulations are capable
`of delivering the corticosteroid dexamethasone to the
`colon of normal subjects. Locally delivered
`corticosteroids may be useful in the treatment of
`ulcerative colitis and Crohn’s disease.
`Pharmacoscintigraphic evaluation is a useful method
`to discriminate between the in vivo behaviour of
`colonic delivery systems.
`
`I N T R O D U C T I O N
`
`Delivery of drugs to the colon is useful in the treatment of
`several colonic diseases (ulcerative colitis and Crohn’s
`disease). Corticosteroids have traditionally formed the
`basis of treating inflammatory bowel disease. However,
`chronic treatment of inflammatory bowel disease with
`
`to : Dr D. R. Friend, Cibus Pharmaceutical
`Correspondence
`Burlingame, CA 94010, USA.
`
`Inc.,
`
`steroids, while often effective, is plagued by a number of
`serious side-effects (e.g. acne, moonface, striae, hyper-
`tension, peptic ulcer, impaired glucose tolerance and
`mood disturbances). Long-term treatment (greater than
`6 months) can lead to osteoporosis, cataracts, necrosis
`and fracture of head and femur, and overt diabetes
`mellitus. If these undesired side-effects could be overcome
`or markedly reduced in both subchronic and chronic
`dosing regimes, corticosteroids would have the potential
`
`# 1997 Blackwell Science Ltd
`
`205
`
`Cosmo Ex 2008-p. 1
`Mylan v Cosmo
`IPR2017-01035
`
`

`

`206
`
`C . J . K E N Y O N e t a l .
`
`of being ideal therapeutic treatments of inflammatory
`bowel disease."
`One approach to improving the use of corticosteroids to
`treat inflammatory bowel disease is through site-specific
`drug delivery. By delivering more drug to the inflamed
`tissues, it is possible to substantially increase the local
`tissue concentrations. Kinetically, local delivery is asses-
`sed using a term called drug delivery index.# By in-
`creasing the drug delivery index, it is possible to deliver
`efficacious amounts of drug from significantly smaller
`doses. As a result of the lower dose, side-effects should be
`reduced relative to traditional approaches to drug adminis-
`tration. This concept has been successfully demonstrated
`in a number of animal studies involving oral administra-
`tion of colonic delivery systems for corticosteroids$, % and
`in humans with delivery via enemas.&
`The general approaches to delivering drugs to the colon
`via the oral route include : (1) enteric coating designed to
`release drug in the more alkaline environment of the
`gastrointestinal
`tract ;'
`(2) bioerodible coatings and
`matrices ;( (3) prodrugs ;) (4) timedkrelease systems ;*
`and (5) sustained release systems that release drug as
`they transit through the small and large intestines."! An
`alternative to oral administration is rectal dosing (sup-
`positories, enemas).""
`A potential matrix material for colonic drug delivery is
`guar gum. Owing to its high viscosity this polysaccharide
`may carry certain drugs to the large intestine without
`appreciable release in the stomach or small intestine.
`Once in the large intestine, the guar gum matrix will be
`degraded by specific enzymes produced by the gut
`microflora (i.e. a-galactosidases and b-mannanases) to
`initiate drug release."#
`In the early stages of product development of novel
`colonic delivery systems, considerable time can be lost in
`establishing the likely potential of any given research
`strategy because of a lack of suitable in vitro or animal
`models. Pilot pharmacokinetic studies in small groups of
`healthy subjects may provide unsuitable end-points for
`the assessment of colonic targeting preparations."$ How-
`ever, pharmacoscintigraphic evaluation provides detail-
`ed information on the in vivo performance of novel oral
`formulations and can be used to focus the product
`development process."%
`The purpose of this study was, therefore, to determine
`the in vivo performance of selected formulations for the
`guar gum colonic drug delivery research programme.
`The investigation was a pharmacoscintigraphic evalu-
`ation of guar gum delivery formulations spanning a
`
`range of release profiles. Disintegration and gastro-
`intestinal transit data determined scintigraphically were
`correlated with the absorption profiles of dexamethasone.
`
`S U B J E C T S A N D M E T H O D S
`
`Subjects
`
`A total of 32 healthy volunteers (18 male and 14 female)
`were enrolled in the study. Each subject underwent a
`medical examination both prior to and following com-
`pletion of the study, during which blood samples for
`haematology and clinical chemistry were taken, and a
`urine sample was provided for urinalysis. The protocol of
`the study was approved by an IRB (Quorn Research
`Review Committee, Leicestershire, UK). Approval to
`administer radiolabelled preparations to healthy volun-
`teers was obtained from the Department of Health,
`London. Prior to recruitment, the nature of the study was
`explained both verbally and in writing to the volunteers,
`and each volunteer provided written consent.
`
`Experimental procedures
`
`This investigation was a double-blind, parallel group
`design in which blocks of eight healthy subjects received
`one of four different treatment regimens. A total of 32
`healthy male or non-pregnant, non-breast feeding female
`subjects participated ; subjects were randomized to one of
`the four different treatment formulations so that each
`subject received only a single dose. The four treatment
`formulations were designed to release dexamethasone to
`varying extents in the gastrointestinal tract. The first
`formulation was an immediate release (IR) tablet while
`the remaining three tablet formulations were designed to
`release drug primarily in the colon and are designated as
`CD
`, CD
`and CD
`. Dexamethasone (U.S.P., micronized)
`"
`#
`$
`was obtained from the Upjohn Co.
`(Kalamazoo,
`Michigan, USA) and the formulations were manu-
`factured by Penn Pharmaceuticals (Tredegar, UK). These
`tablets differed with respect to grade (particle size) of guar
`gum and other excipients used to modulate drug release
`in the gastrointestinal tract. The tablets weighed approxi-
`mately 333 mg and contained 2n7 % dexamethasone (i.e.
`9 mg of dexamethasone per tablet) and 2 mg of sam-
`arium oxide. The tablets were irradiated for 6 min in a
`neutron flux of 10"# n\s.cm#, and in vitro dissolution
`testing demonstrated that neither the addition of the
`samarium oxide nor the neutron activation process
`
`# 1997 Blackwell Science Ltd, Aliment Pharmacol Ther 11, 205–213
`
`Cosmo Ex 2008-p. 2
`Mylan v Cosmo
`IPR2017-01035
`
`

`

`affected the performance of the dosage forms or the
`stability of the drug.
`Anterior and lateral anatomical markers containing
`0n1 MBq **Tcm were taped to the skin over the right lobe
`of the liver. On a single study day, each subject received
`a single formulation radiolabelled with "&$Sm and con-
`taining 9 mg dexamethasone at approximately 08.00 h
`with 240 mL of water. Anterior scintigraphic images
`were recorded at frequent intervals for up to 16 h, using
`a gamma camera (General Electric Maxicamera) with a
`40 cm field of view and fitted with a low-energy parallel
`hole collimator. Images were recorded at approximately
`10-min intervals up to 12 h post-dose and then at
`approximately 30-min intervals until 16 h post-dose.
`Return visits were made to the clinical unit at 24 and 36
`h post-dose to allow the acquisition of further images.
`The volunteers remained moderately active during the
`study period and all images were acquired with the
`subjects standing in front of the gamma camera. The
`images were recorded using a Bartec computer system
`and were stored on optical disk for subsequent analysis.
`Transit and tablet disintegration information related to
`the following parameters were obtained : (a) gastric
`emptying time ; (b) colon arrival time ; (c) small intestinal
`transit time ; (d) anatomical location and time of initial
`tablet disintegration ; and (e) anatomical location and
`time of complete tablet disintegration.
`A standard light
`lunch, dinner and supper were
`provided at 4, 9 and 14 h post-dose, respectively. Each
`subject drank 200 mL of water at 2 h post-dose and fluids
`were allowed ad libitum after lunch. At the end of study
`day 1, subjects were instructed to fast until returning to
`the clinical unit the following morning. Food was only
`allowed ad libitum after the 24-h image and blood sample.
`Venous blood samples were withdrawn via an in-
`travenous cannula or by venipuncture on the following
`time schedule : 0 (pre-dose), 1n0, 2n0, 4n0, 6n0, 8n0, 10n0,
`12n0, 14n0, 16n0, 24n0 and 36n0 h post-dose. The
`samples were left at room temperature for approximately
`30 min until a clot was formed. The samples were then
`centrifuged at approximately 3000 r.p.m. (or 1800 g) for
`7 min at 4 mC. The resulting serum fraction was split into
`two aliquots by pipetting into two pre-labelled poly-
`propylene screw-cap tubes. Samples were flash frozen
`and then stored immediately at k20 mC for radio-
`immunassay (RIA) of dexamethasone concentrations.
`Plasma cortisol levels were also measured by RIA as
`there was some cross-reactivity with cortisol
`in the
`dexamethasone RIA technique.
`
`# 1997 Blackwell Science Ltd, Aliment Pharmacol Ther 11, 205–213
`
`C O L O N I C D E L I V E R Y O F D E X A M E T H A S O N E
`
`207
`
`Analysis of dexamethasone and cortisol
`
`Dexamethasone was measured in serum using a vali-
`dated radioimmunassay method by Phoenix Inter-
`national Life Sciences (St. Laurent, Quebec, Canada).
`Dexamethasone (Reference Standard) for assay devel-
`opment was obtained from Sigma Chemical Co. (St Louis,
`Missouri, USA). Sheep anti-dexamethasone antisera was
`purchased from Guildhay Ltd. (Guildford, Surrey, UK)
`and [1,2,4-$H]-dexamethasone was purchased from
`Amersham Canada Ltd (Oakville, Ontario, Canada).
`Rabbit anti-sheep whole serum was purchased from
`Sigma Chemical Co. and human serum was purchased
`from Scantibodies Laboratory (Santee, California, USA).
`Cortisol
`reference
`standard was purchased from
`American Chemicals Ltd (Montreal, Quebec, Canada).
`Coat-A-Count cortisol RIA kits were purchased from
`Diagnostics Product Corporation, DPC (Los Angeles,
`California, USA) for cross-reactivity evaluation of anti-
`dexamethasone antibody with cortisol.
`The method involved allowing the tritiated dexame-
`thasone and non-radioactive dexamethasone to react
`with the sheep anti-dexamethasone antiserum under
`suitable incubation conditions and for an adequate
`period. Separation of
`free dexamethasone from the
`antibody-bound dexamethasone was achieved by adding
`the second antibody (anti-sheep whole serum) plus
`polyethylene glycol (PEG 8000). The antibody-free frac-
`tion was measured by liquid scintillation counting. The
`concentration of dexamethasone in the unknown sam-
`ples is directly proportional to the interpolated counts of
`the free fraction, obtained from a standard curve con-
`taining known amounts of dexamethasone. The range of
`quantitation of dexamethasone was 202–2504 pg\mL
`with a lower limit of quantitation (LLOQ) of 2020 pg\mL
`and lower limit of detection (LLOD) of 49n5 pg\mL. The
`range of quantitation of cortisol was 10n0–500 ng\mL
`with a LLOQ of 10 ng\mL and a LLOD of 1n2 ng\mL.
`
`Pharmacokinetics
`
`A variety of pharmacokinetic parameters (e.g. area under
`the concentration-time curve (AUC), mean residence
`time (MRT) and mean absorption time (MAT) were
`measured using standard techniques. The amount of
`drug absorbed over time was estimated using a Wager–
`Nelson calculation"& and reference pharmacokinetic para-
`meters."' In order to assess the impact of observed non-
`zero pre-dose concentration values due to cross-reactivity
`
`Cosmo Ex 2008-p. 3
`Mylan v Cosmo
`IPR2017-01035
`
`

`

`208
`
`C . J . K E N Y O N e t a l .
`
`with serum cortisol on the AUCinf, the baseline value was
`subtracted from all time points and the mean corrected
`AUCinf was compared to the mean uncorrected AUCinf.
`The impact of baseline adjustments on the mean AUCinf
`of the uncorrected vs. the corrected concentration values
`for all four formulations was minimal, since this dif-
`ference ranged from 1 to 3 %. Therefore, non-corrected
`data are presented.
`
`Statistical analysis
`
`Analyses of variance were performed on the untrans-
`formed pharmacokinetic parameters listed above, with
`the exception of the ratio AUC
`!−t to AUCinf, elimination
`rate constant
`following intravenous administration
`(kel(iv)), MRT and MAT. The analysis of variance model
`included drug formulation as a factor. A 5 % level of
`significance was used. Each analysis of variance included
`calculation of least-squares means, adjusted differences
`between formulation means and the standard error
`associated with these differences. The above statistical
`analyses were done using the SAS GLM procedure.
`
`R E S U L T S
`
`The tablets were observed to be present in the stomach in
`all 32 individuals in the first scintigraphic image fol-
`lowing dosing (i.e. approximately 1 min post-dose). They
`did not adhere to the oesophageal wall nor subsequently
`obstruct the gut.
`It was noted that a small amount of radioactive marker
`‘ leached ’ from each of the four formulations shortly after
`administration of the preparations. Initially, this marker
`was observed to disperse throughout the gastrointestinal
`tract ; however, it often could not be detected in sub-
`
`Table 1. Transit profiles of the four dexamethasone delivery
`systems evaluated
`
`Gastric
`emptying
`(h post-dose)
`
`Colon
`arrival
`(h post-dose)
`
`Formulation
`
`IR
`CD
`CD
`CD
`
`"
`
`#
`
`$
`
`NA*
`0n9p0n7(cid:139)
`0n6p0n4
`0n9p0n7
`
`NA
`5n2p1n5
`4n2p0n6
`4n9p1n2
`
`Small
`intestinal
`transit
`(h)
`
`NA
`4n3p1n6
`3n6p0n6
`4n1p1n4
`
`* NA l not applicable as there was no discrete transit event due
`to prior disintegration of the tablets.
`(cid:139) Data are meansps.d. (n l 8).
`
`"
`
`sequent images. This material is thought to result from
`progressive erosion of the surface of the tablets due to the
`continual peristaltic action of the gut. For this reason,
`initial disintegration was recorded as the midpoint
`between the two images after which dispersed radioactive
`marker was observed in consecutive images.
`Mean transit data (gastric emptying (GE), colonic arrival
`time and small intestinal transit time (SITT)) for all four
`formulations are summarized in Table 1. Formulation IR
`disintegrated before gastric emptying could be observed
`while the remaining three formulations remained intact
`until reaching the distal small intestine or colon. Table 2
`summarizes the mean disintegration times of all four
`formulations tested.
`Initial disintegration of formulation IR occurred in the
`stomach at 10p17 min (range 1–48 min) post-dose and
`tablets completely disintegrated shortly after adminis-
`tration in six of the eight subjects (25 minp5n16 h ;
`range 1 min–13n8 h (n l 7) post-dose). The result was a
`sharp rise in the serum dexamethasone concentration
`within the first hour after dosing (see Figure 1).
`Formulation CD
`initially disintegrated in the small
`intestine (1n7p1n0 h ; range 0n7–3n8 h ; see Table 2).
`Because initial disintegration occurred in the more
`proximal portion of the intestine, some drug was ab-
`sorbed before reaching the colon (see Figure 1). Colon
`arrival occurred 5n2p1n5 h (range 3n9–7n6 h) post-dose.
`Complete tablet disintegration occurred in the colon at
`7n9p2n7 h (range 5n1p12n8 h ; n l 8) post-dose. Total
`dexamethasone absorption was complete by 16 h post-
`dose as indicated by the Wagner–Nelson plots of percent
`drug absorbed over time. Multiple peaks in the pharma-
`cokinetic profile were observed in subjects 12, 15, 24 and
`30, and the time of the major peak correlated well with
`complete tablet disintegration times in the colon. A single
`extended peak was found in subjects 7 and 27 and
`absorption appeared to be complete before the tablet had
`fully disintegrated.
`There was a large variation for the time of gastric
`emptying among individuals receiving formulation CD
`#
`but the time for colon arrival and the site for initial and
`complete disintegration was consistent. Mean pharma-
`cokinetic data (serum concentrations over time) from
`subjects receiving CD
`are presented in Figure 1. In six of
`#
`the eight subjects who received CD
`, initial disintegration
`#
`of the tablets was observed following colonic arrival
`(5n8p2n3 h (range 2n9–10n5 h) ; see Table 2) post-dose.
`Multiple serum dexamethasone concentration peaks
`were observed in all eight subjects, and significant drug
`
`# 1997 Blackwell Science Ltd, Aliment Pharmacol Ther 11, 205–213
`
`Cosmo Ex 2008-p. 4
`Mylan v Cosmo
`IPR2017-01035
`
`

`

`Table 2. Disintegration profiles of the four
`formulations tested
`
`Formulation
`
`Initial tablet disintegration
`Post-dose (h)
`Post-GE (h)
`
`Complete tablet disintegration
`Post-dose (h)
`Post-GE (h)
`
`C O L O N I C D E L I V E R Y O F D E X A M E T H A S O N E
`
`209
`
`IR
`CD
`"
`CD
`CD
`
`#
`
`$
`
`0n17p0n18* —(cid:139)
`1n7p1n0(cid:140)
`1n2p1n3
`5n8p2n3(cid:142)
`5n2p2n3
`3n6p1n6(cid:139)(cid:139)
`2n7p1n2
`
`2n1p5n2
`7n9p2n7(cid:141)
`12n4p3n2**
`12n2p3n8(cid:140)(cid:140)
`
`2n0p5n0
`6n1p3n2
`7n4p3n6
`8n3p3n9
`
`GE l gastric emptying.
`* Data are meansps.d. (n l 8).
`(cid:139) No discrete emptying of tablet observed due to disintegration in the stomach.
`(cid:140) Tablets disintegrated initially in the stomach (n l 6) or small intestine (n l 2).
`(cid:141) Tablets disintegrated completely primarily in the ascending colon (n l 6), transverse
`colon (n l 1) or the splenic flexure (n l 1).
`(cid:142) Tablets disintegrated initially in the distal small intestine (n l 2), the ascending
`colon (n l 4), the transverse colon (n l 1) or descending colon (n l 1).
`** Tablets disintegrated completely in the ascending colon (n l 1), hepatic flexure
`(n l 2), transverse colon (n l 4) or the descending colon (n l 1).
`(cid:139)(cid:139) Tablets disintegrated initially in the distal small intestine (n l 6), the ileo-caecal
`junction (n l 1) or the ascending colon (n l 1).
`(cid:140)(cid:140) Tablets disintegrated completely in the ascending colon (n l 3), hepatic flexure
`(n l 1) or the splenic flexure (n l 1) ; the position of total disintegration was not
`observed in three of the subjects.
`
`(range 2n3–7n4 h ; in the small intestine in most cases)
`post-dose, with about 23 % of the drug absorbed in the
`small intestine. In seven of the eight subjects receiving
`formulation CD
`, the tablets began to disintegrate in the
`$
`small intestine, while in subject 5 disintegration com-
`menced after the tablet reached the ascending colon.
`Complete tablet disintegration occurred 12n2p3n8 h
`(range 5n9–14n9 h ; n l 5) post-dose. In fact, complete
`disintegration did not occur in the first 16 h post-dose in
`three of the eight subjects who received formulation CD
`,
`$
`and,
`in the remaining five subjects, disintegration
`occurred in the distal colon.
`A summary of the pharmacokinetic parameters from all
`four formulations is shown in Table 3. The largest value
`of tmax, MRT, MAT and smallest value of Cmax and AUCinf
`can be found in formulation CD
`. All three CD formu-
`$
`lations showed reduced AUC relative to the IR for-
`mulation. The relative bioavailability of the delayed-
`release preparations (CD
`, CD
`and CD
`) to the IR
`"
`#
`$
`formulation (excluding subject Nos 17 and 32) based on
`), 37n9 % (CD
`the mean AUCinf were : 56n7 % (CD
`) and
`44n0 % (CD
`).
`$
`Statistical comparisons of the pharmacokinetic para-
`)
`meters (AUC
`!, −t, AUCinf, Cmax, tmax, kel, MRToral and t
`"/#
`were made between the IR formulation and the three CD
`formulations tested. When CD
`, CD
`and CD
`were
`"
`#
`$
`compared individually to the IR formulation, a significant
`difference (P 0n05) was observed for every comparison.
`
`#
`
`"
`
`Figure 1. Mean serum concentrations of dexamethasone over
`time in subjects receiving formulations IR, CD
`, CD
`, or CD
`) ;
`"
`#
`$
`n l 8 per formulation.
`
`release was observed in the colon in subjects 3, 11, 18,
`25 and 31.
`The gastrointestinal transit and disintegration proper-
`ties of
`formulation CD
`are illustrated by a series of
`#
`scintigraphic images from a representative subject in
`Figure 2.
`Serum dexamethasone concentrations in subjects re-
`ceiving formulation CD
`are presented in Figure 1. Initial
`tablet disintegration (see Table 2) occurred 3n6p1n6 h
`
`$
`
`# 1997 Blackwell Science Ltd, Aliment Pharmacol Ther 11, 205–213
`
`Cosmo Ex 2008-p. 5
`Mylan v Cosmo
`IPR2017-01035
`
`

`

`210
`
`C . J . K E N Y O N e t a l .
`
`Figure 2. Gastrointestinal transit and in vivo disintegration of formulation CD
`#
`subject.
`
`as assessed by gamma scintigraphy from a selected
`
`# 1997 Blackwell Science Ltd, Aliment Pharmacol Ther 11, 205–213
`
`Cosmo Ex 2008-p. 6
`Mylan v Cosmo
`IPR2017-01035
`
`

`

`C O L O N I C D E L I V E R Y O F D E X A M E T H A S O N E
`
`211
`
`Table 3. Pharmacokinetic parameters and measures of the four formulations tested
`
`Formulation
`
`IR
`CD
`"
`CD
`#
`CD
`$
`
`AUCinf
`(ng. h\mL)
`
`721p329 (46)*
`462p210 (46)
`311p134 (45)
`353p113 (32)
`
`Cmax
`(ng\mL)
`
`92n7p48n4 (52)
`32n1p15n1 (47)
`18n4p5n4 (30)
`19n1p4n2 (22)
`
`tmax
`(h)
`
`1n8p1n8 (98)
`10n6p5n0 (47)
`12n0p6n9 (58)
`6n0p3n5 (58)
`
`MRT
`(h)
`
`6n8p1n4 (21)
`12n6p3n0 (24)
`15n1p3n7 (25)
`14n4p1n9 (13)
`
`MAT
`(h)
`
`2n6p1n4 (55)
`8n3p3n0 (37)
`10n8p3n7 (35)
`10n1p1n9 (19)
`
`* Data are means (n l 8)ps.d. (%CV) except formulation IR, which included only the six subjects demonstrating immediate
`disintegration and apparent drug release.
`
`parameters influence the gastric emptying of pharma-
`ceutical preparations : the physical size of the solid dosage
`form and whether it is administered to a fed or fasted
`stomach. Following administration in the fasted state,
`single unit dosage forms are treated as indigestible
`material by the stomach and are emptied along with the
`phase 3 activity of the migrating myoelectric complex
`(MMC). Thus, variations in gastric emptying time can be
`large : the average gastric emptying time was 56p42
`min for subjects receiving formulation CD
`. No case of
`"
`prolonged gastric residence was observed in the study
`suggesting that
`the ‘ housekeeper wave ’ efficiently
`emptied the preparations from the stomach.
`Transit data from the CD formulations are in agreement
`with those reported previously for solutions, pellets and
`tablets."( It is generally accepted that small intestinal
`transit is not affected by the digestive state of the
`gastrointestinal tract. Cases of both rapid and prolonged
`small intestinal transit were observed (e.g. subject 17
`(1n3 h) ; subject 7 (6n9 h)) ; similar examples have been
`reported previously in the literature. Both radiological
`studies") and scintigraphic studies"*, #! have demon-
`strated that dosage forms often accumulate at the ileo-
`caecal
`junction before entering the colon. In some
`instances, passage across the ileo-caecal junction is rapid
`and in others stagnation can occur for many hours,
`resulting in extended small intestinal transit times.") This
`effect has been observed for both multiple- and single-
`unit controlled-release systems."(, #" No obvious relation-
`ship between tablet size and transit through the ileo-
`caecal junction has been described.#!
`In several subjects, colon arrival accompanied the
`ingestion of food at lunch time (e.g. subject 19 (CD
`) ;
`"
`subject 3 (CD
`) ; subject 4 (CD
`)). Mass movements
`#
`$
`accompanying eating have been reported previously and
`support the hypothesis that the ileo-caecal region has a
`reservoir function.## The gastrocolonic response, i.e. the
`response of the colon to eating, plays an important role in
`
`Figure 3. Relationship between mean absorption time (MAT) and
`total disintegration time (post-dose) of the four formulations
`tested.
`
`The relationship between various disintegration events
`and the pharmacokinetics, which provides an indirect
`assessment of release and subsequent absorption in the
`gastrointestinal tract was examined. An example of the
`relationship is shown in Figure 3 where the MAT is
`related to the total time to disintegration post-dose. These
`data show that increasing the time to total disinte-
`gration post-dose is directly related to an increase in mean
`absorption time of the drug, which is a measure of the
`effect of the dosage form on dissolution and absorp-
`tion from the gastrointestinal
`tract. Other disinte-
`gration events (see Table 2) also correlate reasonably
`well with MAT across the four formulations tested in this
`clinical study.
`
`D I S C U S S I O N
`
`The basis for successfully delivering a drug specifically to
`the colon relates to the transit properties of the dosage
`form and its drug release characteristics. Two main
`
`# 1997 Blackwell Science Ltd, Aliment Pharmacol Ther 11, 205–213
`
`Cosmo Ex 2008-p. 7
`Mylan v Cosmo
`IPR2017-01035
`
`

`

`212
`
`C . J . K E N Y O N e t a l .
`
`"
`
`#
`
`determining colon arrival of tablet preparations.#$, #% In
`healthy subjects, large single-unit dosage forms transit
`the ascending and transverse colon more quickly than
`solutions or pellets.'
`occurred at
`Complete tablet disintegration of CD
`12n3p3n2 h (range 6n7–15n6 h ; n l 8) post-dose, and
`was more distal to that reported for formulation CD
`. In
`addition, Cmax with this regimen (18n4 ng\mL) was much
`(32n1 ng\mL). As a result,
`lower than that with CD
`"
`systemic adverse events induced by dexamethasone may
`be reduced with this regimen relative to the non-targeted
`formulation.
`The reason for the dichotomy (slow drug release rather
`than the expected immediate release profile) in two of the
`eight subjects receiving the IR formulation is unclear.
`Scintigraphic studies (unpublished) on small hydrophilic
`matrix tablets (e.g. 5–7 mm in diameter) suggest that the
`rate of hydration of hydrophilic polymers is critical to the
`formation of the gel
`layer. Integrity of the gel
`layer
`probably determines the rate of tablet disintegration (the
`more cohesive the gel
`layer, the slower the disinte-
`gration). It is possible that in six of the eight subjects, the
`gel layer did not form quickly, resulting in rapid tablet
`disintegration, while in subjects 17 and 32 hydration of
`the polymer occurred rapidly, thereby significantly im-
`proving tablet
`integrity.
`In conclusion, six subjects
`receiving the IR formulation displayed immediate release
`characteristics while the remaining two, subject Nos 17
`and 32, exhibited delayed release.
`High serum dexamethasone levels were measured for
`the IR formulation over the first 2 h after dosing. In
`contrast, the other three formulations were considerably
`more stable in the upper gastrointestinal tract than the
`IR control formulation. While CD
`began disintegrating
`"
`in the distal small
`intestine, CD
`and CD
`normally
`#
`$
`resisted disintegration until they were in the terminal
`ileum or in the ascending colon. Despite measurable
`amounts of drug absorbed, all three CD forms delivered
`substantial amounts of drug into the colon. All three CD
`formulations released drug in the upper gastrointestinal
`tract. In some cases, premature drug release made
`correlation of tablet disintegration and drug release less
`reliable. Nonetheless, the amounts of drug reaching the
`colon ranged from 72 % to over 82 % of that administered,
`depending on the formulation. It is interesting to note
`that the estimates for the amount of drug delivered into
`the colon assume drug released in the small intestine
`from the CD formulations was absorbed as efficiently as
`that of the IR formulation. The overall performance
`
`Table 4. Summary of performance characteristics of the four
`formulations tested
`
`Formulation
`
`Function
`
`Performance, in vivo
`
`IR
`
`Immediate release
`
`CD
`
`"
`
`CD
`
`#
`
`CD
`
`$
`
`a colonic delivery
`system
`
`a colonic delivery
`system
`
`a colonic delivery
`system
`
`—rapid drug
`release\disintegra-
`tion in stomach
`—highest AUC
`—drug release\
`disintegration in
`ascending colon
`—72 % of drug was
`delivered to colon
`—slow drug release
`—lowest AUC
`— 81 % of drug was
`delivered to colon
`—slow drug release
`— 78 % of drug was
`delivered to colon
`
`characteristics of the four formulations evaluated are
`summarized in Table 4.
`It appears that the more drug release was delayed (or
`prolonged) by the formulation, the less total drug was
`absorbed from the gastrointestinal tract. The reduction in
`amount of drug absorbed over time, as assessed by AUC,
`could be caused by several factors. One explanation is
`that dexamethasone was incompletely released from the
`tablets and hence was unavailable for absorption (the
`drug would then be excreted along with any undissolved
`tablet matrix material). However, scintigraphic data
`suggest that the tablets completely disintegrated before
`excretion and therefore drug should have been available
`for absorption. Assuming that the drug was released
`from the matrix, the reduced AUC may be due to a
`relatively slow rate of absorption from the colon relative
`to that from the small intestine. Many drugs are absorbed
`relatively slowly from the colon. Another contributing
`factor could be incomplete dissolution of dexamethasone
`in the colon. Less water is present in the more distal
`regions of the colon and as a result, there may have been
`insufficient water to adequately dissolve the drug. Finally,
`the drug may have been degraded (metabolized) in the
`lumen by gut microflora. Recovery of dexamethasone
`from stools would help provide information on the fate of
`dexamethasone delivered into the colon. A clinical trial
`in inflammatory bowel disease patients with colonic
`delivery dosage form is needed to determine whether
`reduced glucocorticoid AUC is associated with adequate
`efficacy.
`
`# 1997 Blackwell Science Ltd, Aliment Pharmacol Ther 11, 205–213
`
`Cosmo Ex 2008-p. 8
`Mylan v Cosmo
`IPR2017-01035
`
`

`

`Strategies to reduce the amount of drug released in the
`first 3–4 h include enteric coatings. This approach should
`allow more drug to reach the large intestine. Ideally,
`drug release would continue well into the large bowel.
`This delivery profile would permit treatment of distal
`inflammation, which is common in ulcerative colitis
`patients. It should be noted that these dosage forms were
`designed to provide a sustained or prolonged release of
`drug in the colon, and the scintigraphic data indicate
`that this goal was achieved. For treatment of distal
`ulcerative colitis, delivery will need to be prolonged still
`further than was demonstrated by even the slowest
`release formulation tested herein. However, there may be
`instances wherein release of drug should be relatively
`rapid once the formulation has reached the ascending
`colon. This delivery profile can be accomplished using the
`technology demonstrated herein by adjusting several
`formulation parameters and the judicious use of coatings.
`
`R E F E R E N C E S
`
`1 Lo$ fberg R. New steroids for inflammatory bowel disease.
`Inflamm Bowel Dis 1995 ; 1 : 135–41.
`2 Tozer TN, Friend DR, McLeod AD. Kinetic perspectives on
`colonic delivery. STP Pharma Sci 1995 ; 5 : 5–12.
`3 McLeod AD, Cui N, Friend DR, Tozer TN, Fedorak RN. A novel
`dextran-glucocorticoid prodrug accelerates healing with
`limited adrenal suppression in experimental colitis in rats.
`Gastroenterology 19