0002-9270/93/8808-1188$03.00/ 0
`Tur A MLRICA:-. JouR~AL o~ GASTRO[~(ROLOGY
`Copyright c:i 1993 by Am. Coll. of Gas1roen1erolog}
`
`Vol. 88. No. 8. 1!19J
`Pnnted in U.S.A
`
`Mesalamine Capsules for Treatment of Acti':e Ulcerative
`Colitis: Results of a Controlled Trial
`
`Stephen Hanauer, M.D., Jerrold Schwartz, M.D., Malcolm Robinson, l\LD., Walter Roufail, M.D.,
`Sanjeev Arora, M.D., John Cello, M.D., Michael Safdi, M.D., and the Penta a ·
`tudy Group
`University of Chicago School of Medicine. Chicago. Illinois: No:thwest G_a.~troent~rolnxist;~· Arli11f(to11 lle1ghts. ~llino'.s, HCA
`Presbyterian Hospital, Oklahoma City, Oklahoma; Forsyth Medical Special/S/s. M 111.Hm!·Salc;m. North Ca~olma. Neu England
`Medical Center Boston MassachuseltS: San Francisco General llospual. San /ora11c1sco. Ca/ifomw:
`' Great~r Cincinnati Gastroenterologr Associates. Ci11ci1111ati. Ohw
`
`The efficacy of a capsule formulation of mesalamine
`was assessed in 374 patients with mild to moderately
`active ulcerative colitis. Patients, stratified to pancolitis
`or left-sided disease, received either placebo or mesa(cid:173)
`lamine at 1, 2, or 4 g daily for 8 wk. Efficacy was
`assessed using clinical improvement-physician global
`assessment, sigmoidoscopic index, biopsy score, trips
`to the toilet, and clinical symptoms (abdominal pain,
`urgency, stool consistency, and rectal bleeding)-and
`induction of remission (more stringent criteria for phy(cid:173)
`sician global assessment, sigmoidoscopic index, and
`biopsy score). For physician global assessment of treat(cid:173)
`ment benefit, 79% and 84% of patients on the 2-g and
`4-g doses of mesalamine, respectively, received treat(cid:173)
`ment benefit, compared with 54% on placebo (p s
`0.0002). For the physician global assessment of treat(cid:173)
`ment success, both the 2-g and 4-g doses of mesalamine
`were superior to placebo (57% and 59% of patients,
`p = 0.0021 and 0.0012, respectively), compared with
`36% on placebo. Both the 2-g and 4-g doses produced
`statistically significant macroscopic (endoscopic) im(cid:173)
`provement compared with placebo (p < 0.004). The 4-
`g dose also produced a statistically significant micro(cid:173)
`scopic (histologic) improvement compared to placebo
`(p < 0.002). Significant improvement compared to pla(cid:173)
`cebo was also observed at 2 g and 4 g for the four
`clinical symptoms and trips to the toilet (p s 0.003).
`Oral mesalamine capsules were significantly superior
`to placebo for inducing remjssion, with 29% of patients
`at 2 g and 29% at 4 g achieving remission by physician
`global assessment, compared with 12% on placebo.
`Forty-four percent and 48% of patients receiving 2 g
`and 4 g of mesalamine, respectively, achieved remission
`by sigmoidoscopic index (p < 0.05), compared with
`31 % on placebo. Thirty-nine percent of patients at 4 g
`daily achieved microscopic remission, compared with
`23% on placebo (p < 0.03). Treatment response was
`not affected by extent of disease or prior steroid or
`
`Received Sep/. 17. 1992; accep1ed Feb. 15. 1993.
`
`1188
`
`sulfasalazine therapy. The c data uggest that con(cid:173)
`trolled-release me ·ala mine cap ulcs are a afe and ef·
`fective monotherapy in do e of 2-4 g daily for treating
`mild to moderately active ulcerathe colitis, as well as
`for inducing remission, regardles of prior oral steroid
`or sulfasalazine therapy or extent of di ease.
`
`INTRODUCTION
`
`Since its in trod uction over 40 yr ago, sulfasalazine
`has become the most widely prescribed medication for
`the treatment of inflammatory bowel disease. specifi(cid:173)
`cally ulcerative colitis (UC). Oral sulfasalazine has been
`demonstrated (usi ng various criteria) to be effective for
`the treatment of mild-to-moderate UC and for the
`maintenance of remission ( I). When administered in
`oral doses of 4-6 g dail y for up to I month. oral
`sulfasalazine has been reponed to induce remission in
`35-803 of patients, depending on duration of therapy.
`dose, and the endpoints used to determine.efficacy (1-
`5). This is twice the spontaneous remission rate ob(cid:173)
`served in the placebo groups (1 - 3).
`Just as there is a dose-dependent response rate, there
`is also a dose-dependent rate of adverse effects which
`limit the acute and chronic use of sulfasalazine. Com(cid:173)
`mon adverse efTects include nausea and headache (I. 6.
`7). Historically (in 20 UC trials) adverse effects ba\e
`been noted in 13-603 of patients (4). Sperm abnor·
`malities, in panicular. can be documented in 80% of
`males taking sulfasalazine ( I, 8- 15). Less common and
`more severe adverse efTects include hemolytic anemia.
`neutropenia. hepatitis, pancreatitis. and worsening of
`colitis (I). Most of the serious complications, except for
`rare cases of pancreatitits, pericarditis, or worsening ~f
`colitis (acute intolerance), are related to the sulfapyn·
`dine moiety. All except for the last complication, wol)(cid:173)
`ening of colitis, are side efTects that have been related
`to the sulfapyridinc moiety.
`Sulfasalazine is composed of sulfapyridine and mes·
`alamine (5-aminosalicylic acid) linked by an azo bond.
`
`Mylan Exhibit 1004
`
`

`
`August 1993
`
`M ESALA MI NE CA PSULES FOR TRE ATMENT OF ACUTE UC
`
`11 89
`
`This bond is metabolized by colo n ic bacteria, causing
`the release of the mesalamine and sul fapyridine m ole(cid:173)
`cules (16- 18). In 1977, Azad Kha n et al. (l6) de m o n(cid:173)
`strated definitively that m esala m ine was a thera pe uti(cid:173)
`cally active moiety of sulfasalazine. T h is discovery led
`to the development of mesalamine fo rmulatio ns or
`delivery systems used for both UC a nd Crohn 's d isease.
`These delivery system s include a n e ne m a formula(cid:173)
`tion that has been shown to be effective in the treatment
`of distal UC ( 19). Although effec tive in the treatment
`of active distal disease, the e nema form ula tion of m es(cid:173)
`alamine does not bene fit m ost patie nts with di sease
`proximal to the sple nic flexure (20). Oral adm in istra(cid:173)
`tion of mesalamine is lim ited by a bso rpt ion in the
`proximal small bowel. necessitating p ro tected de li very
`systems for distribution to distal sites o f enterocolonic
`inflammation (2 1 ).
`In general, two types of agent have e vo lved to give
`delayed. targeted delivery o f m esalami ne to the sm aJJ
`bowel and/ or colon: azo-bond prod r ugs a nd no n linked
`agents. Prodrugs, such as olsalazine ( D ipcntum ), join
`two mesalamine mo lecules with an az o bond a nd re(cid:173)
`quire azo-red uctase to release active drug in the colon .
`To date, two types o f no nlinkcd age nts have evolved :
`I) delayed-release form ulatio ns consist of ac rylic-based
`resin-coated rnesa larn inc that dissolves a t a pH greater
`than 7, releasi ng active drug in the d istal ileum and
`colon (20, 21); and 2) a slow-release. capsule formula(cid:173)
`tion encapsulates m esala m ine in ethylcellulose-coated
`beads and delivers active d rug co ntinuo usly to the sm a ll
`and large bowel. indepe nde nt of in testina l pH (22, 23).
`This trial was developed thro ugh the colla bo rative
`efforts of an internationa l group o f cl inic ians to p rovide
`clinically appropriate criteria for s ubjective a nd obje~­
`tive assessment of m esala m ine e ffi cacy a nd safety m
`UC. This study assessed a controlled- release, capsule
`formulation of oral rnesala mine (Pe ntasa, M a rio n Mer(cid:173)
`rell Dow Inc., Kansas City, M O) for t he treatment of
`and induction of rem issio n in patients with mildly to
`moderately active UC.
`
`MATERIALS AND M ETHODS
`
`Each study site received individua l Institutional. R e(cid:173)
`view Board approval, a nd a ll pa tients gave wntten
`informed consent.
`
`Patient selection
`Patients over the age of 18 yr, with mildly to. m~d­
`erately active UC, were selected for entry i nto th1s.tnal
`between March 6. 1987, a nd August 4, 1988. Pauents
`were included if they had a diagnosis o f UC a n? .the
`presence of active disease confirmed by both climcal
`symptoms and colo noscopic evidence of active in~am­
`mation of 5 or m ore on a 15-point index scale. Paue~ts
`were stratified based on exte nt of d isease. Patients WJth
`
`disease distal to the splenic flexure were considered to
`have le ft-sided involvem ent, whereas those with disease
`proxima l to the splenic flexure were classified as having
`pancolitis.
`Continua tion o f therapy with steroids, sulfasalazine,
`or other m esalamine formulations was not allowed
`du ring participatio n in this trial. Prior use of steroid.
`sulfasalazine, or other mesalamine formulations re(cid:173)
`quired a 7-day "washout" prior to baseline evaluations.
`Prior use o f an immunosuppressive agent required a
`90-day washout. Positive stool culture for enteric path(cid:173)
`ogens, the presence of fecal ova and parasites, or Clos(cid:173)
`tridium difficile toxin were also exclusio n criteria. In
`addition , m edications known to have a therapeutic
`effect (sulfasalazine, topical m esalamine, or any corti(cid:173)
`costeroids). mask symptoms (antispasm odics, antidi(cid:173)
`arrheals except loperamide), cha nge absorptio n (choles(cid:173)
`tyramine), or possibly worsen the disease (antibiotics,
`N SA lDs) were prohibited during the trial. Female pa(cid:173)
`tients were required to be taking a medically prescribed
`form of birth control throughout the trial or to not be
`of child-bearing potential. Pregnant or lactating females
`were also excluded fro m participation.
`
`Study design
`T his was a multicenter, double-blind, placebo-con(cid:173)
`trolled , parallel, dose-response trial. At 20 study sites,
`patie nts were stra tified by disease location, pancolitis
`or left-sided disease, a nd were randomized to receive
`eithe r placebo or m esalamine at I, 2, or 4 g daily for 8
`wk.
`
`St udy drug administration
`Patients meeting the ad mission criteria were ran(cid:173)
`domly assigned to one of four treatment regimens:
`pla cebo or L g, 2 g, or 4 g of oral mesalat;ni~e da.ily.
`Study drug was supplied in 250-m g capsules.to 1de~t1cal
`bli ster cards to ensure blinding of both the mvestigator
`a nd the patient. Mesala mine I g daily was administered
`as one active and three p lacebo capsules, qid; mesalam(cid:173)
`ine 2 g daily as two active ~nd two place.bo capsules,
`qid; and m esala mine 4 g daily as four active capsules.
`qid. Placebo was administered as four ~lacebo capsules.
`qid . D osing was contin ued for the entire 8-w~ study.
`Lopera mide (2 m g) was dispe nsed to patients only
`when absolutely necessary for control of diarrhea.
`
`E valuation m ethods and scheduling
`Physician global assessm ent (P GA) is .an arbitr~~ly
`·gned multicomponent m easure of d isease act1v1ty
`es1
`,
`.
`d
`that uses the physician's assessment of 1m.proveme~t .or
`worsening in clinical status based on d1seas~ act1v1ty
`and symptom severity as .c~mpared to baseline. PGA
`was evaluated by the phys1c1a o at we~ks I, 4, a nd 8, ? r
`upon withdrawal, using six categon es. as shown m
`Ta ble I.
`
`

`
`11 90
`
`HANAUER et al.
`
`TABLE I
`Physician Global As~essment Scale
`
`I =Complete relief of symptoms
`2 =Marked improvement of symptoms
`3 =Moderate improvement of symptoms
`4 = Slight improvement of symptoms
`5 = No change in symptoms
`6 = Worseniag of symptoms
`
`A baseline colo noscopy documented the linear extent
`and severity of proctocolonic inflammation. Sigmoid(cid:173)
`oscopic or proctoscopic evaluations were also per(cid:173)
`formed at weeks I, 4, and 8, or upo n withdrawal from
`the study. All endoscopic evaluations were documented
`in the most severe area of disease involvement located
`5- 15 cm fro m the anal verge. T he evaluatio ns focused
`on the presence and severity of erythema, friability.
`granularity/ulceration, mucopus, and the appearance
`of mucosa! vascular pattern. Each of these fi ve cate(cid:173)
`gories was assigned a value fro m 0 to 3 (0 = normal, I
`= mild, 2 = moderate, and 3 = severe), and the five
`scores were totaled to produce an index score of 0- 15
`(with a maximum severity score of 15). Changes in
`sigmoidoscopic index were computed as the last avail(cid:173)
`able sigmoidoscopic score minus baseline.
`A treatment failure assessment was an additio nal
`efficacy parameter used not o nly to allow patients to
`withdraw from the trial if they were not receiving
`therapeutic benefit, but also to compare the percent of
`patients in each treatment group that were not receiving
`therapeutic benefit. This assessment was made to iden(cid:173)
`tify a worsening of disease at 7 days or greater, post
`baseline. A patient with an increase of 5 points in the
`sigmoidoscopic index, a worsening or no improvem ent
`in daily trips to the toilet, and a worsening o r no
`improvement in any symptoms would be classified as
`a "treatment failure" and could be removed from t he
`trial. In addition, patients could leave the st udy as
`treatment failures at or after 14 days if there was no
`improvement in the sigmoidoscopic index a nd patient
`sympto ms. Although treatment failure was utilized d ur(cid:173)
`ing the trial to allow patients to withdraw if they were
`not receiving therapeutic benefit, categorizatio n as a
`treatment failure was independent fro m withdrawal or
`reasons for withdrawal from the trial.
`Biopsies, taken at the area representative of disease
`activity fro m 5 ci:i to 15 cm from the anal verge, were
`collected at baseline and at week 8 or upon withdrawal
`(endpoint). Histological ~bang~s were identified by a
`central reference pathologist, usmg a 4-point categorical
`scale (0 "'.' n~rmal colonic mucosa through 3 = high
`grade, active mflammatory bowel disease).
`A baseline assessment of four clinical sympto ms and
`daily trips to the toilet (any attempt at a bowel move(cid:173)
`ment) was collected at the baseline visit through collab(cid:173)
`the patient.
`oration between the investigator and
`
`Vol. 88, No. 8, 1993
`
`Patients then recorded their impressions of symptom
`severity weekly throughout the trial. Symptoms, includ(cid:173)
`ing stool consistency. rectal bleeding, abdominal/ rectal
`pai n. and urgency, were recorded using a horizontal
`visual analog scale with a possible rating ofO (absence
`of sym ptoms) to I 0 (symptoms as severe as possible).
`Daily trips to the toilet were recorded by patients in a
`diary at baseline and throughout the trial. Mean dailv
`tri ps were computed for the patient at endpoint. using
`a weekly mean of the patient's daily trips to the toilet.
`Physical examinations and evaluation of adverse ef.
`feet reports were also conducted at each visit. Clinical
`laboratory assessments were collected at baseline and
`at endpoint.
`
`£.fficac~1 · parame1ers
`To establish effi cacy. two broad criteria were used:
`clinical improvemen t and induction of remission. Oin(cid:173)
`ical im provement was assessed using PGA, assessment
`of treatment failure, sigmoidoscopic index, biopsy
`score, patient perceptions of severity of the four clinical
`symptoms. and trips to the toilet. Ind uction of remis(cid:173)
`sio n was assessed by mo re stringent criteria for PGA.
`sigmoidoscopic index, a nd biopsy score.
`Clinical impmveinenl. For PGA. two assessments
`(treatment success a nd treatment benefit) were used.
`Treatment success was defined by an assessment at
`endpoint of tho e patients who had complete relief of
`sympto ms (category I) or marked improvement of
`sym ptoms (category 2) only. Treatment benefit was
`defined as a ny im provement at endpoint over baseline,
`including patien ts in categories I and 2. as well as those
`with moderate improvement (category 3) and slight
`improvement (category 4). Efficacy for treatment fail·
`ure was the opposite of therapeutic success, and was
`determ ined by assessing the percentage of patients
`meeting the treatment failu re criteria. Efficacy for sig(cid:173)
`
`moidoscopic index was determined using mean change l
`line to last available record) or not im proved (no change l
`
`from baseline to endpoint on a 15-point scale. For
`biopsy score a nalysis, patients were classified as either
`improved (decrease of at least one category from base-
`
`o r an increase in category). Efficacy for the four clinical
`sym pto ms a nd trips to the toilet was assessed using
`mean cha nge from baseline to endpoint.
`Induction of remission. Induction of remission was
`assessed indi vidually for PGA, sigmoidoscopic index.
`and biopsy score at final visit using stringent criteria: a
`rating of category I (complete relief of symptoms) only
`for PGA, a grade of 0 to 4 points (out of a maximum
`of 15) o n the sigmoidoscopic index, or a grade ofO to
`I on biopsy score (t hought to indicate clinically normal
`mucosa or inactive UC) with a minimum of a one·
`grade decrease fro m baseline to final visit.
`
`

`
`r"'P'•--~~~~~~---------------......... ..........
`
`l11g11.1t 1993
`
`MESALAMINE CA PSULES FOR TREATMENT OF ACUTE UC
`
`11 9 1
`
`S1aus11tal method,
`Tnal size was pcci ftcd before the trial began as 70
`patients per lrealmcnt group, v. hich \\US approx im a tely
`the mean of sample siLc '\ calcula ted for fo ur of the
`response variables. T he e um atcd sample sizes for the
`four\'ariabk s rang~d fro m 23 to 100 patie nts per group
`for SOc; power to dclCCl difTe rcnces between mesalam(cid:173)
`tne and placebo u ing two-sid ed statistica l tests at <~ =
`005.
`,>\II random11.t.'d pauen ts \\ho ingested st udv medi(cid:173)
`cation were included m the a na lysic;. E ndpoi~ts were
`assign1.'d according 10 thl' intc nti o n-10-tn:at principle,
`u~ing the last ob~rva11on carried fo rward . Imputation
`w:1Semplo)cd forda1a missing a t baseline or endpoint.
`Baseline comparab1ht ~ across treatme nt groups was
`trsted b} the Kru!.ka l-\\'a ll ic; tc~t (rnntinuous factors)
`<1nd the • test (ca1cgonc.1l fac tor.,; ).
`Cont.muou'> rc~pon-.c \:ln nble!> were a nalyzed using
`anJl}~IS of co"anancc. C hange from baseline was ana(cid:173)
`lyzed. adju 11ng for basehm.• a nd ce nter. T he ra nk trans(cid:173)
`formauon wa..c, cmplo)cd whe n there wa evide nce that
`lhc data \\Crc not normall y d istributed (24).
`D1chotomou
`respo nse \ <l riabl c~ (two categories)
`\.\ere anaJ~7cd using log1'>tic rcgrcs io n. adjusting for
`ba~hne (when a baseline existed) and ce nter.
`All statisucal tests were two-sided. In addi tion to
`testing diITercnces between placebo and each treatment
`group, a three-dose linea r tre nd test using orthogonal
`contrasts was employed 10 dcterm i ne whether there was
`an mcrcasing response wi th increasing dose for the three
`mcsalamine treatment groups (25).
`Seven of the 20 sites enrolled J 0 o r fewer patients.
`Before the studv was unblinded, fi ve sites were pooled
`into one group.and two sites were combined into an(cid:173)
`other group for purposes of statistical analysis. The
`olher 13 sites enrolled between 15 and 44 pati ents.
`
`R ESULTS
`
`Basl!/111e demographics
`Baseline demographic characteristics for the 374 pa(cid:173)
`uents who were randomized and received drug are
`shown in Table 2 and did not difTer significantly among
`treatment groups.
`
`Early discontinuation of therapJ'
`The number of pati~nts withdrawing from the trial
`prematurely was significantlv sm a ller in the 2-g ( P =
`0.008) and the 4-g treatment groups (p = 0.002) tha~
`in the placebo treatment group (Fig. 1 ). The number .0
`1
`patients from the 1-g group withdrawing from tbe tn~
`prematurely (23/92) was smaller than that number 10
`the placebo group (30/ 90); the difference across all
`treatments was statistically significa nt ( P = o.oo5).
`
`Clinical improvement
`Physician global assessment. For the PGA assess(cid:173)
`ment that measured treatment benefit, 7 13
`(p =
`0 .0 .191 ), 79% (p = 0.0002), and 84 3 (p < 0.000 1) of
`patients, respectively, on the 1-g, 2-g, and 4-g doses
`were assessed by the physician to have received treat(cid:173)
`me nt benefit, compared to 54% on placebo. A three(cid:173)
`dose trend across the three active mesalamine doses
`was also significant [p = 0.0267 (Fig. 2)].
`For the PGA assessment that measured treatment
`success, both the 2- and 4-g doses of mesalamine were
`signjficantly superior to placebo, with 573 ( p = 0.002 1)
`and 59% (p = 0.0012) of patients, respectively, achiev(cid:173)
`ing complete relief or marked improvement of symp(cid:173)
`toms, in contrast to 36% on placebo. Treatment success
`was achieved in 453 of patients receiving the 1-g dose,
`which was not significantly different from placebo
`( p = 0.1681 ). However, a three-dose trend across the
`three mesalamine doses was marginally significant [p =
`0.0610 (Fig. 2)).
`Sigmoidoscopic index. Mean change in sigmoido(cid:173)
`scopic index was statistically significant for the 2-g and
`4-g doses with mean respective decreases (improve(cid:173)
`ment) of 4.3 and 5.0, in contrast to a 2.5-point decrease
`with placebo [p = 0.0033 and 0.0001, respectively
`(Table 3)). For the 1-g treatment group, the mean
`change in the sigmoidoscopic index (3.4) was not sig(cid:173)
`nificantly different than placebo ( p = 0.1 600). Regard(cid:173)
`less, a significant dose-response relationship was de(cid:173)
`tected across the three active mesalamine doses (p =
`0.0108), indicating a trend for improvement with an
`increase in mesalamine dose.
`Treatment fa ilure. Nine percent of patients in the 4-
`g treatment group were classilied as treatment failures
`in contrast to 183, 173 , and 22 % in the 2-g, l-g, and
`placebo groups, respectively. A dose-resp~nse relation(cid:173)
`ship was not detected across the three active mesalam(cid:173)
`ine doses.
`Biopsy score. A statistically significant decrease (im(cid:173)
`provement) in biopsy score compared to placebo was
`observed in 58% of patients at the 4-g dose, compared
`to 37% on placebo (p = 0.0015). Neither the 1- or 2-g
`dose was statistically significantly different from pla(cid:173)
`cebo with 42% and 41 3, respectively, improvi ng over
`baseline. A significant dose-response relationship was
`detected across the three active doses (p = 0.03~6).
`indicating a trend for improvement as the mesalamrne
`dose increased.
`Clinical improvement by disease location. For the
`PGA assessment that measured treatment success, t.here
`were no significant differences in response '.or e1t~er
`the 2- or 4-g dose of mesalamine between patients with
`pancolitis and those wi~h. distal UC ( P > 0.6915). For
`pancolitis patients recervmg the 2-g and 4-g dose, ~e­
`spectively, 6 t 3 and 523 experienced complete rehef
`
`

`
`1192
`
`HANAUER et al.
`
`Vol. 88, No. 8, 1993
`
`TABLE 2
`Baseline Characteristics*
`
`Placebo
`(n = 90)
`
`54% (49/ 90)
`46% (41 / 90)
`
`6% (5/ 90)
`94% (85/90)
`
`69% (62/90)
`31 % (28/90)
`
`28% (25/ 89)
`72% (64/89}
`
`l g
`(n = 92)
`
`53% (49/ 92)
`47% (43/ 92)
`
`7% (6/ 92)
`93% (86/ 92)
`
`65% (60/ 92)
`35% (32/ 92)
`
`30% (27/ 9 1)
`70% (64/ 91)
`
`35% (32/ 92)
`65% (60/ 92)
`39.9 ± 13.2
`(n = 92}
`7.1 ±7.9
`(n = 92)
`
`2g
`(n = 97)
`
`5 1 % (49/ 97)
`49% (48/ 97)
`
`7% (7/ 96)
`93% (89/96)
`
`68% (66/ 97)
`32% (3 1/ 97)
`
`2 1 % (20/ 97)
`79% (77/97)
`
`41 % (40/ 97)
`59% (57/97)
`40. 1 ± 14.6
`(n = 97)
`5.9 ± 6. 1
`(n = 97)
`
`4 g
`(n = 95)
`
`44% (42/95)
`56% (53/ 95)
`
`6% (6/95)
`94% (89/95)
`
`723 (68/95)
`28% (27/95}
`
`29% (27/92)
`71 % (65/92)
`
`40% (38/94)
`60% (56/ 94)
`40.9 ± 13.0
`(n = 95)
`7.9 ± 8.0
`(a= 95)
`
`Sex
`Male
`Female
`Smoking
`Yes
`No
`Location
`Distal
`Pan col
`Recent oral steroid use
`Yes
`No
`Recentt SASP* use
`Yes
`No
`Age (yr) (mean ± SD)
`
`Years of UC disease (mean ± SD)
`
`42% (38/90)
`58% (52/90)
`39.6 ± 13.4
`(n = 90)
`7. 1 ± 7.0
`(n = 90)
`*No statistical differences detected across treatment groups for any baseline characteristics.
`t Recent = use of SASP or oral steroids within 30 days of starting study medication.
`* Sulfasalazine.
`
`35
`
`30
`
`25
`
`33(30)
`
`1(1)
`
`12111>
`
`25123)
`
`4(41
`
`-
`
`lnsu!fic.lmtt Therapou!IC: Efteci
`
`c:J Advet&l!Eventf
`1mon;unen1 CondlbOn
`
`El ~~ltl/
`CJ 0010!
`
`1 gram
`Placebo
`Trearmenl
`n-90
`n:92
`•suute-=sbr llgf'llfica"rt cornparod to placebO. p< 05
`F1G. I. Summary of early terminations [percent (n) of patients].
`
`2 gram
`n-97
`
`or marked improvement of symptoms in contrast to
`32% on placebo. In comparison, 55 % and 62% of
`patients with distal disease receiving the 2-g and 4-g
`doses, respectively, experienced treatment success as
`measured by PGA, compared with 37% on placebo.
`Differences in mean change in sigmoidoscopic index
`between patients with pancolitis and distal disease were
`not statistically significant for the 2-g and 4-g doses
`[p > 0.7963 (Table 4)]. Also, no statistically significant
`differences were noted in the number of treatment
`failures between patients with pancolitis and those with
`distal UC (p > 0. 7325). In the 2-g and 4-g groups,
`respectively, 16% and 15% of pancolitis patients were
`classified as treatment failures, compared with 29%
`receiving placebo; in the group of patients with distal
`
`100
`
`90
`
`80
`
`70
`
`60
`
`~ i
`
`: 50
`~
`,
`411
`
`30
`
`20
`
`10
`
`0Bonf11'1
`
`o -
`- """"""""
`
`5 4(49)
`
`36 32)
`
`71(65)*
`
`45(41 )
`
`79(77)*
`
`114(80)"
`
`57(55)*
`
`59(56)"
`
`-
`
`'
`IU"'m
`n.-92
`n.90
`•saius.riCollfly •IO"'I~ comoeted '° pli&CebO P< 05
`FIG. ~- Physicia!'l ~obal assessment (PGA): Percent (n) of patients
`who achieved rem1ss1o n. treatment success o r treatment benefit by
`PGA.
`'
`
`T ABLE 3
`Efficacy R esults.for Sigmoidoscopic Index
`
`Sigmoidoscopic Index
`
`Treatment•
`
`Placebo (n = 90)
`I g (n = 92)
`2 g (n = 97)
`4 g (n = 95)
`
`Baseline
`(mean± SE)
`10.3 ± 0.30
`9.9 ± 0.29
`10.2 ± 0.26
`10.3 ± 0.27
`
`Endpoint
`(mean+ SE)
`
`7.8 ± 0.54
`6.6 ± 0.49
`5.9 ± 0.47
`5.3 ± 0.45
`
`Change
`(mean + SE)
`- 2.5 ± 0.45
`-3.4 ± 0.45
`-4.3 ± 0.43t
`-5.0 + 0.44t
`
`• A three-dose trend across the three active mesalamine doses was
`statisticall y significant, p < 0.05.
`t Statistically significant, compared with placebo, p < 0.05.
`
`

`
`August 199 3
`
`MESALAMINE CAPSULES FOR TREATMENT OF ACUTE UC
`
`11 93
`
`TABLE 4
`Ej)icac.v Res11/1sfor Sigmoidoscopic Index by Disease l ocal ion
`(11 = 374)*
`
`Treatment*
`
`Baseline
`(mean ± SE)
`
`Change
`Endpoint
`(mean ±SE) (mean ± SE)
`
`Sigmoidoscopic Index (Dis1a/ UC)
`10.2 ± 0.36
`7.9 ± 0.65
`Placebo (n = 62)
`10.0 ± 0.35
`I g (n = 60)
`7.1 ± 0.64
`9.9 ± 0.32
`5.7 ± 0.58
`2 g(n = 66)
`4g(n=68)
`10.2±0.3 1
`5.2±0.51
`Sigmoidoscopic Index (Panco/i1is)
`Placebo (n = 28)
`10.8 ± 0.57
`7.7 ± 1.02
`lg(n=32)
`9.7±0.50
`5.6±0.70
`2g(n = 31)
`11.0 ± 0.40
`6.4 ± 0.84
`4g(n=27)
`10.7 ±0.54
`5.8±0.92
`
`-2.4± 0.54
`-3. 1 ±0.56
`-4.4 ± 0.53
`- 5. 1 ±0.52
`
`-2.8 ± 0.8 1
`-4.0± 0.76
`-4.3 ± 0.77
`-4.7 ± 0.83
`
`•The test for differences in treatment effect on sigrnoidoscopic
`index between patients with distal UC and patients with pancolitis
`was not statistically significant (p = 0.7963).
`
`disease, 18% and 7% receiving 2 g and 4 g of mesalam(cid:173)
`ine were categorized as treatment fai lures, in contrast
`to 19% receiving placebo. Finally, there were no signif(cid:173)
`icant differences in decreases in biopsy score for either
`the 2-g or 4-g doses between patients with pancolitis
`and those with distal UC ( p > 0.4061 ). For pancolitis
`patients receiving the 2-g and 4-g dose, respectively,
`35% and 59 % experienced a decrease in biopsy score
`in contrast to 29% on placebo. In comparison, 44%
`and 57% of patients with distal disease receiving the 2-
`g and 4-g doses experienced a decrease in biopsy score,
`compared with 40% o n placebo.
`Trips to the toilet and clinical symptoms. Compared
`with placebo, the 2- and 4-g doses of mesalamin~ pro(cid:173)
`duced statistically significant improvement in tnps to
`the toilet and in all four clinical symptoms (p = 0.0040
`and p = 0.00 10, respectively). Mean decreases of 23%
`at 2 g and 30% at 4 g for trips to the toilet were reported,
`compared to a 12% decrease on placebo. Mean de(cid:173)
`creases of 35% and 39% at 2 g and 4 g, respectively,
`were noted for abdominal/rectal pain, compared with
`a 5% decrease with placebo. Similar significant de(cid:173)
`creases were also recorded for the other clinical symp(cid:173)
`toms, compared with placebo (Table 5). For the l-g
`
`treatment group, on ly stool consistency was signifi(cid:173)
`cantly improved, compared with placebo (p = 0.0 148).
`Improvement in trips to the toiJet (p = 0. 1606), rectal
`bleeding (p = 0.2002), abdominal/rectal pain (p =
`0.2249), and urgency ( p = 0.1214) did not reach statis(cid:173)
`tical significance for this dose, compared with placebo.
`A statisticalJy significant dose-response relationship was
`detected across the three active mesalamine doses for
`trips to the toilet, rectal bleeding, and abdominal/rectal
`pain (p < 0.05) and was borderline for stool consistency
`(p = 0.0552) and urgency (p = 0.0825).
`
`Induction of remission
`PGA (Table 6, Fig. 2). Both the 2-g (p = 0.0024) and
`4-g ( p = 0.0019) doses of mesalamine were statistically
`significant for inducing remission by PGA, compared
`with placebo. Twenty-nine percent of patients in the 2-
`g group and 29% of patients in the 4-g group achieved
`remission as assessed by the physician, compared to
`12% on placebo. Twenty-one percent of patients in the
`1-g group achieved remission by the PGA parameter;
`however, this did not reach statistical significance when
`compared to placebo (p = 0.0925).
`Sigmoidoscopic index (Table 6). The 2-g (p = 0.0469)
`and 4-g (p = 0.0140) doses of mesalamine were statis(cid:173)
`tically significant for inducing remission by sigmoido(cid:173)
`scopic index, compared with placebo. Forty-four per(cid:173)
`cent and 48% of patients in the 2-g and 4-g groups,
`respectively, achieved endoscopic remission, compared
`to 3 1 % on placebo. Forty percent of patients in the 1-
`g group reached endoscopically defined remission
`which was not signi ficantl y different than placebo (p =
`0.1637).
`Biopsy score (Table 6). T he 4-g dose of mesalamine
`was statistically significant (p = 0.02 13) for induci ng
`remission by biopsy score, compared with placebo.
`Thirty-nine percent of patients achieved histologic re(cid:173)
`mission compared with 23% on placebo. Treatment
`with either the 2-g or l-g doses of mesalamine did not
`produce remission by this parameter that was statisti(cid:173)
`cally significant compared to placebo.
`
`T ABLE 5
`Mean and(%}* Change in Clinical Symploms and Trip.110 Toi/el
`
`Clinical Symptom
`
`Placebo
`(n - 90)
`-0.7 (12%)
`- 1.4(26%)
`-1.7 (36%)
`-0.2 (5%)
`1.2 (21 3 )
`
`I g
`(n = 92)
`-I.I (15%)
`- 2.6* (423)
`-2.3 (47%)
`-0.5 (13%)
`1.7 (29%)
`
`Trips to toilett
`Stool consistency
`Rectal bleedingt
`Abdominal/ rectal paint
`Urgency
`<
`• 3 reflects the percentage decrease (improvement) in sympt~m severi.ty:
`.
`'fi
`t Three-dose trend across three active mesalamine doses achieved stausucal sigm icance. P
`*Statistically significant compared with placebo, P < 0.05.
`
`4g
`(n = 95)
`-17*(30%)
`- 3.5t (65%)
`-3.2* (67%)
`-l.4t (39%}
`-2.5t(50%1
`
`2g
`(n = 97)
`1.5* (23%)
`-3.2* (573)
`-2.9* (58%)
`-l.3H35%)
`-2.2* (39%)
`
`0 05
`·
`·
`
`

`
`1194
`
`HANAUER et al.
`
`TABLE 6
`t.ffecacy Res11/1sfor J11d11ctiu11 of Remission [Percem (11) of Patiellls
`in Remission]
`
`Treatment
`
`Parameter
`
`Placebo
`(n =90)
`12% (1 1)
`PGA
`31 % (28)
`Sigmoidoscopy
`23% (21)
`Biopsy
`•Statistically significant. compared with placebo. P < 0.05.
`
`4g
`2g
`lg
`(n = 95)
`(n =97)
`(n = 92)
`21 % (19) 29% (28)* 293 (28)*
`40% (37) 443 (43)* 483 (46)*
`393 (37)*
`28% (26) 29% (28)
`
`Subgroup analyses
`The demographic subgroups (years ~ince UC d_iag(cid:173)
`nosis, age, gender, smoking status, disease location.
`recent oral steroid use, or recent sulfasalazine use) were
`examined for the consistency of treatment effect (inter(cid:173)
`action) with each of the nine efficacy parameters. None
`of the subgroups were found to be consistent predictors
`of treatment response. In addition, the treatment-by(cid:173)
`site interaction was not statistically significant for any
`of the efficacy parameters for clinical improvement
`and/or induction of remission, indicating that the mag(cid:173)
`nitude of tTeatment effects was not significantly differ(cid:173)
`ent between investigative sites.
`Compliance
`Of the 374 patients who ingested study medication,
`338 (90%) were considered to be medication compliant,
`using the following criteria: compliance was <:::: 703 for
`the duration of the study, patients had not been off
`medication for more than 2 days prior to final visit,
`and patients consumed study medication for at least 4
`days prior to terminating study participation.
`Adverse reactions to therapy
`A total of 32 patients discontinued participation in
`the trial due to an adverse reaction (treatment- or
`nontreatment-related): 11 receiving placebo and five,
`nine, and seven, respectively, recei