354
`
`ARTICLE
`
`(/)
`
`0::: w
`0
`0:::
`0
`(/)
`0
`
`CJ
`_J
`<( z
`0
`1-u
`z
`=>
`LL
`
`Randomized Trial of 2 Delayed-Release Formulations of
`Linaclotide in Patients With Irritable Bowel Syndrome
`With Constipation
`
`, Susan M. Fox, PhD5
`, David S. Reasner, PhD4
`, Wilmin Bartolini, PhD3
`William D. Chey, MD1, Gregory S. Sayuk, MD2
`,
`Wieslaw Bochenek, PhD5
`, Ramesh Boinpally, PhD5, Elizabeth Shea, PhD3
`, Kenneth Tripp, PhD6 and Niels Borgstein, MD3
`
`INTRODUCTION: Immediate-release (IR) formulation of linaclotide 290 µg improves abdominal pain and constipation
`(APC) in patients with irritable bowel syndrome (18S) with constipation. Delayed-release (DR)
`formulations were developed on the premise that targeting the ileum (delayed-release formulation 1
`[DRl]) or ileocecal junction and cecum (MD-7246, formerly DR2) would modulate linaclotide's
`secretory effects while preserving pain relief effects.
`
`METHODS:
`
`RESULTS:
`
`This phase 2b study randomized patients with I8S with constipation to placebo or 1 of 7 once-daily
`linaclotide doses (DRl 30, 100, or 300 µg; MD-7246 30, 100, or 300 µg; or IR 290 µg) for 12 weeks.
`Key efficacy endpoints were change from baseline in abdominal pain and complete spontaneous bowel
`movement frequency, and 6/12-week combined APC+ 1 responder rate.
`
`Overall, 532 patients were randomized; mean age was 45.1 years, and most were women (83.3%) and
`White ( 64. 7%). All linaclotide DR 1 and MD-7246 groups experienced greater improvements in abdominal
`pain from baseline and vs placebo throughout treatment. Linaclotide DR 1 and IR led to numerically greater
`improvements from baseline in complete spontaneous bowel movement frequency and higher APC+ 1
`responder rates compared with placebo; MD-7246 results were similar to placebo. Diarrhea was the most
`common adverse event with DRl and IR; rates were similar between MD-7246 and placebo.
`
`DISCUSSION:
`
`Altering the site of drug delivery in the intestine might uncouple linaclotide's pain relief from secretory
`effects. Persistent, modest abdominal pain improvement with limited impact on bowel symptom
`parameters, as seen across MD-7246 doses, warrants further study of MD-7246 as a novel treatment for
`abdominal pain, regardless of I8S subtype.
`
`SUPPLEMENTARY MATERIAL accompanies this paper at http://links.lww.com/AJG/B699, http://links.lww.com/AJG/B700, http://links.lww.com/AJG/B70l , and http://
`links.lww.com/AJG/B702
`
`Am J Gastroenterol 2021;116:354-361 . https:!/doi.org/ 10.14309/ajg.0000000000000967
`
`INTRODUCTION
`Irritable bowel syndrome (IBS) is a functional gastrointestinal
`(GI) disorder characterized by recurrent abdominal pain associ(cid:173)
`ated with altered bowel habits (1,2) . IBS affects approximately
`11 % of adults globally (3) and significantly reduces quality oflife
`and work productivity (4,5). Although the predominant bowel
`habit varies from constipation to diarrhea, abdominal pain is the
`unifying and most bothersome symptom across IBS subtypes
`(1,6,7). Numerous therapies are used to treat IBS, but few provide
`relief of both abnormal bowel habits and abdominal pain (8).
`
`Linaclotide is a guanylate cyclase (GC)-C agonist currently
`approved as an immediate-release (IR) formulation for IBS with
`constipation (IBS-C) and chronic idiopathic constipation (9).
`Linaclotide acts locally within the GI tract, enhancing secretion
`and resulting in increased luminal water content and accelerated
`bowel transit (10-13). Preclinical evidence suggests that linaclo(cid:173)
`tide inhibits the activity of pain-sensing nerves by increasing
`extracellular cyclic guanosine monophosphate released through
`the basolateral membrane of intestinal epithelial cells (14-17). In
`phase 3 trials, linaclotide IR 290 µg has demonstrated efficacy in
`
`1Division of Gastroenterology and Hepatology, Michigan Medicine, Ann Arbor, Michiga n, USA; 2Division of Gastroenterology, Washington University School of
`Med icine, St. Louis, Missouri, USA; 3 1ronwood Pharmaceutica ls, Inc., Boston, Massachusetts, USA; 41mbria Pharmaceuticals, Boston, Massachusetts, USA;
`5AbbVie, Inc., Madison, New Jersey, USA; 6Cyclerion Therapeutics, Cambridge, Massachusetts, USA. Correspondence: Wilmin Bartolin i, PhD.
`E-mail : wbartoli ni@ironwoodpharma.com.
`Received May 21, 2020; accepted August 27, 2020; published online October 15, 2020
`
`The American Journal of GASTROENTEROLOGY
`
`VOLUME 116 I FEBR UARY 2021 www.amjgastro.com
`
`Copyright © 2020 by The American College ofGastroenterology. Unauthorized reproduction of this article is prohibited.
`
`Bausch Health Ireland Exhibit 2059, Page 1 of 8
`Mylan v. Bausch Health Ireland - IPR2022-00722
`
`

`

`Delayed-Release Formulations of Linaclotide
`
`355
`
`(/)
`
`0:::: w
`0
`0::::
`0
`(/)
`0
`
`(9
`_J
`
`<( z
`0
`1-u
`z
`=:)
`LL
`
`treating both the constipation and abdominal pain symptoms
`associated with IBS-C (18,19) with a favorable safety profile (20) .
`Visceral hypersensitivity is recognized as a central mechanism
`for abdominal pain in IBS (21). The hypersensitivity typical ofIBS
`is likely complex and might be present throughout the GI tract;
`however, colonic sensory afferents appear key to the development
`and persistence of IBS-related abdominal pain (22,23).
`Two delayed-release (DR) formulations of linaclotide are
`under development to target more distal regions of the GI tract,
`possibly leading to differential effects on abdominal pain and
`bowel function. Delayed-release formulation (DR) 1 was
`designed to release linaclotide in the ileum to enhance pain relief
`and preserve the drug's secretory effects, whereas DR2 (hereafter
`called MD-7246) was designed to release linaclotide more distally
`in the ileocecal junction and cecum, to relieve pain originating in
`the colon and minimize secretory effects in the small bowel (24).
`The objective of this study was to evaluate the safety, efficacy, and
`dose response of these 2 DR formulations when administered
`orally to patients with IBS-C.
`
`METHODS
`Study design
`A 12-week, randomized, double-blind, double-dummy, placebo(cid:173)
`controlled, parallel-group phase 2b clinical trial was conducted at
`71 study centers in the United States between October 2015 and
`September 2016. Patients discontinued prohibited medications
`for 2: 14 days (24 hours for laxatives) during the screening period.
`Eligible patients entered the pretreatment period and provided
`daily and weekly symptom assessments with an electronic diary
`(eDiary). Eligible patients were randomized to placebo or 1 of 7
`once-daily linaclotide doses (DRl 30, 100, and 300 µg; MD-7246
`30, 100, and 300 µg; or IR 290 µg [positive control]) for 12 weeks.
`Table 1 (see Supplementary Digital Content 1, http://links.lww.
`com/ AJG/B699) reports randomization details. Study visits oc(cid:173)
`curred every 4 weeks (Figure 1).
`The trial was designed, conducted, and reported in compliance
`with ethical principles set forth by the Declaration of Helsinki and
`Good Clinical Practice guidelines. The research protocol was ap(cid:173)
`proved by an institutional review board at each study site, and the
`trial was registered with ClinicalTrials.gov (NCT02559206).
`
`Patients
`Inclusion criteria were as follows: patients aged 18 years or older
`and who met the Rome III criteria for IBS-C ( 6); patients who had
`hard/lumpy stools with 2: 25% of bowel movements (BMs) and
`loose (mushy)/watery stools with < 25% of BMs without antidi(cid:173)
`arrheal medications or laxatives, and experienced < 3 BMs per
`week for the 2: 12 weeks during which the IBS-C diagnosis was
`established; and patients who had average worst abdominal pain
`2: 3.0 (11-point scale; 0 = none, 10 = worst possible), '.S l0
`spontaneous BMs (SBMs; BMs occurring without laxative, sup(cid:173)
`pository, or enema use), and '.S 6 complete SBMs (CSBMs; SBMs
`associated with a sensation of complete evacuation) during the 14
`days before treatment.
`Patients were excluded if they reported loose/watery stools
`without laxatives for > 25% of BMs in the 12 weeks before
`screening; had a Bristol Stool Form Scale score of7 with any SBM
`during the 14 days before randomization; and used rescue therapy
`on the day before or day of randomization. Other key exclusion
`criteria were structural alterations or conditions that could
`
`impact GI motility and a history of chronic conditions that could
`be associated with abdominal pain, discomfort, or constipation.
`
`Efficacy assessments and endpoints
`Patients recorded the number of BMs and severity of abdominal
`pain, bloating, and discomfort (11-point scales; 0 = none, 10 =
`worst possible) daily using eDiary. For each BM, stool consistency
`(7-point Bristol Stool Form Scale), sensation of complete evacu(cid:173)
`ation (yes/no), and use of rescue therapy were recorded.
`Key efficacy endpoints. Key efficacy endpoints were weekly change
`from baseline ( CFB) in abdominal pain (primary endpoint), weekly
`CFB in CSBM frequency, and a combined abdominal pain and
`constipation (APC) response. Weekly abdominal pain scores were
`calculated by averaging daily scores during each week. Weekly
`CSBM frequency was the total number of CSBMs each week. A
`patient was an APC + 1 responder if they met the following weekly
`criteria for 2: 6 of 12 weeks of treatment: (i) a reduction of 2: 30% in
`the average abdominal pain score compared with the baseline av(cid:173)
`erage; and (ii) an increase from baseline of 2: 1 in the CSBM weekly
`rate for that week (18,19). A patient was an APC+ 1 sustained
`responder if they met the responder definition and were a weekly
`responder for 2: 2 of the last 4 weeks of treatment.
`Additional efficacy endpoints. Additional endpoints included
`12-week CFB in abdominal discomfort and bloating, and abdominal
`pain and adequate relief responder rates. The 12-week CFB
`values were defined as the average of nonmissing values over
`the treatment period minus the baseline value. A patient was a
`6/12-week abdominal pain responder if they met the weekly
`abdominal pain responder criteria for 2:6 weeks of the 12-week
`treatment period; the more stringent 9/12-week abdominal pain
`responder definition required 2: 9 weeks. Table 2 (see Supplementary
`Digital Content 2, http://links.lww.com/AJG/B700) reports addi(cid:173)
`tional endpoints.
`
`Safety assessments
`Adverse events (AEs) and serious AEs (SAEs) were documented
`at each study visit. The investigator at each study site assessed
`the severity and causal relationship of AEs to study drug.
`Physical examinations, vital sign measurements, and standard
`clinical laboratory tests were also performed.
`
`Statistical analysis
`The sample size was determined based on the results of the pre(cid:173)
`vious linaclotide IR study (18,19), assuming higher doses of each
`DR formulation would show CFB in abdominal pain similar to
`historical values for linaclotide IR and lower doses would show
`CFB approximately 50% of previous treatment differences. Under
`these assumptions, 65 patients per treatment group (520 total
`patients) would have 81 % power based on a 2-sided linear trend
`test within each formulation at a type I error of 0.05.
`Efficacy analyses were based on an intent-to-treat population
`(i.e., randomized patients who received 2: 1 dose of study drug) .
`No adjustments were made for multiplicity. NominalPvalues for
`pairwise comparisons are provided for descriptive purposes only.
`For key efficacy endpoints, weekly least-squares (LS) mean CFB
`values are graphed. Overall treatment effects were evaluated using
`a mixed model repeated measures framework, with fixed effects
`for the baseline value of the specified measure, week, treatment
`group, geographic region, and week-by-treatment group. The
`overall dose response within each DR formulation was assessed
`with an overall trend test (i.e., linear contrast). Responder rates
`
`© 2020 by The American College of Gastroentero logy
`
`The American Journal of GASTROENTEROLOGY
`
`Copyright © 2020 by The American College ofGastroenterology. Unauthorized reproduction of this article is prohibited.
`
`Bausch Health Ireland Exhibit 2059, Page 2 of 8
`Mylan v. Bausch Health Ireland - IPR2022-00722
`
`

`

`356
`
`Chey et al.
`
`(/)
`
`0::: w
`0
`0:::
`0
`(/)
`0
`
`(9
`_J
`
`<( z
`0
`1-u
`z
`=>
`LL
`
`Screening
`period
`
`Pretreatment
`period
`
`Up to 21 days
`
`14 to 21 days
`
`Treatment
`period
`
`Day 1 to Day 85
`
`Week4visit
`(Day 29 ± 3)
`
`Week8visit
`(Day 57 ± 3)
`
`Screening
`visit
`(Day-42 to
`Day -1 5)
`
`Pretreatment
`visit
`(Day - 21 to
`Day -1 4)
`
`Randomization
`visit
`(Day 1)
`
`Week 12/
`end-of-treatment
`period visit
`(Day 85 ± 3)
`
`No t reatment
`
`30 •9 LIN DR1 tablet and PBO capsu le
`
`100 •9 LIN DR1 tablet and PBO capsule
`
`300 •9 LIN DR1 tablet and PBO capsule
`
`30 •9 MD-7246 tablet and PBO capsule
`
`100 •9 MD-7246 tablet and PBO capsu le
`
`300 •9 MD-7246 tablet and PBO capsu le
`
`290 •9 LIN IR capsule and PBO capsule
`
`PBO tablet and PBO capsule
`
`Figure 1. Stud y design. DRl , delayed-release formu lation l ; IR , immediate-release fo rmu lation ; LIN , linac lotide; PBO , placebo.
`
`were compared between treatment groups using the Cochran(cid:173)
`Mantel-Haenszel test, controlling for geographic region; dose
`response was assessed with Cochran-Mantel-Haenszel correla(cid:173)
`tion statistics. Additional 12-week CFB endpoints were evaluated
`over the entire treatment period using an analysis of covariance
`model with fixed effects for treatment group, geographic region,
`and baseline value of the specific measure. Pairwise comparisons
`between each linaclotide dose and placebo were performed for all
`analyses.
`
`RESULTS
`Patient disposition, demographics, and baseline characteristics
`Of 989 patients screened, 532 were randomized and received 2: 1
`dose of study drug (intent-to-treat population), and 448 (84.2%)
`completed
`the 12-week treatment period (see Figure 1,
`Supplementary Digital Content 3, http:/ /links.lww.com/ AJG/
`B701). The mean age was 45.1 years, and mean body mass index
`was 30.0 kg/m 2
`• Patients were mostly women (83.3%) and White
`(64.7%). Treatment groups were generally balanced about de(cid:173)
`mographics and baseline clinical characteristics (Table 1).
`
`IR group (LS mean difference from placebo: 0.992 [95% CI:
`0.228-1.756]), followed by the DRl 300-µg group (0.662 [95%
`CI: -0.104 to 1.428]) (Figure 3). The changes in CSBM frequency
`seemed to show a dose-response trend across the DRl doses
`( trend test, P = 0.07). By contrast, CSBM frequencies were similar
`between the 3 MD-7246 dose groups and placebo over the
`treatment period, and no dose response was observed ( trend test,
`P = 0.42).
`APC+ 1 responder rates. In the linaclotide DRl dose groups, the
`percentages of patients who were APC + 1 responders were higher
`than those in the placebo group (odds ratio [OR] [95% CI]: 1.39
`[0.61-3.17], 1.27 [0.57-2.85], and 2.39 [1.10-5.19] for DRl 30 µg,
`100 µg, and 300 µg, respectively), and a dose response was ob(cid:173)
`served (correlation test, P < 0.03) (see Table 2, Supplementary
`Digital Content 2, http://links.lww.com/AJG/B700). For linaclo(cid:173)
`tide IR, the OR (95% CI) vs of placebo group was 1.71 (0.79-3.70).
`APC+ 1 responder rates were similar between the 3 MD-7246
`groups and placebo group (ORs ranging from 0.89 to 1.13), with
`no observed dose response (correlation test, P = 0.86). APC+ 1
`sustained responder rates showed similar trends across all doses.
`
`Key efficacy endpoints
`CFB in abdominal pain. All linaclotide DRl and MD-7246 groups
`experienced improvements from baseline in abdominal pain
`throughout the treatment period (Figure 2). Among linaclotide
`DRl doses, a dose response was observed for CFB in abdominal
`pain (trend test, P < 0.03), with the 300-µg group showing the
`greatest improvement over the entire treatment period (LS
`mean difference from placebo [95% confidence interval (CI)] :
`-0.771 [-l.419to -0.123]). TheLSmeandifferencefromplacebo
`was -0.569 (95% CI: -1.214 to 0.076) for linaclotide IR and
`ranged from -0.455 to -0.258 for the 2 lower DRl doses and all
`MD-7246 doses, with no dose response seen across MD-7246 doses
`(trend test, P = 0.55).
`CFB in CSBM frequency. Over the entire treatment period, the
`greatest increases in CSBM frequency were seen in the linaclotide
`
`Additional endpoints
`The results for additional abdominal and bowel symptom end(cid:173)
`points showed trends across the doses of the 2 formulations that
`were similar to the trends seen with the primary abdominal and
`bowel symptom efficacy endpoints (see Table 2, Supplementary
`Digital Content 2, http://links.lww.com/AJG/B700). Abdominal
`pain response rates were highest for the linaclotide DRl 300-µg
`group, followed by the linaclotide DRl 30-µg and MD-7246
`300-µg groups. For the most stringent abdominal pain responder
`threshold (i.e., 9/12-week abdominal pain responder), ORs (95%
`Cis) vs placebo group were 2.49 (1.14-5.43) for DRl 300 µg, 1.84
`(0.83-4.07) for DRl 30 µg, 1.62 (0.75-3.54) for linaclotide IR, and
`1.45 (0.65-3.25) for MD-7246 300 µg. For the less stringent
`threshold (i.e., 6/12-week abdominal pain responder), ORs vs
`placebo group ranged from 0.85 to 1.11 for all doses, except DRl
`
`The American Journal of GASTROENTEROLOGY
`
`VOLUME 116 I FEBRUARY 20 2 1 www.amjgastro.com
`
`Copyright © 2020 by The American College ofGastroenterology. Unauthorized reproduction of this article is prohibited.
`
`Bausch Health Ireland Exhibit 2059, Page 3 of 8
`Mylan v. Bausch Health Ireland - IPR2022-00722
`
`

`

`Delayed-Release Formulations of Linaclotide
`
`357
`
`Table 1. Summary of patient demographic and baseline clinical characteristics
`
`Demographics
`Age, yr, mean (SD)
`BMI, kg/m 2 , mean (SD)
`Women , n (%)
`Race, n (%)
`
`White
`Other
`Hispanic or Latino
`Clinical cha racteristics,
`mean (SD)
`SBMs/wk
`CSBMs/wk
`Stool consistency (BSFS)3
`Abdomina l pain
`Abdominal discomfort
`Abdomina l bloating
`
`PBO
`(n = 66)
`
`LIN IR
`290µg
`(n = 66)
`
`LIN DRl
`30µg
`(n = 67)
`
`LIN DRl
`100 µg
`(n = 67)
`
`LIN DRl
`300µg
`(n = 67)
`
`MD-7246
`30µg
`(n = 67)
`
`MD-7246
`100 µg
`(n = 66)
`
`MD-7246
`300µg
`(n = 66)
`
`45.4 (14.7)
`30.1 (7.1)
`53 (80.3)
`
`44.1 (14.4)
`
`29.7 (5.9)
`53 (80.3)
`
`44.8 (14.9)
`29.7 (7.9)
`59 (88.1 )
`
`44.7 (13.7)
`29.3 (7.5)
`59 (88. 1)
`
`46.5 (12.7)
`30.4 (7 5)
`55 (82.1)
`
`42.3 (12.6)
`29.5 (69)
`50 (74.6)
`
`47.9 (12.8)
`29.5 (65)
`52 (78.8)
`
`45.2 (14.4)
`
`31.5 (7.5)
`62 (93.9)
`
`38 (57.6)
`28 (42.4)
`10 (15.2)
`
`43 (65.2)
`23 (348)
`14 (21.2)
`
`45 (67.2)
`22 (32.8)
`16 (23.9)
`
`47 (70. 1)
`20 (29 9)
`17 (25.4)
`
`41 (61.2)
`
`26 (38.8)
`16 (23.9)
`
`42 (62.7)
`
`25 (37.3)
`18 (26.9)
`
`40 (60.6)
`26 (39.4)
`15 (22.7)
`
`48 (72.7)
`
`18 (27.3)
`21 (31.8)
`
`1.66 (1 .26)
`0.29 (0.56)
`2.25 (1 .13)
`6.41 (1.85)
`6.58 (1.72)
`6.67 (2.16)
`
`1.78 (1.24)
`
`1.62 (1.16)
`
`0.34 (0.61)
`2.21 (0.99)
`6.18 (1.77)
`
`6.29 (1.67)
`6.40 (1.91 )
`
`0.35 (0.56)
`2.35 (1.13)
`6.20 (1.59)
`6.45 (1.38)
`6.67 (1.41)
`
`1.49 (1 .09)
`0.22 (0.46)
`2.28 (0.90)
`6.18 (1.67)
`6.34 (1 .64)
`6.52 (1.74)
`
`1.73 (1.24)
`
`1.70 (1.04)
`
`0.27 (0.54)
`2. 17 (1.06)
`6.38 (1.82)
`6.70 (1.68)
`699 (1.55)
`
`0.35 (062)
`2.30 (1.01)
`6.07 (168)
`6.35 (1.61)
`6.52 (162)
`
`1.60 (1.14)
`0.31 (061)
`2.35 (0.90)
`6.48 (1 53)
`6.67 (1.51)
`6.91 (1 57)
`
`1.65 (1 .19)
`0.33 (0.56)
`2.18 (0.93)
`6.09 (1.68)
`6.36 (1 .61)
`6.56 (1.66)
`
`BMI , body mass index; BSFS, Bristol Stool Form Scale; CSBM, complete spontaneous bowel movement; DRl, delayed-release formulation l ; IR, immediate-release
`formulation; LIN, linaclotide; PBO, placebo; SBM, spontaneous bowel movement.
`•samplesizesforstoolconsistencywereasfollows: n = 59(placebo), n = 62 (LIN IR 290 µg), n = 62 (LIN DRl 30 µg), n = 56(LIN DRl lOOµg), n = 61 (LIN DRl 300 µg),
`n = 61 (M D-7246 30 µg), n = 58 (MD-7246 100 µg), and n = 55 (M D-7246 300 µg).
`
`(/)
`
`0:::: w
`0
`0::::
`0
`(/)
`0
`
`(9
`_J
`<(
`z
`0
`I-u
`z
`=:)
`LL
`
`300 µg (OR [95% CI]: 2.03 [0.99-4.16]) andDRl 30 µg (OR [95%
`CI]: 1.44 [0.70-2.98]).
`
`Safety
`No patients from the linaclotide DRl or MD-7246 groups ex(cid:173)
`perienced SAEs, and no deaths occurred during the study.
`Treatment-emergent AEs (TEAEs) experienced by 2: 2% of pa(cid:173)
`tients during the treatment period are reported in Table 3 (see
`Supplementary Digital Content 4, http:/ /links.lww.com/ AJG/
`B702). In general, the frequency of TEAEs was similar between
`the linaclotide DRl and IR groups, and the frequency of TEAEs
`for the MD-7246 group was similar to the placebo group.
`Diarrhea was the most common TEAE in the linaclotide DRl
`and IR groups (Figure 4). Diarrhea was reported by 2 (3.0%), 5
`(7.5%), and 7 (10.4%) patients in the linaclotide DRl 30-µg, 100-µg,
`and 300-µg groups, respectively, and 9 patients (13.6%) in the IR
`group. The frequency of diarrhea was generally lower in the MD-
`7246 and placebo groups, having been reported by none, 1 (1.5%),
`and 2 (3.0%) patients in the MD-7246 30-µg, 100-µg, and 300-µg
`groups, respectively, and 1 patient (1.5%) in the placebo group. Two
`patients each (3.0%) in the DRl 100-µg and 300-µg groups and 4
`patients (6.1 %) in the IR group discontinued the study drug due to
`diarrhea (Figure 4). Two other TEAEs led to study drug discontin(cid:173)
`uation in the D Rl and MD-7246 groups: headache (D Rl 30 µg [ n =
`l] and DRl 100 µg [n = l]) and nausea (DRl 100 µg [n = l] and
`MD-7246 100 µg [n = l]). Additional TEAEs leading to
`discontinuation in the IR group included abdominal pain, defecation
`urgency, and rectal hemorrhage (for each, n = 1).
`
`One patient in the placebo group experienced SAEs of pneu(cid:173)
`monia and sepsis, and 1 patient in the linaclotide IR group ex(cid:173)
`perienced gastroenteritis. These SAEs were considered unrelated
`to the study drug by the study investigator, and all SAEs resolved
`without study drug discontinuation. There were no clinically
`meaningful differences between treatment groups in the in(cid:173)
`cidence of abnormal vital signs or laboratory parameters.
`
`DISCUSSION
`Physicians and patients alike recognize the need for effective
`therapies to manage the APC symptoms of IBS-C. The 2018
`American College of Gastroenterology Monograph on the man(cid:173)
`agement of IBS recognized only 3 evidence-based treatment op(cid:173)
`tions for patients with IBS-C (8). However, no available
`pharmacotherapies for IBS are recognized as pure "visceral anal(cid:173)
`gesics," h aving isolated benefit for abdominal pain without a dis(cid:173)
`cernable impact on bowel habits.
`Linaclotide's analgesic and secretory effects and its modulation
`of bowel function might be mediated by 2 distinct pathways, each
`initiated by GC-C agonism within the GI tract. Distinct pathways
`would suggest the possibility of reformulating linaclotide to un(cid:173)
`couple the pain benefit from its bowel effects. Net flux of water into
`the luminal bowel, in response to linaclotide' s activation of GC-C,
`is greatest in the more proximal portions of the small bowel (e.g.,
`duodenum and jejunum) (13). Two DR formulations oflinaclotide,
`DRl and MD-7246, were developed with the intention of exploring
`the possibility of differential effects oflinaclotide on secretion and
`pain modulation based on delivery to more distal segments of the
`bowel. It was hypothesized that by bypassing GC-C receptors in the
`
`© 2020 by Th e Ameri can College of Gastroenterology
`
`The American Journal of GASTROENTEROLOGY
`
`Copyright © 2020 by The American College ofGastroenterology. Unauthorized reproduction of this article is prohibited.
`
`Bausch Health Ireland Exhibit 2059, Page 4 of 8
`Mylan v. Bausch Health Ireland - IPR2022-00722
`
`

`

`358
`
`Chey et al.
`
`(/)
`
`0::: w
`0
`0:::
`0
`(/)
`0
`
`(9
`_J
`
`<( z
`0
`I-u
`z
`=>
`LL
`
`a
`
`-"
`
`b
`
`-0- PBO
`-
`UN IR290µg
`-L J- LIN DR1 30 µg
`-◊- LIN DR1 100 µg
`- b.- U N DA1 300 µg
`
`LS mean difference
`vs. PBO (95% Cl)
`
`--0.569(-1 .214, 0.076)
`--0.297(--0.942, 0.348)
`--0.286(--0.936, 0.363)
`--0.771 (-1.419,--0.123)
`
`-0- PBO
`-+- LI N IA 290 µg
`-■- MD-7246 30 µg
`-♦- MD-7246 100 µg
`-4- MD-7246300µ.g
`
`LS mean difference
`vs. PBO (95% Cl)
`
`--0.569(-1.214, 0.076)
`--0.455(-1 .102,0.191)
`--0.299(--0.947, 0.348)
`--0.258(--0.905,0.389)
`
`\
`
`--,
`' .._,.
`'-.'\
`' ~ ...
`'\ --<t-...
`',
`'-"':t-::..-==•
`'•---
`:
`....... _____ -■ .............
`
`Time(weeks)
`
`Time (weeks)
`
`Figure 2. Weekly CFB in abdom inal pa in. Treatment effects of (a) LIN DRl 30 µ,g, 100 µ,g, and 300 µ,g and (bl MD-7246 30 µ,g, 100 µ,g, and 300 µ,g are
`shown with respect to LIN 290 µ,g IR and PBO , with 12-week LS mean differences vs PBO reported in each legend. All 8 treatmentgroups were eva luated in
`a single model, wherein treatment group was inc luded as a fixed effect. CFB, change from baseline; Cl, confidence interva l; DRl, delayed-release
`formu lation l ; IR , immediate-release fo rmu lation; LI N, linac lotide; LS, least-squa res; PBO, placebo.
`
`proximal small bowel, these novel DR formulations of linaclotide
`might have less fluid secretion and, thus, a limited or no effect on
`bowel habits. At the same time, these DR preparations were
`expected to retain antinociceptive effects through modulation of
`sensory afferent pain signaling originating in the colon.
`
`Indeed, this hypothesis seems to have been confirmed in this
`phase 2b trial in patients with IBS-C. DRl, targeting the ileum,
`demonstrated similar results at the 300-µg dose when compared
`with the linaclotide IR formulation for each of the 3 key efficacy
`endpoints (i.e., weekly CFB in abdominal pain and CSBM
`
`a
`
`-0- PBO
`-+- LIN IA290µg
`-Ll- LIN DR1 30 µ.g
`-◊- LIN DR1 100 µg
`LIN DA1 300 µ.g
`- t:, -
`
`LS mean difference
`vs. PBO (95% Cl)
`
`0.992 (0.228, 1.756)
`0.048(--0.716, 0.812)
`0.299(-0.469, 1.067)
`0.662(-0.1 04, 1.428)
`
`b
`
`-0- PBO
`....... LIN IA 290 µg
`-■- MD-7246 30 µg
`-♦- MD-7246100 µg
`-A- MD-7246 300 µg
`
`LS mean difference
`vs. PBO (95% Cl)
`
`0.992(0.228, 1.756)
`0.161 (-0.604, 0.925)
`-0.095 {-0.861, 0.670)
`-0.246(-1.011, 0.519)
`
`nme{weeks)
`
`nme(weeks)
`
`Figure 3. Weekly CFB in CSBM frequency. Treatment effects of (a) LI N DR 130 µ,g, 100 µ,g, and 300 µ,g and (bl MD-7246 30 µ,g, 100 µ,g, and 300 µ,g are
`shown with respect to LIN 290 µ,g IR and PBO, with 12-week LS mean differences vs PBO reported in each legend .Al l 8treatmentgroupswereeva luated in a
`single model, wherein treatment group was included as a fixed effect. CFB , change from baseline; Cl, confidence inte rva l; CSBM , complete spontaneous
`bowel movement; DRl, delayed-release formu lation l; IR , immediate-release fo rmu lation; LIN , linaclotide; LS, least-squares; PB O, placebo.
`
`The American Journal of GASTROENTEROLOGY
`
`VOLUME 116 I FEBRUARY 2021 www.amjgastro.com
`
`Copyright © 2020 by The American College ofGastroenterology. Unauthorized reproduction of this article is prohibited.
`
`Bausch Health Ireland Exhibit 2059, Page 5 of 8
`Mylan v. Bausch Health Ireland - IPR2022-00722
`
`

`

`Delayed-Release Formulations of Linaclotide
`
`359
`
`(/)
`
`0:::: w
`0
`0::::
`0
`(/)
`0
`
`(9
`_J
`
`<( z
`0
`1-u
`z
`=:)
`LL
`
`16
`
`14
`
`12
`
`10
`
`~
`,,;
`5l
`~
`<l.
`
`13.6
`
`□ Diarrhea TEAE leading to discontinuation
`
`10.4
`
`7.5
`
`6.1
`
`3.0
`
`3.0
`
`Placebo
`(n = 66)
`
`IR 290 µg
`(n = 66)
`
`30 µg
`(n = 67)
`
`100 µg
`(n = 67)
`
`DR1
`
`300 µg
`(n = 67)
`
`30 µg
`(n = 67)
`
`MD-7246
`
`Figure 4. Pe rcentage of patients with diarrhea TEA Es, including those leading to discontinuation. The proportion of patients with dia rrhea TEA Es leading to
`discontinuation (grey) is shown as a subset of the total popu lation with ::::: 1 diarrhea TEAE in each treatment group. DR 1, delayed-release fo rmu lation 1; IR,
`immed iate-release formu lation ; TEAE, treatment-eme rgent adverse event.
`
`frequency, and APC+ 1 responder rate). Furthermore, linaclotide
`DRl led to greater improvements compared with placebo for the
`additional endpoints of abdominal bloating, abdominal discom(cid:173)
`fort, abdominal pain responder rate, and adequate relief responder
`rate. MD-7246 demonstrated a signal for relief of abdominal pain
`compared with placebo, with absolute reductions in abdominal
`pain from baseline (a key endpoint) in all 3 dose groups over the
`treatment period. The anatomic origin of abdominal pain in IBS
`remains uncertain. Several studies have clearly demonstrated
`greater rectosigmoid sensitivity in patients with IBS compared with
`healthy volunteers using barostat or balloon distention protocols
`(25,26). Yet, considerable variation in perception can be discerned
`among patients with IBS-C (27). Furthermore, detectable hyper(cid:173)
`sensitivity of the rectosigmoid might be modulated by physiologic
`activities in the small intestine, such as the consumption of a meal
`(28). The similar levels of pain relief observed in this study with IR
`290 µg, DRl 300 µg, and MD-7246 suggest that GC-C activation
`and subsequent release of extracellular cyclic guanosine mono(cid:173)
`phosphate into the distal small bowel and colon might be sufficient
`to elicit the desired nociceptive effects of GC-C agonism. Trends
`seen for MD-7246 vs placebo for the key abdominal endpoint, plus
`those seen for the additional abdominal endpoints, support the
`antinociceptive properties expected for this formulation and sug(cid:173)
`gest that a portion oflinaclotide' s analgesic properties might also be
`mediated by GC-C action in the more proximal small intestine. At
`the same time, MD-7246 exerted little effect on BM frequency, with
`no dose response for CSBM frequency or APC+ 1 responder rates.
`Data from this study, with sample sizes appropriate for phase 2b
`exploration, suggest that linaclotide's treatment effects could in(cid:173)
`deed be modulated by targeting drug delivery to specific parts of the
`GI tract. With its pharmacologic effects limited to visceral anal(cid:173)
`gesia, MD-7246 presents an intriguing option for the management
`of nonconstipated subtypes of IBS.
`Both the linaclotide DRl and MD-7246 formulations were well
`tolerated. The TEAE profile for linaclotide DRl was consistent with
`the established safety profile for linaclotide IR 290 µg (18-20), with
`diarrhea being the most common AE. TEAEs, including diarrhea,
`were similar between the MD-7246 and placebo groups and were
`relatively lower than those in the linaclotide D Rl and IR groups. The
`diarrhea rate in the IR 290-µg group was lower than that seen in the
`pivotal phase 3 trials (19.5%-19.7% vs 2.5%-3.5% for placebo)
`
`(18,19). The reason for this is unknown, although a possible expla(cid:173)
`nation is that, unlike the previous studies, this study did not include a
`2-week study visit. No SAEs were reported in patients receiving
`linaclotide DRl or MD-7246.
`This study has several notable strengths, including the en(cid:173)
`rollment of a carefully phenotyped patient population who met
`the Rome III criteria for IBS-C. Patients with potentially impor(cid:173)
`tant clinical confounders (e.g., structural disease) were excluded.
`Endpoints examined in this study, including abdominal pain
`using an 11-point numerical rating scale, followed US Food and
`Drug Administration guidance for clinical trials in IBS (29).
`However, this study also has limitations. Although drug delivery
`to specified bowel segments was expected based on the in vitro
`release profiles of the DR formulations in biorelevant dissolution
`media, verification of drug delivery using sampling or imaging
`techniques was not performed in vivo. It is conceivable that the
`precise location oflinaclotide delivery with the DR formulations
`might have varied based on the intestinal bacterial and bio(cid:173)
`chemical milieu of individual patients and/or with different rates
`of intestinal transit. In addition, results for this study must be
`considered relative to the limited sample size (66-67 patients per
`treatment arm) and limited inferential statistical analyses, which
`focused on the primary endpoint and evaluation of dose trends.
`In conclusion, in this phase 2b study of2 novel linaclotide DR
`formulations, DRl seemed to have similar efficacy for im(cid:173)
`provements in bowel function and abdominal pain compared
`with the commercially available IR preparation. Importantly,
`MD-7246 maintained improvement in abdominal pain relief
`relative to placebo with little impact on bowel symptoms and
`very low rates of treatment-associated diarrhea. MD-7246
`should, therefore, be examined as an option for treating
`IBS-related abdominal pain without altering bowel habits. Fur(cid:173)
`ther MD-7246 studies across the spectrum of IBS subtypes, in(cid:173)
`cluding those with diarrhea predominance, should be considered
`based on these observations.
`
`CONFLICTS OF INTEREST
`Guarantor of the article: Wi