Aliment Pharmacol Ther 1998; 12: 591–603.
`
`Review article: issues in oral administration of locally acting
`glucocorticosteroids for treatment of inflammatory bowel disease
`
`D. R. FRIEND
`CIBUS Pharmaceutical Inc., California, USA
`
`Accepted for publication 24 February 1998
`
`SUMMARY
`
`Inflammatory bowel diseases are treated in some cases
`by local administration of anti-inflammatory drugs.
`Local delivery of drugs in the colon following oral
`administration may lead to improved efficacy/side-effect
`profiles and may improve patient compliance. This
`review covers a number of
`issues important in the
`design of oral delivery systems of glucocorticosteroids
`for local therapy of colonic inflammation. The choice of
`specific glucocorticosteroids is based on the drug’s
`physicochemical and pharmacological properties. The
`
`administered
`conditions under which an orally
`glucocorticosteroid (or other drug) must be delivered
`to treat ulcerative colitis are also discussed. These
`conditions
`include variations
`in local pH,
`transit
`throughout
`the gastrointestinal
`tract,
`the potential
`role of gut microflora, and drug dissolution in both
`the healthly and diseased large intestine. The effective
`delivery of
`topically-active glucocorticosteroids
`in
`ulcerative
`colitis and Crohn’s
`colitis patients
`is
`complex, but
`if
`successful
`could improve
`their
`usefulness.
`
`INTRODUCTION
`
`Delivery of drugs to the colon is potentially useful in the
`treatment of colonic inflammatory bowel disease (IBD)
`(i.e. ulcerative colitis (UC), Crohn’s colitis). In addition,
`many sustained or extended release dosage forms
`deliver substantial amounts of drug into the large
`intestine, from which it is absorbed. Delivering peptide
`and protein drugs into the colon may enhance their oral
`bioavailability also.
`Several comprehensive reviews are available on meth-
`ods used in oral colonic drug delivery.1–6 This review
`presents information on choice of drug and on design-
`ing a colonic drug delivery system to improve treatment
`of IBD using topically active glucocorticosteroids (GCSs).
`Topically active GCSs are still largely unproven in their
`ability to substantially improve treatment of IBD with a
`
`Correspondence to: Dr D. R. Friend, CIBUS Pharmaceutical, Inc., PO Box
`1226, Menlo Park, CA 94026-1226, USA. E-mail: drfriend1@aol.com
`
`concomitant reduction in side effects. The reasons for
`this lack of clear success could, in part, be due to the
`lack of appropriate delivery systems. The primary focus
`of this review is the treatment of ulcerative colitis,
`although Crohn’s disease is also considered.
`The location of disease in UC at diagnosis is shown in
`(A) of Figure 1, while the location of inflammation in
`Crohn’s patients at diagnosis is shown in (B). As stated
`below, the site of inflammation in UC extends to the
`more proximal regions of
`the colon over time.
`In
`Crohn’s disease, between 30 and 40% of patients have
`significant colonic involvement.7 Thus many colonic
`drug delivery systems may be useful
`in treating a
`number of Crohn’s patients in addition to many UC
`patients.
`The development of a colon targeted preparation with
`reduced side-effect profiles requires a thorough under-
`standing of the drug substance as well as an appreci-
`ation of the role disease will play while the delivery
`system moves through the gastrointestinal tract. This
`review presents information on these and other issues
`
`(cid:211) 1998 Blackwell Science Ltd
`
`591
`
`Cosmo Ex 2028-p. 1
`Mylan v Cosmo
`IPR2017-01035
`
`

`

`592
`
`D. R. FRIEND
`
`Figure 1. Localization of inflammation at diagnosis in 783
`patients with UC (A) and in 195 patients with Crohn’s disease (B).
`These data indicate the percentage of patients with involvement of
`each respective colonic segment. From ref. 7.
`
`that are important in the development of an effective
`local delivery system for treatment of colonic inflam-
`mation with GCSs in IBD patients.
`
`GLUC OCORTICOSTEROIDS IN T HE TREATMENT
`OF IBD
`
`GCSs have been used to treat IBD for many years and
`are the mainstay in the treatment of active UC.8 GCSs
`are still used extensively to control symptoms of IBD
`because more specific therapeutics have not been
`developed. The mechanism of action of GCSs has been
`extensively reviewed.9–12 GCSs affect many cell types
`involved in the pathogenesis of IBD.13–15 Early con-
`trolled trials of GCS therapy in UC patients showed that
`70–90% of those treated with cortisone or prednisolone
`improved compared with 30–40% of those treated with
`placebo.16, 17 Their primary limitation has been side-
`effects (see below).
`The pharmacokinetics of many GCSs are well-known.
`In general, plasma concentrations of GCSs vary consid-
`erably after oral ingestion of the same dose by normal
`volunteers and patients.18 Many GCSs have elimination
`half-lives from plasma of 2–4 h, whereas the biological
`half-life ranges from 18 to 36 h.19 Recent developments
`in the use of GCSs to treat IBD suggest that absorption
`may not be necessary, and in light of the known side-
`effect profiles, is probably undesirable.
`The long-term use of GCSs in the treatment of IBD
`produces
`iatrogenic Cushing syndrome leading to
`adrenal atrophy, osteoporosis, aseptic necrosis of the
`
`head of the femur, weight gain with redistribution of fat
`to the truncal areas, ‘moon face’, ‘buffalo hump’, and
`many other side-effects.20 Thus, a goal of GCS therapy
`of IBD would be a high local anti-inflammatory effect in
`the mucosa with negligible amounts of the active agent
`absorbed into the systemic circulation to avoid systemic
`side-effects. Reaching such a goal would permit evalu-
`ation of GCS-based therapy in maintaining remission of
`IBD.
`
`Local effects of glucocorticosteroids
`
`The systemic side-effects of GCSs have long been
`evident. This problem, plus the knowledge that many
`UC patients have distal disease, has led to the develop-
`ment of rectally applied dosage forms. Enema dosage
`forms have been used extensively to treat distal
`ulcerative colitis and proctosigmoiditis. GCSs have been
`used effectively in these dosage forms to obtain a local
`therapeutic effect.21–26 One goal of
`this
`route of
`administration was to reduce the side-effects often
`observed with systemic GCS therapy.27–29 By delivering
`the steroid to the site of action (inflamed mucosa),
`smaller doses can theoretically be administered, and as
`a result a lower total systemic exposure is possible. Local
`therapy of distal colitis has proved effective using
`retention enemas of hydrocortisone and other GCSs,
`such as betamethasone.30–31 However, topically applied
`steroids possessing systemic activity, such as hydrocor-
`tisone hemisuccinate or prednisolone-21-phosphate,
`exert side-effects following rectal administration and
`absorption.32–33
`rectal preparations
`Another
`issue in the use of
`(enemas,
`foams, suppositories) is the ability of these
`dosage forms to gain access to only a portion of the
`distal colon. A recent study has pointed out
`the
`limitations of two budesonide enema preparations in
`reaching the splenic flexure.34 Similar studies with
`rectal foams and enemas containing 5-amino salicylic
`acid show that only the rectosigmoid region can
`consistently be reached in patients with active UC.35
`While foams and suppositories are better accepted by UC
`patients than enemas,36 rectal preparations are gener-
`ally unpopular even in the motivated IBD sufferer.37
`
`Topically active glucocorticosteroids in IBD
`
`The inability to dissociate the anti-inflammatory effects
`from the unwanted side-effects of oral GCSs has led to
`
`(cid:211) 1998 Blackwell Science Ltd, Aliment Pharmacol Ther 12, 591–603
`
`Cosmo Ex 2028-p. 2
`Mylan v Cosmo
`IPR2017-01035
`
`

`

`REVIEW: ORAL ADMINISTRATION OF GLUCOCORTICOSTEROIDS FOR IBD
`
`593
`
`Table 1. GCSs—potentially useful in the treatment of UC
`
`Systemic*
`
`Hydrocortisone
`Prednisone
`
`Prednisolone
`Dexamethasone
`Methylprednisolone
`Betamethasone
`Cortisone acetate
`Triamcinolone
`Fluoromethalone
`Desoximetasone
`Fludrocortisone acetate
`Cortisone
`
`Topical
`
`Budesonide
`Prednisolone sodium
`metasulphobenzoate
`Tixocortol pivalate
`Prednisolone sodium phosphate
`Flunisolide
`Triamcinolone acetonide
`Flucinonide
`Desonide
`Beclomethasone dipropionate
`Fluticasone propionate
`Hydrocortisone acetate
`
`* Most systemically acting GCSs can be administered topically with
`exception of non 11-hydroxylated forms such as prednisone and
`cortisone.
`
`the development of topically active compounds. These
`compounds are active as administered, but following
`absorption into the systemic circulation they are
`metabolized to inactive compounds.
`Table 1 lists two groups of GCSs based on their mode of
`action (systemic or topical). Most of the systemically
`active GCSs listed in Table 1 can be administered
`topically with therapeutic effect; most of the compounds
`listed as topically active in Table 1 are metabolized upon
`absorption. A discussion of the pharmacokinetics of
`topically active GCSs has been published.38
`The pharmacokinetics of local delivery of GCSs have
`been analysed and described in terms of a selective
`advantage or drug delivery index.39 Based on calcula-
`tions of selective advantage, topically active GCSs can be
`used to reduce the side-effects associated with the oral
`administration of systemically active GCSs. The feasibil-
`ity of colonic delivery of GCSs to increase local tissue
`concentrations and lower potential side-effects has been
`demonstrated in several animal studies using the
`prodrug approach.40–42
`The issue of systemic versus topically active GCSs is
`important; to-date, the place of topically active GCSs in
`the treatment of IBD is relatively unproven. Delivering
`systemically active GCSs (e.g. prednisolone) to the colon
`will probably be efficacious. However, the potential for
`side-effects will still exist unless the total dose admin-
`istered is lowered. Swartz & Dluhy19 have estimated
`that the daily dose of prednisolone above which adrenal
`suppression occurs is 5–6 mg/day. Even with a colon
`
`(cid:211) 1998 Blackwell Science Ltd, Aliment Pharmacol Ther 12, 591–603
`
`targeted preparation, it is unknown if efficacy can be
`adequately maintained at such a dose. Normally, the
`daily dose of prednisolone required to control symptoms
`of UC ranges from 40–60 mg per day.8 Data from
`enema preparations suggest that even though local or
`topical application of systemically active GCSs is effica-
`cious, some side-effects are still observed. Thus, the use
`of topically active GCSs in oral colon-targeted prepara-
`tions must be considered.
`One of the early compounds developed for topical
`treatment of IBD was prednisolone metasulphobenzoate.
`Following rectal administration, significantly less drug
`is absorbed compared to prednisolone-21-phosphate,
`with similar therapeutic effect.21, 29 Another compound
`used in Europe for local treatment of distal colitis is
`tixocortol pivalate. This compound is a cortisol deriva-
`tive with low receptor affinity (and relatively low
`potency). The compound is poorly bioavailable follow-
`ing oral administration due to metabolism upon
`absorption.43 Both of
`these compounds have been
`generic for a number of years and are not approved
`for human use in the USA.
`Budesonide has probably been studied more extensive-
`ly than any other topically active GCS for the treatment
`of IBD. Budesonide is the active ingredient in the nasal
`inhalation product Rhinocort (Astra USA Inc.). Pulmi-
`cort (Astra USA Inc.) has recently been approved by the
`FDA for treatment of asthma. Budesonide is also
`available in Europe and Canada as an enema prepara-
`tion (Entocort; Astra, Lund) and some countries as an
`oral treatment for Crohn’s disease (Entocort CIR).
`Budesonide is a relatively potent GCS with low oral
`bioavailability due to extensive first pass hepatic
`metabolism. This drug is metabolized via oxidative and
`reductive pathways in the liver to six metabolites,
`predominately 6b-hydroxy-budesonide and 16a-hy-
`droxy-prednisolone.44, 45 Both metabolites have very
`little anti-inflammatory activity compared with bud-
`esonide. About 90% of an orally administered dose of
`budesonide is metabolized in the liver, leading to an oral
`bioavailability of 10–15%. Budesonide is metabolized
`2–4 times more rapidly than beclomethasone depropi-
`onate (see below) and 2–3 times faster than trimacino-
`lone acetonide.46 Table 2 lists some of the pharmaco-
`kinetic parameters and measures following single and
`multiple doses by enema or oral controlled ileal release
`budesonide dosage forms. The doses most commonly
`used for treatment of Crohn’s disease and UC range from
`6 to 15 mg/day.
`
`Cosmo Ex 2028-p. 3
`Mylan v Cosmo
`IPR2017-01035
`
`

`

`594
`
`D. R. FRIEND
`
`Table 2. Pharmacokinetic measures and parameters of single and multiple doses of budesonide in healthy volunteers and patients with
`distal UC, proctitis or Crohn’s disease
`
`Study participants (n)
`
`Daily dose
`
`Cmax
`(nmol/L)*
`
`tmax
`(h)
`
`AUC
`(nmol/L h)
`
`Systemic
`availability
`(%)
`
`References
`
`Enema
`Healthy volunteers (10)
`Healthy volunteers (15)
`Patients with UC or P (24)
`Patients with UC or proctitis (24)
`
`Oral
`Healthy volunteers (12)
`Healthy volunteers (12)
`Healthy volunteers (12)
`Patients with Crohn’s disease (18)
`Patients with Crohn’s disease (18)
`
`2 mg/100 mL
`2 mg/100 mL
`2 mg/100 mL
`2 mg/100 mg · 28 days(cid:160)
`
`3 mg · 5 days(cid:160)
`9 mg · 5 days(cid:160)
`15 mg · 5 days(cid:160)
`4.5 mg
`4.5 mg b.d. · 8 weeks
`
`3
`2.1
`2.5
`
`1.9
`5.3
`8.0
`4.1
`3.2
`
`1.2
`1.3
`1.3
`1.2
`
`2.3
`2.7
`3.1
`4.7
`4.4
`
`13.6
`9.7
`11.6
`
`11.5
`37.0
`60.2
`28.5
`21.1
`
`13.0
`15.2
`
`118
`119
`120
`120
`
`121
`121
`121
`122
`122
`
`* Multiply by 0.431 for lg/L.
`(cid:160) Pharmacokinetic parameters were measured after last dose of budesonide, which was administered once daily.
`
`The pharmacology and therapeutic efficacy of bud-
`esonide in the treatment of IBD,47 asthma and rhinitis48
`have been reviewed. Budesonide has greater topical
`potency in skin vasoconstriction assays than beclomet-
`hasone, betamethasone valerate, desonide, flunisolide,
`hydrcortisone butyrate, prednisolone, fluocinolone, pre-
`dnacinolone or triamcinolone acetonide, and has sim-
`ilar activity to betamethasone dipropionate.49 Numer-
`ous
`studies have demonstrated the usefulness of
`budesonide in treating both Crohn’s disease and UC
`when administered topically via an enema preparation
`or as an oral targeted release preparation.48, 50 In most
`studies, the efficacy of budesonide has been compared
`with that of oral prednisolone. Recent in vitro data also
`suggest that once-daily dosing of budesonide should be
`sufficient in the treatment of IBD.51
`The potential of budesonide to treat UC has been
`studied in several clinical trials. As with other trials
`performed with budesonide, the efficacy and side-effect
`profiles were compared with those of oral predniso-
`lone.50 A dose of 10 mg of budesonide daily, formulated
`in enteric coated sustained-release pellets, was com-
`pared with 40 mg of prednisolone daily for 9 weeks in a
`randomized, double-blind controlled trial. Mean endo-
`scopic scores improved significantly in both groups, but
`without any significant differences. Morning plasma
`cortisol
`levels were suppressed in the prednisolone
`group, but were unchanged relative to baseline in the
`budesonide group.50
`The key finding in this study is the lack of cortisol
`suppression in the budesonide group. Based on previous
`
`trials with budesonide, it was expected that there would
`be some, however minor, suppression of plasma cortisol
`levels. The reasons for this lack of suppression may be:
`(i) incomplete release of budesonide or (ii) the rate of
`absorption was slower compared with other formula-
`tions tested previously with budesonide,
`leading to
`lower peak blood levels. No pharmacokinetic data were
`available to assess the relationship between serum
`concentrations of budesonide and plasma cortisol
`levels.50 The authors noted that the most inflamed
`portion of the colon in UC may be exposed to the drug
`intermittently and only before defaecation.
`The delivery of budesonide from another pH-dependent
`delivery system under development in Europe has been
`studied.52–54 These efforts have focused on delivery of
`budesonide to the ileocecal region for the treatment of
`Crohn’s disease.
`An important aspect of colonic delivery of GCSs is the
`potential to use these drugs in maintenance therapy.
`One study to-date has addressed the potential of oral
`budesonide in maintaining remission of Crohn’s dis-
`ease.55 At a daily dose of 6 mg over 1 year, budesonide
`prolonged the time to relapse in patients with Crohn’s
`disease of the ileum and proximal colon.55 This effect
`was noted for up to 6 months, after which budesonide
`was no better than placebo. Considerably more data are
`required to determine the place of topically active GCSs
`in maintenance therapy of IBD.
`There are several other GCSs potentially useful in the
`treatment of UC when formulated in a colon-targeted
`drug delivery system. These are fluticasone propionate,
`
`(cid:211) 1998 Blackwell Science Ltd, Aliment Pharmacol Ther 12, 591–603
`
`Cosmo Ex 2028-p. 4
`Mylan v Cosmo
`IPR2017-01035
`
`

`

`REVIEW: ORAL ADMINISTRATION OF GLUCOCORTICOSTEROIDS FOR IBD
`
`595
`
`beclomethasone dipropionate and flunisolide. Fluticas-
`one propionate is a topically active GCS developed by
`Glaxo. The drug is under patent until 2003 in the USA.
`The other two compounds are generic but are not
`indicated for use in UC or Crohn’s disease in the USA.
`Fluticasone propionate is practically insoluble in water
`but is soluble in a number of non-polar organic solvents.
`The oral bioavailability using labelled and unlabelled
`drug is less than 1% primarily due to incomplete
`absorption and pre-systemic metabolism in the gut and
`liver.56 In contrast, the majority of fluticasone delivered
`into the lungs is absorbed systemically. Binding of
`fluticasone propionate to human GCS receptors is 18
`times greater than dexamethasone, nearly twice that of
`beclomethasone 17-monopropionate (the active metab-
`olite of beclomethasone dipropionate), and over three
`times that of budesonide. Total blood clearance of
`fluticasone is equivalent to hepatic blood flow.57 After
`oral administration, 87–100% of fluticasone propionate
`is excreted in the faeces. Of this, 9–20% is unchanged
`fluticasone propionate after a 1 mg dose; in contrast,
`54–75% of a 16 mg dose is excreted unchanged in the
`faeces.57 It should be noted that
`these data were
`obtained in two subjects. Nonetheless, these results
`illustrate the need to enhance the dissolution of drug in
`the colon to maximize local bioavailability. Compared
`with budesonide, fluticasone propionate dissolves 80
`times more slowly in human bronchial fluid (6 min vs.
`8 h).58
`The anti-inflammatory effect of fluticasone propionate
`has been examined in both Crohn’s disease59 and
`UC.60, 61 In the evaluation of Crohn’s disease, fluticas-
`one (20 mg/day orally in four divided doses over
`3 weeks) significantly reduced disease activity without
`any significant changes
`in plasma cortisol
`levels.
`However, this study was uncontrolled, and as a result
`the most meaningful parameter measured was the
`cortisol levels. No information on the dosage form was
`provided in this study.59
`Fluticasone propionate was evaluated in UC patients
`with active left-sided or pan colitis.60 Two hundred and
`five patients were evaluated in this study. The investi-
`gators concluded that fluticasone propionate (20 mg/
`day) was not as effective as prednisolone (40 mg/day),
`although there was a complete absence of suppression
`of corticoadrenal axis by fluticasone. Despite the fact
`that this compound is topically active, the dosage form
`was not designed, per se, to deliver the drug to the colon,
`particularly the more distal regions. It was hoped that
`
`(cid:211) 1998 Blackwell Science Ltd, Aliment Pharmacol Ther 12, 591–603
`
`relatively large amounts of the drug would reach the
`colon and eventually be excreted in the faeces, and that
`a colonic delivery system was unnecessary.55 Another
`trial with fluticasone was performed in patients with
`active distal UC.61 The conclusion of this study was that
`after 4 weeks of therapy, fluticasone propionate (5 mg
`four times daily) was ineffective in controlling active
`distal disease. Again, the dosage form in this placebo-
`controlled study was not a colon targeted preparation.
`A number of studies have indicated that fluticasone
`does not lower plasma cortisol
`levels or urinary-free
`cortisol excretion even following an oral dose of
`16 mg.57 The plasma concentration of fluticasone at
`which a 50% reduction in plasma cortisol results has
`been estimated at 0.134 ng/mL for total fluticasone
`propionate and 0.013 ng/mL for unbound drug.62
`When fluticasone propionate is administered orally at
`up to 16 mg,
`less than 0.130 ng/mL total drug is
`detected unchanged in plasma.
`topically
`Beclomethasone dipropionate is another
`active GCS potentially useful in the treatment of UC.
`This generic compound has been used to treat asthma
`and rhinitis and has also been examined in the topical
`therapy of IBD.27, 28 It has been studied less extensively
`than budesonide for the treatment of IBD, nonetheless it
`does possess similar properties compared with bud-
`esonide. Data from one pilot study in humans has been
`published wherein the delivery system (an enterically
`coated hydroxypropylmethylcellulose tablet) released
`drug in the distal small intestine and proximal large
`intestine.63 Such a release profile makes it suitable for
`treatment of Crohn’s disease but
`less so for UC.
`Beclomethasone, also, is very poorly soluble in water,
`possibly creating bioavailability problems for a solid oral
`dosage form.
`
`COLONIC DRUG DELIVERY FOR IBD
`
`There are several approaches to delivering drugs to the
`large intestine.1–6 These approaches include enteric
`coatings designed to release drugs in the more alkaline
`environment of the lower gastrointestinal tract, bioer-
`odible coatings and matrices, prodrugs, timed-release
`systems and sustained-release dosage forms that release
`drug as they pass through the small and large intestine
`(more suitable for Crohn’s disease than UC). Specific
`features of these delivery systems are addressed below
`in the context of factors that can affect their perfor-
`mance.
`
`Cosmo Ex 2028-p. 5
`Mylan v Cosmo
`IPR2017-01035
`
`

`

`596
`
`D. R. FRIEND
`
`Luminal pH
`
`Both commercial and experimental delivery systems
`rely on differences in pH along the alimentary canal to
`control
`the site of drug release. Therefore,
`it
`is
`important to understand these differences and what
`role disease state may play in the performance of these
`systems in patients. Local pH levels can affect intestinal
`enzyme activities and hence the performance of delivery
`systems that rely on these enzymes to trigger drug
`release from biodegradable dosage forms and from
`prodrugs.
`The pH of normal subjects along the gastrointestinal
`tract is reasonably well characterized. The pH in the
`terminal ileum has been measured in normal ambulant
`human subjects at 7.5 (cid:139) 0.4 (n (cid:136) 66).64 Several
`studies show that luminal colonic pH is acidic due to
`the presence of short chain fatty acids (produced by
`fermentation of dietary fiber by colonic microflora).
`There is a drop in pH to a mean of 6.4 (cid:139) 0.4 in the
`caecum. The pH remains between 6.4 and 7.0 from the
`ascending colon to the left colon.64 Mucosal pH in the
`colon tends to be higher than luminal pH: from the right
`to the left colon the pH has been measured at
`7.05 (cid:139) 0.32 (n (cid:136) 21) to 7.38 (cid:139) 0.59 (n (cid:136) 21).65
`The luminal pH in patients with IBD can be lower than
`that measured in normal volunteers. Intraluminal pH
`was measured in six patients with active UC using
`radiotelemetry capsules. There was considerable vari-
`ability in the group studied. In several patients, colonic
`pH ranged from 5.0 to neutral.66 However, in three
`patients, the pH values were considerably lower: 2.3,
`2.9 and 3.4.66 Similar data have been collected in four
`Crohn’s disease patients. In this study, the minimum
`colonic pH was found to be 5.3 (cid:139) 0.3 in the right colon
`but lower colonic pHs were recorded in the more distal
`regions of the colon.67
`
`The use of enteric coatings to delay the release of drugs
`in the gastrointestinal tract and the reliability of this
`approach to colon targeted drug delivery have been
`addressed in normal volunteers.68, 69 The dosage forms
`used were coated with Eudragit S (Ro¨hm Pharma,
`Weiterstadt, Germany), an acrylic resin designed to
`dissolve at pH 7.0 or above. The material is used in both
`commercial and many experimental dosage forms
`designed to delay release of a drug until it reaches the
`terminal ileum or ascending colon. In accordance with
`the measured pH values along the length of
`the
`gastrointestinal tract, the tablets rapidly disintegrated
`at sites ranging from the ileum to the splenic flexure (see
`Table 3). From data presented, it can be concluded that
`the site of disintegration varied from the ileum to the
`splenic flexure. The authors concluded that such data
`demonstrate a lack of site specificity. The spreading of
`the disintegrated material was aboral with no back
`mixing of the colonic contents observed.69
`
`Intestinal transit
`
`The movement of a dosage form through the gastroin-
`testinal tract in IBD patients needs to be understood
`when designing oral-based colonic targeted prepara-
`tions. If transit is very rapid, particularly through the
`colon, drug release may be incomplete resulting in loss
`of efficacy. Likewise, excessively slow transit can lead to
`inadequate amounts of drug reaching the distal portions
`of the gut.
`The role of the dosage form (size, shape, density) and
`fasted versus fed conditions on gastrointestinal transit
`are relatively well understood in normal subjects.
`Gastric emptying in normal subjects has been tho-
`roughly reviewed.70 Normally, tablets and pellets empty
`from the fasted stomach within 60–90 min, while
`
`Table 3. Gastrointestinal transit times (h) for Eudragit S coated tablets (from reference 69)
`
`Vol.
`
`Gastric
`emptying
`
`Time in
`lower SI*
`
`Time
`through ICJ(cid:160)
`
`SI transit
`time
`
`Time in AC(cid:224)
`
`Main position at 12 h
`
`1
`3.0
`1.8
`6.1
`3.1
`2.1
`2
`1.6
`> 15
`> 15
`> 13
`—
`3
`1.4
`5.5
`8.7
`7.3
`0.5
`10.8–12.0§
`4
`6.1
`3.6
`5.1
`—
`5
`2.0
`1.1
`4.9
`2.9
`0.2
`6
`1.3
`> 12
`> 12
`> 10
`—
`5.0–5.5§
`7
`1.3
`0.0
`4.0
`0
`* SI (cid:136) small intestine; (cid:160) ICJ (cid:136) ileocecal junction; (cid:224) AC (cid:136) ascending colon; § Range as tablet disintegrated prior to ICJ.
`
`Hepatic flexure
`ICJ
`Splenic flexure
`AC
`Hepatic flexure
`ICJ
`Hepatic flexure
`
`(cid:211) 1998 Blackwell Science Ltd, Aliment Pharmacol Ther 12, 591–603
`
`Cosmo Ex 2028-p. 6
`Mylan v Cosmo
`IPR2017-01035
`
`

`

`REVIEW: ORAL ADMINISTRATION OF GLUCOCORTICOSTEROIDS FOR IBD
`
`597
`
`tablet dosage forms can be delayed
`emptying of
`considerably in the fed state. The literature predicts
`that standard size tablets should empty with the
`migrating myoelectric complex.71 It is unlikely that
`disease activity significantly alters gastric emptying in
`either the fed or fasted state. In any case, colon-targeted
`dosage forms can be enterically coated to prevent drug
`release in the stomach, assuming such coatings dissolve
`when the tablets, beads or pellets enter the duodenum.
`Generally,
`transit of most pharmaceutical dosage
`forms through the small intestine is relatively constant,
`even when they are administered with food. The
`average transit time of both pellets and tablets has
`been measured at about 3 (cid:139) 1 h.72 However, deviations
`from this average transit time have been noted; one
`study measured a mean small intestinal transit of 8 h
`(although the authors of this paper admit they may
`have misjudged location of the radiotelemetry capsule,
`which was 24 mm in length).73
`The transit of multiparticulate dosage forms (e.g. small
`pellets) normally involves an initial regrouping of the
`pellets prior to entrance and spreading in the colon.74–77
`Under fasting conditions, the ileum acts as a reservoir
`with colonic filling curves (based on scintigraphic
`imaging of a 1 mL solution of 99mTc-diethyltriamino-
`pentaacetic acid (DTPA)) showing long plateaus and low
`slopes indicative of episodic colonic inflows and wide
`spreading of the marker in the colon. In another study,
`transit through the ileocecal junction was unaffected by
`tablet size (3, 6, 9 and 12 mm diameter).78 In another
`study, filling of the colon (from the ileum) has been
`characterized by bolus movements followed by plateaus
`(periods of no movement).79 Emptying of the caecum
`and ascending colon had a half-time of 87.6 (cid:139) 27.0
`min, indicating that nearly all the material present in
`this region would be emptied in 4–5 h. Following transit
`into the transverse colon, transit slows, suggesting that
`this region may be one of faecal storage.80 Another
`study found that transit through the ascending and
`transverse colon was segmental; transit through the
`descending and sigmoid colon was relatively rapid
`indicating these regions are primarily faecal conduits.81
`Total colonic transit times of radio-opaque markers was
`26.2 (cid:139) 8.3 h.81 The transit of tablets or non-disinte-
`grating capsules through the colon tends to be more
`rapid than that of pellets greater than 1.0 mm in
`diameter.82–84
`In another normal volunteers study, segmental transit
`through the colon of a radio-opaque marker was
`
`(cid:211) 1998 Blackwell Science Ltd, Aliment Pharmacol Ther 12, 591–603
`
`assessed.85 Whole colonic transit times were 39 (cid:139) 5 h
`(n (cid:136) 38). Mean transit time in the right colon was
`13.8 h and 14.1 h in left colon (the authors divided the
`ascending, transverse and descending colon into two
`roughly equal portions of right and left). Transit through
`the rectosigmoid region was 11 h on average.85
`A transit study of different size tablets through the
`colon showed that smaller tablets (3 and 6 mm) were
`retained in the ascending colon longer than larger
`tablets were (9 and 12 mm).78 This finding is consistent
`with that of Hardy et al.84 who showed that large
`(25 mm · 9 mm), single unit dosage forms moved
`significantly faster through the proximal colon than
`did small pellets (0.5–1.8 mm).86 The rapid transit rates
`of larger size dosage forms through the colon has been
`called ‘streaming’. Where distinct
`liquid and solid
`phases are present in the colon (which can occur in
`both a healthy and a diseased colon), the two phases
`can move at different rates so that there is streaming of
`contents in the colon.87 As a result, larger units may
`transit the colon more rapidly than smaller units, as
`observed in several studies.78, 84, 86 Streaming has been
`postulated to occur in diverticular disease,87 while its
`role in IBD is unknown. If streaming does occur in
`patients with active IBD, it could lead to very slow or
`very rapid movement of dosage forms through the colon
`with a potential effect on drug delivery characteristics.
`Gastrointestinal transit in IBD can vary from that
`observed in healthy subjects, because mucosal inflam-
`mation disturbs the normal mechanism of absorption
`and transit.88 A common pathophysiological attribute
`of UC patients is diarrhoea. The diseased colon presents
`abnormalities in fluid and electrolyte absorption and
`secretion.89 The physical consistency of stools is also
`important clinically and can potentially affect drug
`delivery to the colon.
`The net effect of transit abnormalities has been studied
`using gamma scintigraphy. Transit data following oral
`administration of a radiolabelled liquid preparation in
`UC patients during active and quiescent disease are
`summarized in Table 4.90 A number of
`interesting
`observations can be made based on the results of this
`study. The gastric half-emptying time and mouth to
`caecum transit times of a radiolabelled meal were about
`the same regardless of disease site and activity. Whole
`gut
`transit times were quite long in all
`instances,
`ranging from 56 to 78 h depending on location and
`activity of UC. The difference between mouth-to-caecum
`and whole gut transit times gives relatively long colonic
`
`Cosmo Ex 2028-p. 7
`Mylan v Cosmo
`IPR2017-01035
`
`

`

`598
`
`D. R. FRIEND
`
`Table 4. Transit measurement and stool output during active and quiescent disease (paired data)* (from reference 90)
`Total colitis (n (cid:136) 6)
`Distal colitis (n (cid:136) 8)
`
`Active
`
`Quiescent
`
`P
`
`Active
`
`Quiescent
`
`P
`
`Half-time for gastric emptying (min)
`Mouth-to-caecum transit (min)
`Whole gut transit (h)
`Mean daily stool weight (g)
`Mean daily stool frequency
`
`46 (cid:139) 22*
`298 (cid:139) 62
`64 (cid:139) 22
`253 (cid:139) 72
`4.1 (cid:139) 1.0
`
`41 (cid:139) 6
`310 (cid:139) 60
`56 (cid:139) 29
`159 (cid:139) 62
`1.9 (cid:139) 1.0
`
`N.S.
`N.S.
`N.S.
`< 0.02
`< 0.002
`
`51 (cid:139) 23
`313 (cid:139) 87
`68 (cid:139) 49
`192 (cid:139) 99
`3.0 (cid:139) 0.8
`
`78 (cid:139) 33
`293 (cid:139) 82
`78 (cid:139) 55
`144 (cid:139) 79
`1.2 (cid:139) 0.7
`
`< 0.05
`N.S.
`N.S.
`< 0.05
`< 0.001
`
`* Data are means (cid:139) S.D.
`
`transit times ranging from 50 to 70 h. Stool weight was
`sign