CLINICAL—ALIMENTARY TRACT
`
`GASTROENTEROLOGY 2012;143:1473–1481
`
`CLINICALAT
`
`Teduglutide Reduces Need for Parenteral Support Among Patients With
`Short Bowel Syndrome With Intestinal Failure
`PALLE B. JEPPESEN,* MAREK PERTKIEWICZ,‡ BERNARD MESSING,§ KISHORE IYER,储 DOUGLAS L. SEIDNER,¶
`STEPHEN J. D. O’KEEFE,# ALASTAIR FORBES,** HARTMUT HEINZE,‡‡ and BO JOELSSON§§
`
`*Department of Medical Gastroenterology, Rigshospitalet, Copenhagen, Denmark; ‡Department of General Surgery and Clinical Nutrition, Warszawa, Poland; §Hopital
`Beaujon Service de Gastroenterologie et Assistance Nutritive, Clichy, France; 储Mount Sinai Medical Center, New York, New York; ¶Vanderbilt University Medical
`Center, Nashville, Tennessee; #University of Pittsburgh Medical Center, Pittsburgh, Pennsylvania; **University College London, London, UK; ‡‡Nycomed, a Takeda
`Company, Nycomed GmbH, Konstanz, Germany; and §§NPS Pharmaceuticals, Bedminster, New Jersey
`
`See Covering the Cover synopsis on page
`1403; see editorial on page 1416.
`
`BACKGROUND & AIMS: Teduglutide, a glucagon-like
`peptide 2 analogue, might restore intestinal structural
`and functional integrity by promoting growth of the mu-
`cosa and reducing gastric emptying and secretion. These
`factors could increase fluid and nutrient absorption in
`patients with short bowel syndrome with intestinal failure
`(SBS-IF). We performed a prospective study to determine
`whether teduglutide reduces parenteral support in pa-
`tients with SBS-IF. METHODS: We performed a 24-week
`study of patients with SBS-IF who were given subcutane-
`ous teduglutide (0.05 mg/kg/d; n ⫽ 43) or placebo (n ⫽
`43) once daily. Parenteral support was reduced if 48-hour
`urine volumes exceeded baseline values by ⱖ10%. The
`primary efficacy end point was number of responders
`(patients with ⬎20% reduction in parenteral support vol-
`ume from baseline at weeks 20 and 24). RESULTS: There
`were significantly more responders in the teduglutide
`group (27/43 [63%]) than the placebo group (13/43 [30%];
`P ⫽ .002). At week 24, the mean reduction in parenteral
`support volume in the teduglutide group was 4.4 ⫾ 3.8
`L/wk (baseline 12.9 ⫾ 7.8 L/wk) compared with 2.3 ⫾ 2.7
`L/wk (baseline 13.2 ⫾ 7.4 L/wk) in the placebo group
`(P ⬍ .001). The percentage of patients with a 1-day or
`more reduction in the weekly need for parenteral support
`was greater in the teduglutide group (21/39 [54%]) than in
`the placebo group (9/39 [23%]; P ⫽ .005). Teduglutide
`increased plasma concentrations of citrulline, a biomarker
`of mucosal mass. The distribution of treatment-emergent
`adverse events that led to study discontinuation was sim-
`ilar between patients given teduglutide (n ⫽ 2) and pla-
`cebo (n ⫽ 3). CONCLUSIONS: Twenty-four weeks of
`teduglutide treatment was generally well tolerated in
`patients with SBS-IF. Treatment with teduglutide re-
`duced volumes and numbers of days of parenteral
`support for patients with SBS-IF;
`
`Scan the quick response (QR) code to the left with
`your mobile device to watch this article’s video ab-
`stract and others. Don’t have a QR code reader? Get
`one by searching ‘QR Scanner’ in your mobile de-
`vice’s app store.
`
`Short bowel syndrome (SBS) results from surgical re-
`
`section, congenital defect, or disease associated loss of
`absorption. The concomitant malabsorptive spectrum of
`SBS is wide, and patients with SBS are heterogeneous
`because of large variations in remnant bowel anatomy and
`function. Patients with intestinal insufficiency are able to
`compensate for their malabsorption by physiologic or
`pharmacologic adaptation,1,2 whereas supplemental par-
`enteral support (PS; parenteral nutrition and/or intrave-
`nous [PN/IV] fluids) is required to maintain fluid, elec-
`trolytes, trace elements, vitamins, and nutrient balances in
`patients with SBS with intestinal failure (SBS-IF).3,4 Treat-
`ments aim to maximize remnant intestinal absorptive
`capacity; to minimize the symptoms of malabsorption;
`and to avoid, minimize, or eliminate the need for PS,
`thereby alleviating the daily burden of this debilitating
`condition. Hormonal therapies focusing on enhancing
`the structural and functional integrity of the remaining
`intestine are emerging. Glucagon-like peptide 2 (GLP-2), a
`peptide secreted from the intestinal L cells after food
`ingestion, ameliorates the pathophysiologic consequences
`of SBS. GLP-2 administration inhibits gastric acid secre-
`
`Abbreviations used in this paper: AE, adverse event; FCE, fluid com-
`posite effect; GLP-2, glucagon-like peptide 2;
`IF,
`intestinal failure;
`PN/IV, parenteral nutrition and/or intravenous; PS, parenteral support;
`SBS, short bowel syndrome; TEAE, treatment-emergent adverse event.
`© 2012 by the AGA Institute
`0016-5085/$36.00
`
`Keywords: Clinical Trial; Glucagon-Like Peptide 2; Gastro-
`intestinal Disorder; Therapy.
`
`Watch this article’s video abstract and others at http://
`tiny.cc/j026c.
`
`Page 1
`
`

`
`1474 JEPPESEN ET AL
`
`GASTROENTEROLOGY Vol. 143, No. 6
`
`Table 1. Key Inclusion and Exclusion Criteria
`
`Inclusion criteria
`
`Exclusion criteria
`
`SBS resulting from intestinal failure caused by a major intestinal
`resection (eg, injury, cancer, Crohn’s disease, vascular disease,
`volvulus)
`At least 12 continuous months of PS dependency (PN and/or IV fluids)
`before signing informed consent
`PS required ⱖ3 times weekly to meet caloric, fluid, or electrolyte needs
`Patients with Crohn’s disease had to be in clinical remission for ⱖ12
`wk before dosing
`
`Cancer within last 5 y
`Body mass index ⬍15 kg/m2
`Inflammatory bowel disease on immunosuppressant therapy that
`has been introduced or changed within last 3 mo or treatment
`with biologics within last 6 mo
`Previous use of teduglutide
`Previous use of native GLP-2 or human growth hormone within 6
`mo before screening
`⬎4 SBS-related hospital admissions within 12 mo or hospital
`admission within 30 d before screening
`
`CLINICAL AT
`
`tion and motility,5,6 stimulates intestinal blood flow,7
`increases intestinal barrier function,8 and enhances nutri-
`ent and fluid absorption in preclinical and clinical mod-
`els.9-12
`Teduglutide, a dipeptidyl-peptidase degradation-resis-
`tant GLP-2 analogue, has been demonstrated to enhance
`structural and functional integrity of the remaining intes-
`tine in SBS. Open-label, uncontrolled studies in adult
`patients with SBS have suggested clinically meaningful
`reductions of fecal excretions of wet weight (⬃700 to
`1000 g/d) and energy (⬃1 MJ/d) after treatment with
`GLP-2 and teduglutide.10,11,13,14 A recently published, ran-
`domized, placebo-controlled phase 3 study investigated
`whether teduglutide, by increasing intestinal absorption,
`could facilitate PS reductions in patients with SBS-IF.12
`Contrary to the expectations of a dose response, a 0.10-
`mg/kg/d dosage did not meet the primary end point of PS
`reduction, but significant findings from the ad hoc anal-
`ysis of a 0.05-mg/kg/d dosage in that study suggested that
`these differences could be explained by the limitation of
`PS volume reductions to no more than 10% of baseline
`levels, beginning only at the fourth week of dosing, along
`with a trend toward larger baseline PS volume require-
`ments in the 0.10-mg/kg/d group. Therefore, the primary
`objective of this study, the largest double-blind, random-
`ized, placebo-controlled trial performed in patients with
`SBS-IF, was to evaluate whether teduglutide at the 0.05-
`mg/kg/d dosage and with a protocol allowing for earlier
`(ie, at second week of dosing) and more aggressive PS
`reductions of 10% to 30% of baseline levels of PN/IV fluid
`could reduce PS volume in these patients.
`
`Materials and Methods
`All authors had access to the study data and have re-
`viewed and approved the final manuscript.
`
`Patients
`After receiving approval from local institutional review
`boards or medical ethics committees, centers screened patients
`of both sexes who were 18 years of age or older and who had a
`history of SBS that resulted in a dependency on PS for a period
`of at least 12 months before the start of the study. PS depen-
`dency was defined as at least 12 continuous months of PS
`required at least 3 times weekly to meet their caloric, fluid, or
`electrolyte needs because of ongoing malabsorption. The proto-
`
`col did not specify whether previous attempts at weaning had to
`be made.
`. Al-
`Inclusion and exclusion criteria are listed in
`though patients with neoplasms could be included in the study,
`patients with ongoing radiation enteritis or the presence of
`damaged enteral tissue due to radiation enteritis were excluded,
`along with any condition or circumstance that, in the investiga-
`tor’s opinion, put the patient at undue risk or jeopardized the
`integrity of the study results, including the presence of any of
`the excluded disease states described in the Supplementary Ma-
`terials (Supplementary Table 1).
`Patients were not categorized by whether they were receiving
`parenteral nutrition vs intravenous fluids alone. Throughout the
`study, patients were requested to maintain habitual diet and
`fluids, and no new medications were started or ongoing treat-
`ments changed during the stabilization period or throughout
`the 24-week treatment period unless deemed medically neces-
`sary. Patients who completed the 24-week treatment period were
`offered entry into an open-label extension study, the results of
`which will be the subject of a separate report.
`
`Study Design
`In this multinational, multicenter, randomized, double-
`blind, placebo-controlled, parallel-group, 2-stage, phase 3 study
`(Figure 1), patients were recruited from 27 sites in 10 countries
`across Europe and North America. Stage 1 consisted of a screen-
`ing visit and optimization and stabilization periods. After
`screening, eligible patients underwent a PS optimization period,
`if needed, of up to 8 weeks to achieve a stable target urine output
`of 1.0 to 2.0 L/d. This range, at the higher end of normal output
`(⬃1.5 L/d), was considered to minimize the risk for dehydration-
`related complications (eg, renal calculi) without provoking hy-
`perhydration in a population that is prone to diarrhea and
`
`Figure 1. Study design.
`
`Page 2
`
`

`
`CLINICALAT
`
`December 2012
`
`TEDUGLUTIDE AND SBS 1475
`
`Table 2. PN/IV Adjustments Based on 48-Hour Urinary Output
`
`Urine outputa
`
`PN/IV action
`
`⬍1.0 L/d or target based on stabilized urine output
`ⱖ1.0 L/d but less than baseline
`0% to ⬍10% increase over baseline
`ⱖ10% increase over baseline
`
`Increase PN/IV by ⱖ10% (wk 2) or to previous level
`If patient is dehydrated or inadequately nourished, increase PS; if not, maintain PS
`Maintain PS
`Reduce PS by ⱖ10% of stabilized baseline level up to a clinically appropriate
`amount (maximum of 30%)
`
`aBaseline urine output is the urine volume obtained during the stabilization period before initiating treatment.
`
`dehydration. All patients then underwent a 4- to 8-week stabi-
`lization period during which PS usage was to match prescribed
`PS, and oral fluid intake and urine volume could not deviate
`⬎25% from the optimized levels. Although the osmolarity and
`oral intake were not strictly controlled, patients were asked to
`keep intake as constant as possible. Patients who were not stable
`could repeat stage 1 once.
`Stage 2 began when the patients demonstrated PS volume
`stability. Eighty-six patients were randomized in a 1:1 ratio to
`placebo or teduglutide 0.05 mg/kg/d (administered once daily
`subcutaneously into the abdomen, thigh, or arm, at approxi-
`mately the same time each day) for 24 weeks (Figure 1). Ran-
`domization was performed according to a computer-generated
`interactive response system and was stratified at 2 levels of
`baseline PS volume (ⱕ6 or ⬎6 L/wk). The postrandomization
`study evaluations and visits were scheduled at weeks 1, 2, 4, 8,
`12, 16, 20, and 24.
`All patients were required to record PS volume, 48-hour oral
`fluid intake and urinary output, and study drug dosing infor-
`mation in an electronic diary. PS volume was recorded daily and
`48-hour oral fluid intake and urinary output was recorded dur-
`ing the optimization and stabilization periods and at weeks 2, 4,
`8, 12, 16, 20, and 24 during the treatment period. If there was a
`change in oral intake, the clinician considered whether to adjust
`PS volume. Attempts to reduce PS volume were made at every
`visit before week 24.
`
`Optimization and Stabilization
`During the optimization period, patients were assessed
`at planned intervals (weeks 2, 4, 6, and 8, ⫾3 days) for hydration
`and nutrition. PS was adjusted in targeted increments of ⱖ10%
`of the volume at the previous visit. Immediately before each
`scheduled visit, 48-hour oral fluid intake and urine output were
`measured. The measurement included 1 day on and 1 day off PS,
`unless the PS was infused daily. Blood and urine samples were
`collected at each visit to evaluate hydration and nutrition. A
`targeted urine output of 1.0 to 2.0 L/d was used to determine if
`patients required optimization or could enter the stabilization
`period.
`Stability was defined as actual PS usage matching the pre-
`scribed PS, baseline 48-hour oral fluid intake and urine output
`volumes within ⫾25% of the respective 48-hour volumes, and
`urine output volume of 2 to 4 L per 48 hours. No further PS
`adjustments were permitted during the stabilization period.
`The purpose of the PS optimization period was to ensure that
`all patients received and tolerated a stable minimal level of PS
`before treatment, with adequate hydration as indicated by urine
`output. Patients who failed to remain stable for at least 4
`consecutive weeks immediately before randomization were to
`start the optimization period again. Those patients who failed to
`stabilize after 2 attempts could not proceed and were not ran-
`domized.
`
`Efficacy and Safety
`During the treatment period, PS adjustments were
`targeted to be ⱖ10% but ⬍30% of stabilized PS level. Patients
`were required to remain compliant with the prescribed PS
`throughout the study, with all adjustments based on the
`actual PS volume infused. Patients were assessed at planned
`intervals (baseline and weeks 2, 4, 8, 12, 16, and 20) for
`hydration and nutrition. Before all scheduled visits, 48-hour
`oral fluid intake and urinary output measurements were
`taken and included 1 day on and 1 day off PS, unless the PS
`was infused daily.
`Reductions in PS volumes by 10% to 30% of baseline PS levels
`were allowed if the 48-hour urinary volumes exceeded the base-
`line values by ⬎10%. Oral intake during these 48-hour balances
`was to be constant. Determination of the amount of PS volume
`reduction was based on 48-hour urinary output, according to
`the algorithm described in
`. The decision of whether to
`stop a day of PS, reduce the percentage volume of all days that
`PS was administered, or change the relative PN/IV constituents
`of the PS or whether total PS weaning was possible was based on
`the investigator’s clinical judgment and the personal preference
`of the patient.
`Interim safety evaluations 1 week after PS reductions ensured
`that PS reductions were well tolerated. This assessment of nour-
`ishment and hydrational status was based on repeated 48-hour
`urine collections and a clinical evaluation that included clinical
`signs and symptoms of dehydration, change in body weight,
`reviews of the recorded oral fluid intake, blood samples (hemat-
`ocrit, creatinine, blood urea nitrogen), and urine sodium. Be-
`cause the injection site reactions or stomal changes that are
`known to occur with GLP-2 and teduglutide might have un-
`blinded the observer, the clinician assessing and adjusting PS
`volume was required to be different from the one conducting
`the physical examination and assessing safety. If the reduced PS
`volume was well tolerated, the new weekly PS volume was main-
`tained until the next visit; if not, the previously tolerated PS
`volume was resumed. Patients could be rechallenged at the next
`visit if adequate hydration and nutrition requirements were met.
`The primary efficacy end point was the percentage of patients
`who demonstrated a response at week 20 and maintained that
`response at week 24 (responder). A response at a given visit was
`defined as the achievement of a 20% to 100% reduction from
`baseline in weekly PS volume. The secondary efficacy end points
`included the percentage and absolute change in PS and the
`number of patients who stopped PS and their time of discon-
`tinuation.12
`Exploratory end points included response by visit, reduction
`in days on PS, change from baseline in plasma concentrations of
`citrulline (an amino acid produced by enterocytes and used here
`as a biomarker of remnant enterocyte mass),15 and change in the
`fluid composite effect (FCE).12 The FCE is a putative measure of
`
`Page 3
`
`

`
`1476 JEPPESEN ET AL
`
`GASTROENTEROLOGY Vol. 143, No. 6
`
`the combined effects of teduglutide on intestinal fluid absorp-
`tion, that is, not only on PS volume reduction, but also on the
`ability to reduce oral fluid intake and increase urine output
`volume. The FCE was a summation of the increase in urine
`production, reduction in PN/IV volume, and reduction in oral
`fluid intake (L/wk), calculated as a baseline measurement of the
`individual components. The FCE was also calculated for each
`scheduled postbaseline visit using the following equation: reduc-
`tion in PS volume (L/wk) ⫹ reduction in oral fluid intake
`volume (L/wk) ⫹ increase in urine output volume (L/wk).
`Clinical evaluations (vital signs, physical examinations, and
`electrocardiograms), adverse event (AE) monitoring, and labora-
`tory tests (hematology, serum chemistries, and urinalysis) were
`assessed. Safety assessments also included body weight, 48-hour
`urine output, antibodies to teduglutide, and any required endo-
`scopic evaluations. In patients with colon, a baseline colonos-
`copy was required for inclusion to rule out the presence of
`polyps or active intestinal disease.
`Statistical Analysis
`Eighty-six patients were randomized in a 1:1 ratio to
`detect differences in responder rates between teduglutide 0.05
`mg/kg/d and placebo groups of 35% vs 6%, respectively, based on
`the response rates reported in the earlier phase 3 study12 (␣ ⫽
`.05, 2-sided test and power ⫽ 90%). Grounded on these assump-
`tions, nQuery Advisor (version 6.0, Statistical Solutions, Saugus,
`MS) based on Fisher exact test was used to calculate the power.
`The number and percentage of responders are presented here
`by treatment group. The intent-to-treat analysis compared the
`event rates for the 2 treatment groups using the Cochran-
`Mantel-Haenszel test statistics adjusted for the randomization
`stratification variable (ⱕ6 or ⬎6 L/wk of PS volume at baseline).
`The percentage and absolute change in PS volume from baseline
`to the last dosing visit as well as all scheduled visits starting at
`week 4 are presented by treatment group using descriptive sta-
`tistics. Treatment group differences were compared using an
`analysis of covariance model with effects for treatment and
`baseline PS volume, with the potential for the interaction of the
`2 variables also included as an effect. Safety analyses were de-
`scriptive.
`
`Results
`Patients
`From November 2008 to January 2011, one hun-
`dred and thirty-two patients who signed informed con-
`sent forms were screened, 86 were randomized, and 78
`completed the dosing period (Supplementary Figure 1).
`There were no significant differences between treatment
`groups regarding demographic characteristics and medi-
`cations at baseline (Table 3). A total of 39 patients (17 in
`the teduglutide group and 22 in the placebo group) re-
`quired optimization of PN/IV volume before entering the
`stabilization period; 26 patients in the teduglutide group
`and 21 in the placebo group went directly from screening
`to stabilization. However, 12 of these (6 in each group)
`failed to remain stable for a full 4-week period and re-
`quired a return to the optimization period. Overall, mean
`time spent in the optimization stage was 19 ⫾ 23 days for
`the teduglutide group and 23 ⫾ 24 days for the placebo
`group. No patients were weaned off PS during the opti-
`mization period.
`
`Efficacy
`Primary efficacy end point. The primary efficacy
`end point was the responder rate. There were 27/43 (63%)
`responders in the teduglutide group and 13/43 (30%) in
`the placebo group (P ⫽ .002). Small bowel length did not
`appear to be a predictor of response. Responder rate was
`higher for patients without colon in continuity (com-
`pared with patients with colon in continuity); however,
`findings did not achieve statistical significance.
`Secondary end points. At all visits, change from
`baseline in actual PS volume was greater in the teduglu-
`tide group than in the placebo group (Figure 2). At Week
`24, the mean ⫾ SD PS volume reduction in the teduglu-
`tide group was 4.4 ⫾ 3.8 L/wk from a baseline of 12.9 ⫾
`7.8 L/wk vs 2.3 ⫾ 2.7 L/wk from a baseline of 13.2 ⫾ 7.4
`L/wk in the placebo group. The difference in absolute
`change between the treatment groups was statistically
`significant at week 8 (P ⫽ .011) and remained significant
`through week 24 (P ⬍ .001). The percentage reduction in
`actual PS volume at week 24 was 32% ⫾ 19% in the
`teduglutide group vs 21% ⫾ 25% in the placebo group.
`The difference in percentage change between the treat-
`ment groups was significant at week 12 (P ⫽ .028) and
`remained significant through week 24 (P ⫽ .030). No
`patients were completely weaned from PS at week 24.
`Selected exploratory end points. The percentage
`of patients with response (20%⫺100% PS reduction vs
`baseline) was higher in the teduglutide group than in the
`placebo group at all visits. At week 24, 30/39 teduglutide
`patients (77%) demonstrated response vs 18/39 placebo
`patients (46%; P ⫽ .01). The percentage of patients with a
`1-day or more reduction in weekly actual PS use at week
`24 was higher in the teduglutide group (54%, n ⫽ 21/39
`[13 with 1 day off; 8 with ⱖ2 days off]) than in the
`placebo group (23%, n ⫽ 9/39; [6 with 1 day off; 3 with ⱖ2
`days off]; P ⫽ .005).
`Oral fluid intake was significantly higher in patients
`receiving placebo compared with those receiving teduglu-
`tide at weeks 12, 20, and 24 (Figure 3). At all visits, greater
`reduction in FCE was seen in the teduglutide group than
`in the placebo group. At week 24, the mean ⫾ SD reduc-
`tion in the teduglutide group was 5.4 ⫾ 6.0 L/wk vs 1.1 ⫾
`4.3 L/wk in the placebo group (P ⬍ .0006).
`Teduglutide resulted in a significant increase in plasma
`citrulline concentration from baseline levels. At baseline,
`mean ⫾ SD plasma citrulline concentration values were
`18.4 ⫾ 9.5 0 ␮mol/L and 17.5 ⫾ 9.0 ␮mol/L in the
`teduglutide and placebo groups, respectively. At 24 weeks,
`the mean ⫾ SD increase over baseline in plasma citrulline
`concentration was 20.6 ⫾ 17.5 ␮mol/L in the teduglutide
`group vs 0.7 ⫾ 6.3 ␮mol/L in the placebo group (P ⱕ
`.0001). Over 24 weeks, patients receiving teduglutide had
`a nonsignificant increase in body weight of 1.0 ⫾ 3.7 kg
`compared with baseline (P ⫽ .10), whereas patients receiv-
`ing placebo had a decrease in body weight of ⫺0.6 ⫾ 2.8
`kg (P ⫽ .20).
`
`CLINICAL AT
`
`Page 4
`
`

`
`December 2012
`
`TEDUGLUTIDE AND SBS 1477
`
`CLINICALAT
`
`P value
`
`.694a
`
`.759a
`
`.829b
`
`.694b
`
`.100b
`
`.277a
`
`.608a
`
`.332a
`
`.019b
`
`.504a
`
`.707a
`
`.388a
`
`1.000c
`
`Table 3. Demographic Characteristics and Medication at Baseline
`
`Placebo (n ⫽ 43)
`
`Teduglutide, 0.05
`mg/kg/d (n ⫽ 43)
`
`Overall (N ⫽ 86)
`
`Age, mean (SD), y
`Range
`BMI, mean (SD), kg/m2
`n
`Range
`Women, n (%)
`Cause of major intestinal resection, n (%)
`Vascular disease
`Crohn’s disease
`Volvulus
`Injury
`Cancer
`Other
`Intestinal anatomy or remnant small bowel
`length unknown, n
`Patients with stoma, n
`Types of stoma, n (%)
`Jejunostomy
`Ileostomy
`Colostomy
`Other (duodenostomy; jejunostomy ⫹
`ileostomy)
`Colon in continuity, n (%)
`Overall remnant small bowel length, mean
`(SD), cm
`
`n
`Median
`Range
`Remnant small bowel length in patients with
`jejunostomy/ileostomy, mean (SD), cm
`
`n
`Median
`Range
`Remnant small bowel length in patients with
`colon in continuity, mean (SD), cm
`
`n
`Median
`Range
`Remnant colon, n (%)
`⬎25%⫺50%
`⬎50%⫺75%
`⬎75%⫺100%
`Time since last small bowel resection, mean, y
`n
`⬍1
`ⱖ1 to ⬍2
`ⱖ2 to ⬍5
`ⱖ5
`Time receiving PS, mean (SD), y
`Median
`Range
`Parenteral volume, mean (SD), mL/d
`Median
`Range
`Time receiving PS, mean (SD) d/wk
`Median
`Range
`Parenteral volume stratification
`Parenteral volume ⱕ6 L/wk, n (%)
`Receiving PS 3/4/5/6/7 d/wk, n
`Parenteral volume ⬎6 L/wk, n (%)
`Receiving PS 3/4/5/6/7 d/wk, n
`
`49.7 (15.6)
`18–82
`22.3 (3.1)
`43
`17.5–28.6
`24 (56)
`
`16 (37)
`8 (19)
`6 (14)
`4 (9)
`2 (5)
`7 (16)
`3
`
`17
`
`5 (29)
`9 (53)
`1 (6)
`2 (12)
`
`23 (54)
`68.7 (63.9)
`
`40
`48.0
`5–343
`122.8 (81.6)
`
`13
`130
`40–343
`43.3 (31.5)
`
`22
`32.5
`5–100
`
`5 (12)
`8 (19)
`10 (23)
`7.9
`43
`0
`6
`17
`20
`5.9 (5.7)
`3.9
`1.0–25.8
`1929 (1026)
`1771
`514–5000
`5.9 (1.5)
`7.0
`3.0–7.0
`
`7 (16)
`4,1,0,1,1
`36 (84)
`2,2,2,6,24
`
`50.9 (12.6)
`22–78
`22.5 (3.2)
`42
`17.6–9.8
`22 (51)
`
`13 (30)
`10 (23)
`3 (7)
`4 (9)
`1 (2)
`12 (28)
`3
`
`21
`
`11 (52)
`6 (29)
`4 (19)
`0 (0)
`
`26 (61)
`84.4 (64.6)
`
`40
`70.0
`15–250
`137.7 (70.9)
`
`15
`120
`45–250
`52.4 (31.8)
`
`25
`50
`15–140
`
`14 (33)
`6 (14)
`3 (7)
`6.9
`42
`1
`7
`15
`19
`6.8 (6.3)
`3.6
`1.0–24.7
`1844 (1057)
`1714
`124–4714
`5.6 (1.7)
`7.0
`3.0–7.0
`
`8 (19)
`7,0,1,0,0
`35 (81)
`3,4,2,4,22
`
`50.3 (14.1)
`18–82
`22.4 (3.1)
`85
`17.5–29.8
`46 (54)
`
`29 (34)
`18 (21)
`9 (11)
`8 (9)
`3 (4)
`19 (22)
`6
`
`38
`
`16 (42)
`15 (40)
`5 (13)
`2 (5)
`
`49 (57)
`76.5 (64.4)
`
`80
`57.5
`5–343
`130.8 (75.0)
`
`28
`125
`40–343
`48.1 (31.6)
`
`47
`38
`5–140
`
`19 (22)
`14 (16)
`13 (15)
`7.4
`85
`1
`13
`32
`39
`6.3 (6.0)
`3.9
`1.0–25.8
`1887 (1036)
`1764
`124–5000
`5.8 (1.6)
`7.0
`3.0–7.0
`
`15
`11,1,1,1,1
`71
`5,6,4,10,46
`
`Page 5
`
`

`
`1478 JEPPESEN ET AL
`
`Table 3. Continued
`
`Concomitant medication
`Antidiarrheals, n (%)
`Antisecretory agents, n (%)
`
`GASTROENTEROLOGY Vol. 143, No. 6
`
`Placebo (n ⫽ 43)
`
`Teduglutide, 0.05
`mg/kg/d (n ⫽ 43)
`
`Overall (N ⫽ 86)
`
`P value
`
`16 (37)
`22 (51)
`
`22 (51)
`25 (58)
`
`38 (44)
`47 (55)
`
`BMI, body mass index.
`aP value is calculated using the general linear model method, with corresponding variables as dependent variables and treatment group as the
`independent variable.
`bP value is based on the ␹2 contingency table using the exact method.
`cP values for overall treatment comparisons are base on the Fisher exact test for categorical variables and on a 1-way analysis of variance with
`effect for treatment for continuous variable.
`
`CLINICAL AT
`
`Safety and Tolerability
`The safety population consisted of 85 patients who
`received ⱖ1 doses of study drug. Total exposure to study
`drug was similar in the 2 study groups. There were no
`deaths during the study. The number of patients with
`AEs, serious AEs, treatment-emergent AEs (TEAEs), or
`discontinuations due to treatment-emergent serious AEs
`was comparable between treatment groups. Treatment-
`emergent serious AEs were reported by 27 patients (15/42
`patients [36%] vs 12/43 patients [28%] in the teduglutide
`group vs placebo group, respectively). Sixty-nine patients
`were reported to have ⱖ1 TEAEs during the study (35/42
`[83%] vs 34/43 [79%]). Five patients were reported to have
`experienced TEAEs leading to discontinuation during the
`study (2/42 [5%] vs 3/43 [7%]); none were serious.
`The most frequently reported TEAEs in the teduglutide
`group were of gastrointestinal origin, such as abdominal
`pain, nausea, gastrointestinal stoma complication, or ab-
`dominal distension (Supplementary Table 2). Two treat-
`ment-emergent serious AEs reported in the teduglutide
`group were deemed related to the study treatment (acute
`cholecystitis and small intestinal stenosis); both resolved.
`No major findings were reported in the laboratory/
`chemistry or hematology tests of the teduglutide-treated
`vs placebo patients. Six patients from the teduglutide
`group and none from the placebo group developed anti-
`teduglutide antibodies after the start of study drug. An-
`
`Figure 2. Mean (SE) absolute reduction in parenteral support.
`
`tibodies were non-neutralizing and without evidence of
`decreased effect on PS volume reduction or evidence of
`systemic hypersensitivity or other clinically significant se-
`quelae.
`
`Discussion
`In this phase 3 study, teduglutide met the primary
`efficacy end point, defined as the percentage of patients
`who achieved a 20% to 100% reduction in weekly PS
`volume at weeks 20 and 24 compared with baseline. There
`were 27/43 (63%) responders in the teduglutide group and
`13/43 (30%) in the placebo group (P ⫽ .002). When
`addressing the proabsorptive effects of teduglutide, im-
`provements in intestinal fluid absorption under ideal cir-
`cumstances should be counterbalanced by equivalent
`PN/IV volume reductions.
`One of the observations in the previous phase 3 study
`was that, in addition to significant reductions in PS vol-
`ume in patients assigned to teduglutide 0.05 mg/kg/d
`(and nonsignificant reductions in those receiving the 0.10-
`mg/kg/d dosage), there were also small decreases in PS
`volume in patients randomized to placebo.12 This sug-
`gests that appropriate day-to-day fluid management with
`strict adherence to the weaning algorithm can have an
`intrinsic benefit and that treatment with teduglutide 0.05
`mg/kg/d provides an important additional benefit.
`Compared with the previous phase 3 study, this study
`protocol allowed earlier (after 2 weeks vs 4 weeks) and
`more aggressive (10%⫺30% vs 10%) PS volume reduc-
`tions.12 This was intended to minimize fluid accumula-
`tion (edema), reductions in oral intake, and increases in
`urine production encountered in the previous phase 3
`study in relation to treatment with teduglutide 0.10 and
`0.05 mg/kg/d, which negatively affected the primary study
`end point.12
`At week 24, after these protocol adjustments, a larger
`mean PS volume reduction, of 4.4 ⫾ 3.8 L/wk (32%
`change from baseline mean of 12.5 L/wk) was observed
`compared with a reduction of 2.5 ⫾ 2.3 L/wk (27% change
`from baseline mean of 9.6 L/wk) in the teduglutide
`0.05-mg treatment group in the previous study, which
`might explain this study’s higher responder rate. In this
`study, the effects of teduglutide on PS volume reduction
`more closely reflected the FCE (5.4 ⫾ 6.0 L/wk). The FCE
`
`Page 6
`
`

`
`CLINICALAT
`
`December 2012
`
`TEDUGLUTIDE AND SBS 1479
`
`Figure 3. Summary of changes
`in (upper left panel) oral fluid in-
`take, from baseline mean, 1888
`(SD 883) mL/d for teduglutide,
`1637 (603) mL/d for placebo; (up-
`per right panel) urine volume, from
`baseline mean 1350 (216) mL/d
`for teduglutide, 1372 (269) mL/d
`for placebo; (lower left panel) par-
`enteral volume,
`from baseline
`mean 1777 (1096) mL/d for te-
`duglutide, 1899 (1077) mL/d for
`placebo; and (lower right panel)
`fluid composite effect, from base-
`line mean 2317 (1651) mL/d for te-
`duglutide, 2176 (1270) mL/d for
`placebo. Mean (SE) values for
`changes are shown.
`
`at week 24 was in the same magnitude in both studies
`(4.4 ⫾ 7.3 L/wk).
`The high placebo response in this study (30% vs 5% in
`the previous study) might be explained by addressing the
`components of the FCE (Figure 3). In the previous study,
`at week 24, patients receiving placebo had no significant
`changes in the oral fluid intake or urine output volume
`during 48-hour measurements. Therefore, the PS volume
`reduction (0.9 ⫾ 1.4 L/wk) was in the same magnitude as
`the FCE (0.8 ⫾ 6.2 L/wk) in the previous study. In the
`current study, where protocol modifications encouraged
`earlier and more aggressive PS reductions, significantly
`larger PS reductions were also achieved in patients receiv-
`ing placebo (2.3 ⫾ 2.7 L/wk), but subsequently these
`patients had to increase their oral fluid intake signifi-
`cantly (1.6 ⫾ 3.6 L/wk; P ⫽ .009) to maintain urine
`production and hydration constant (Figure 3).
`At week 24, the FCE was 1.1 ⫾ 4.3 L/wk in this study in
`the patients receiving placebo, which closely reflected
`findings in the first phase 3 study. In addition, because
`the oral fluid intake and urine output were not constant
`over time, the observed effects of teduglutide on absolute
`and relative PS volume reduction might imprecisely esti-
`mate the true effect of teduglutide on intestinal fluid
`absorption. Therefore, the FCE might better illustrate the
`summarized effects of teduglutide as also demonstrated
`
`in the 3-week, phase 2, metabolic balance study in pa-
`tients with SBS, in which fecal wet weight excretion was
`reduced from baseline by 711 ⫾ 734 g/d (P ⫽ .001) and
`intestinal wet weight absorption was increased by 743 ⫾
`477 g/d (P ⬍ .001).11 Although FCE is not a standard
`measure, it can be of value in the clinical assessment of
`patients with SBS as another potential indicator, along
`with urine output alone or determination of enteral bal-
`ance (ie, oral fluid intake minus stool volume) when
`making weaning decisions.
`The PS volume reductions from baseline of 32% ⫾ 19%
`(4.4 ⫾ 3.8 L/wk) demonstrated in this study provide
`correspondingly less time connected to the central line, if
`the infusion rate is the same. Because patients with SBS
`can spend up to 16 hours per day connected to the central
`line, this will liberate considerable time for unhindered
`daytime activities or undisturbed sleep.16 It is appeal