609566

research-article2015
PENXXX10.1177/0148607115609566Journal of Parenteral and Enteral NutritionSigalet et al

Original Communication
Journal of Parenteral and Enteral
Nutrition
A Safety and Dosing Study of Glucagon-Like Peptide Volume XX Number X
Month 201X 19
2 in Children With Intestinal Failure 2015 American Society
for Parenteral and Enteral Nutrition
DOI: 10.1177/0148607115609566
jpen.sagepub.com
hosted at
online.sagepub.com
David L. Sigalet, MD, PhD1; Mary Brindle, MD, MPH1; Dana Boctor, MD2;
Linda Casey, MD3; Bryan Dicken, MD3; Sonia Butterworth, MD4;
Viona Lam, BScN1; Vikram Karnik, MD1; Elaine de Heuvel, BSc1;
Bolette Hartmann, PhD5; and Jens Holst, PhD5

Abstract
Background and Aims: A glucagon-like peptide 2 (GLP-2) analogue is approved for adults with intestinal failure, but no studies of
GLP-2 have included children. This study examined the pharmacokinetics, safety, and nutritional effects of GLP-2 in children with
intestinal failure. Methods: Native human GLP-2(1-33) was synthesized following good manufacturing practices. In an open-label trial,
with parental consent, 7 parenteral nutritiondependent pediatric patients were treated with subcutaneous GLP-2 (20 g/kg/d) for 3 days
(phase 1) and, if tolerated, continued for 42 days (phase 2). Nutritional treatment was directed by the primary caregivers. Patients were
followed to 1 year. Results: Seven patients were enrolled (age: 4.0 0.8 years; bowel length, mean SEM: 24% 4% of predicted). All
were parenteral nutrition dependent since birth, receiving 44% 5% of calories by parenteral nutrition. GLP-2 treatment had no effect
on vital signs (blood pressure, heart rate, and temperature) and caused no significant adverse events. Peak GLP-2 levels were 380 pM
(day 3) and 295 pM (day 42), with no change in half-life or endogenous GLP-2 levels. Nutritional indices showed a numeric improvement
in Z scores and citrulline levels; the Z score was maintained while citrulline levels returned to baseline once GLP-2 was discontinued.
Conclusions: GLP-2 was well tolerated in children, with a pharmacokinetic profile similar to that of adults. There were no changes in
endogenous GLP-2 release or metabolism. These results suggest that GLP-2 ligands may be safely used in pediatric patients; larger trials
are suggested to investigate nutritional effects. (JPEN J Parenter Enteral Nutr. XXXX;xx:xx-xx)

Keywords
pharmacokinetics; necrotizing enterocolitis; adaptation; dipeptidyl peptidase IV

Clinical Relevancy Statement
While intestinal failure is a common and serious problem in
pediatric patients, there are no pharmacologic therapies at
present that will increase intestinal nutrient absorption or has-
ten the process of intestinal adaptation. Glucagon-like peptide
2 (GLP-2 [1-33]) is an enterotropic hormone that induces
many of the findings of adaptation, and a long-acting ana-
logue been licensed for use in adults with intestinal failure. No
previous studies have examined the clinical or pharmacologic
kinetics of GLP-2 in the pediatric population. This study
shows that treatment with exogenous native human GLP-2
therapy was well tolerated in this population of parenteral
nutritiondependent children. There were no episodes of fluid
retention or edema, as seen in adults. The pharmacokinetics of
GLP-2 in this population were similar to that reported in
adults but with possibly a more rapid breakdown. Giving
these children exogenous GLP-2 did not change the produc-
tion of endogenous GLP-2; we also found that the production
of endogenous GLP-2 was significantly reduced from that
seen in normal children. The implications for clinical practice
are that GLP-2 (and potentially GLP-2 analogues) may safely
be considered for expanded clinical trials in children, poten-
tially infants, with intestinal failure. Guidelines for a phase III
trial are suggested, which would include a forced reduction in
From 1Pediatric Surgery, Alberta Childrens Hospital, University
of Calgary, Calgary, Canada; 2Pediatric Gastroenterology, Alberta
Childrens Hospital, University of Calgary, Calgary, Canada; 3Surgery
(Pediatric), Stollery Childrens Hospital/University of Alberta,
Edmonton, Canada; 4Pediatric Surgery, British Columbia Childrens
Hospital, Vancouver, Canada; and the 5NNF Center for Basic
Metabolic Research, Department of Biomedical Sciences, University of
Copenhagen, Copenhagen, Denmark.
Financial disclosure: These studies were supported by the Professorship
in Pediatric Surgical Research of the Alberta Childrens Hospital
Research Foundation and by Sidra Medical and Research Center, Doha,
Qatar.
Received for publication June 25, 2015; accepted for publication August
31, 2015.
Corresponding Author:
David Sigalet, MD, PhD, Department of Surgery, Alberta Childrens
Hospital/University of Calgary, 2888 Shaganappi Trail NW, Calgary, AB
T3B 6A8, Canada.
Email: dsigalet@sidra.org

Downloaded from pen.sagepub.com at UNIVERSITAET OSNABRUECK on February 3, 2016

2 Journal of Parenteral and Enteral Nutrition XX(X)
parenteral nutrition therapy, a longer treatment period, and a
crossover study design, to control for endogenous adaptation
during the study period.
Introduction
The ability to provide nutrition support in the form of paren-
teral nutrition (PN) has brought about a revolution in the
care of infants and children with intestinal disease.1,2
However, the care of these patients, at present, is primarily
supportive; no pharmacologic therapies exist that can
improve the function of the residual intestine.3,4 This process
of increasing the nutrient-absorptive capacity of the remnant
intestine is known as adaptation, which occurs naturally in
response to enteral nutrient stimulation. Glucagon-like pep-
tide 2 (GLP-2 [1-33]) is a highly conserved enteroendocrine
hormone that appears to be a regulator within this system.5,6
It is synthesized by the enteroendocrine L-cells of the small
intestine, which are most numerous in the terminal ileum.
GLP-2 is released in response to undigested nutrients, espe-
cially free long-chain fatty acids, in the intestinal lumen and
thus is an endogenous signal of insufficient nutrient-absorp-
tive capacity.4,5 Mechanistically, GLP-2 activates the GLP-2
receptor, which is primarily expressed on enteroendocrine
cells, the enteric neuronal system, and myofibroblasts of the
intestine and not on the epithelium directly.7 GLP-2 acts
acutely to slow proximal motility, increases mesenteric
blood flow, reduces gastric secretions, and over time acts as
a trophic agent for the small intestinal mucosa, improving
nutrient absorption.7-9 In enterally fed animals subjected to
massive resection, GLP-2 levels were highly correlated with
spontaneous adaptation, suggesting that GLP-2 plays an
active role in regulating this response.9 In models of massive
small bowel resection, maintained on PN, treatment with
GLP-2 prevents body weight loss and results in maintenance
of bowel mass and improved villous and crypt architecture,
nutrient transporter expression, and nutrient absorption.10-12
This can be thought of as a feedback system or axis, simi-
lar to the insulin-glucose axis.
In adult human studies, GLP-2 and its analogue teduglutide
have been shown to improve nutrient and fluid absorption.13-15
In infants, the GLP-2 axis of enteroendocrine cell production
of GLP-2 in response to feeds is highly active; postprandial
levels of GLP-2 are very high (up to 450 pM/L) but low in
infants not tolerating feeds.16 In infants with SBS, serum
GLP-2 concentrations correlate with residual small intestinal
length, intestinal absorption, and ultimately survival.17 Given
that infants with intestinal failure from either anatomic short
bowel syndrome or gastroschisis are at increased risk of devel-
oping long-term intestinal failure and have low levels of GLP-
2, we hypothesized that exogenous GLP-2 therapy would
improve intestinal function in this population. However, there
are no studies examining the safety, metabolism, or physio-
logic effects of GLP-2 in the pediatric population.
Downloaded from pen.sagepub.com at UNIVERSITAET OSNABRUECK on February 3, 2016
The present study was proposed as a phase III trial. The
primary aim was to examine the safety and pharmacokinetics
of native GLP-2 in children with anatomic short bowel syn-
drome or intestinal failure. Because the pharmacokinetics are
unknown in the pediatric population, this study focused on
children >1 year of age. The dose used was 20 g/kg, based on
adult studies and a preliminary study done in weaning piglets
that showed that pharmacokinetics and metabolism in this
model were similar to those seen in adult humans.18 The sec-
ondary aim of the study was to obtain preliminary data to
determine the effect size on the nutrition effects of GLP-2 ther-
apy, to guide future investigations.
Methods
With institutional ethic board approval (NCT01573286; Health
Canada Reference GLP-2-01 150979; University of Calgary
Conjoint Health Ethics Board 21691), families of children
being supported with long-term PN were approached to par-
ticipate in the study. The inclusion criteria specified that
patients needed to be >1 year old but <18 years, with intestinal
failure postsurgery (resection or repair of gastroschisis).
Intestinal failure was defined as a requirement for >30% of
calories by PN >3 months from the last intestinal surgery.3 The
exclusion criteria were related to comorbidities (either sys-
temic diseases or intestinal mucosal) that would interfere with
the potential for the recovery of normal intestinal function.
Inclusion and Exclusion Criteria
Inclusion criteria included the following:
Patients >1 year of age (corrected gestational age) but
<18 years old
Anatomic SBS, with <40% of expected bowel length
and a requirement for >30% of calories by PN, >3
months (90 days) from last intestinal surgery; or gas-
troschisis, with a requirement for >30% of calories by
PN and >3 months from last intestinal surgery
Exclusion criteria included the following:
Significant extraintestinal disease (eg, grade IV intraven-
tricular hemorrhage, severe hypoxic encephalopathy)
Significant cardiovascular, hemodynamic, or respira-
tory instability as noted by requirement for dopamine
>4 mcg/kg/min, high-frequency ventilatory support,
extracorporeal membrane oxygenation
Hepatic disease defined as direct bilirubin >100 mol/L
(5.2 mg/dL)
Renal disease defined as blood urea nitrogen >80 or
creatinine >90 mol/L (1.5 mg/dL)
Inborn errors of metabolism necessitating protein
restriction or other special diet
Sigalet et al 3
Figure 1. Timeline of study investigations and interventions.
Vertical arrows indicate timing of blood sampling for glucagon-
like peptide 2 (GLP-2) and citrulline levels s.c., subcutaneous.
Ongoing sepsis syndrome, as noted by refractory
hypotension, thrombocytopenia, acidosis, and/or
bacteremia
Primary motility defect, such as intestinal
pseudo-obstruction
Absorptive defects (eg, microvillus inclusion disease)
Females who are postpubertal must agree to comply
with measures to prevent pregnancy during the study
phase
Coagulopathy that precludes the use of subcutaneous
injections
Allergy to GLP-2 or any of the constituents of the
GLP-2 IC-115 preparation
Subjects
Subjects were primarily home PN patients; during the trial,
they were treated following the standard nutrition care protocol
of the institutions intestinal rehabilitation team. Patients were
enrolled at 3 sites, following a common GLP-2 treatment pro-
tocol. After the study, patients were followed according to their
local protocols, with a clinic visit for study-specific review at
1, 6, and 12 months posttherapy. At enrollment, demographics
and anatomic data were extracted from the chart, including
operative measurements of bowel length, and nutrition param-
eters, including monitoring of liver function and routine PN
bloodwork (complete blood counts, electrolytes, creatinine,
urea, aspartate transaminase, alanine transaminase, bilirubin,
gamma-glutamyl transferase, protein, and albumin levels; see
Figure 1). During the study, there was no attempt to change the
nutrition support of the patients, aside from the expected
advancement of feeds mandated by the primary care team.
Other medications, including the use of antibiotics and motility-
enhancing agents, were also controlled by the primary care
team, independent of the study.
Downloaded from pen.sagepub.com at UNIVERSITAET OSNABRUECK on February 3, 2016
Study Design
A quasi-experimental interrupted time series design was used
with a convenience sample. The therapeutic agent, native human
GLP-2, was produced as a lyophilized powder by solid-state syn-
thesis (CS Bio, Menlo Park, CA; >98% purity) and prepared as a
sterile solution in alkalinized saline (0.9% NaCl alkalinized to pH
8.59.5 by addition of 0.05M NaOH). Powdered GLP-2 was dis-
solved in individual vials (1.5 mg in 1.5 mL) following good
manufacturing practices by the experimental therapeutics pro-
gram of the British Columbia Cancer Agency, Vancouver, Canada
(lot IC115, May 2009). The peptide was stored as a frozen solu-
tion at 20C until used. Stability was demonstrated by repeated
testing with an identical mixing-and-freezing sequence, with
mass spectroscopy at 6-month intervals following manufacture,
with minimal loss of potency (>90% of nominal peptide concen-
tration when tested out to December 2015; testing done by
Maxxam Corporation, Burnaby, Canada).
Phase I
Phase I was the initial 3 days of GLP-2 administration, to exam-
ine the safety and pharmacokinetics of GLP-2. Plasma citrulline
levels and the production of endogenous GLP-2 (fasting and
postprandial) were assessed before the trial. The dose required to
deliver 20 g/kg/d of GLP-2 subcutaneously (n = 7; Figure 1)
was determined on a weekly basis for each patient based on the
measured body weight and by adding one half of the body
weight gained over the previous week. The appropriate doses of
compound were prepared by the central pharmacy of the
University of Calgary, in 1-mL syringes, and then refrozen.
Individual doses were thawed within 1 hour of use; all injections
were given with a 0.22-m filter (Millipore, Etobicoke, Canada)
and a 23-gauge needle (Becton Dickinson, Mississauga, Canada)
via a subcutaneous Insuflon catheter (Unomedical, Lejre,
Denmark). Injection site preparation was done following stan-
dard protocols for administration of subcutaneous medications;
sites were rotated at least weekly. To compensate for the dead
space of the filter and the needle, each injection of GLP-2 was
followed by a flush of 0.6 mL of alkalinized saline (pH 89),
also prepared by the experimental pharmacy. Patients were
injected twice daily at approximately 0800 and 1700, typically
before meals. During the initial 3 days of the treatment, vital
signs were assessed before the injection and at 15, 30, 60, 120,
and 180 minutes following it. The childs temperature, appetite,
nutrition indices inclusive of general well-being, caloric intake,
eating, stooling patterns, complaints of nausea, and inspection of
the injection site were monitored daily to provide information
about the physiologic effects and safety of the therapy. This
included weight, Z scores, tolerance of enteral nutrition, require-
ment for PN, and stool output. Biochemical parameters included
electrolytes, creatinine, and liver function studies, following the
protocol of the PN monitoring in the 7 days prior to the study. A
pharmacokinetic study was done on day 3, modified for use in
4 Journal of Parenteral and Enteral Nutrition XX(X)

children to minimize the volume of blood required. Samples Table 1.Demographics.

were taken prior to GLP-2 administration and at 60, 90, and 180
Parameter n or Mean SEM
minutes postinjection. Blood was drawn with a microcollection
technique (0.5 mL/sample), immediately aliquoted into tubes Sample size 7
containing diprotin A (0.1 mM; Sigma-Aldrich Inc, St Louis, Birth weight, g 1960 1060
MO) and aprotinin (500 KU/mL blood; Bayer Inc, Toronto, Gestational age
Canada), and kept on ice until processed. Samples were centri- At birth, wk 32 4
fuged at 2500 g for 10 minutes at 4C and the plasma stored at At surgery, wk 42 16
80C until analysis. Plasma GLP-2 concentrations were mea- Small bowel length
sured with a GLP-2-specific radioimmunoassay with an intra- At final surgery, cm 55.9 10.0
assay coefficient of variation of 5% as previously described.19 Predicted at final surgery, cm 360.4 52.3
At final surgery, % predicted 15.7 3.2
Serial transverse enteroplasty 5
Phase II Ileocecal valve resected 6
Once children tolerated phase I, they moved into phase II and Colon resected 3
Diagnosis
continued to receive therapy for a further 39 days. Vital signs
Gastroschisis 0
and monitoring of systemic end points (appetite, pain or nau-
Atresia/atresia + gastroschisis 3
sea following injection, general well-being, and stooling pat-
Necrotizing enterocolitis 3
terns) were assessed as the patients returned for their routine
Other (embolic infarct) 1
follow-up, typically done biweekly. Weights and nutritional
intake were recorded by the families using a diary twice
weekly. All adverse events were reviewed by a data safety Table 2. Vital Signs Postinjection.a
monitoring board, which examined the potential causality of
Vitals: Minutes
GLP-2 therapy to the adverse event. The ongoing participa-
Postinjection Day 1 Day 2 Day 3
tion of the patient in the study was at the discretion of the
board. Plasma citrulline levels and the production of endoge- Temperature, C
nous GLP-2 (fasting and postprandially) with a repeat phar- 0 36.8 0.3 37.0 0.0 36.7 0.5
macokinetic profile were evaluated in the final week. 15 36.9 0.2 36.8 0.1 36.8 0.0
60 37.0 0.3 37.1 0.2 36.9 0.1
Heart rate, beats/min
Poststudy Monitoring 0 122 27 128 31 132 26
Patients were typically seen at monthly intervals by their primary 15 124 10 130 5 132 12
care team. Study-specific repeat assessments of physiologic and 60 127 0 122 9 123 0
nutrition parameters, biochemical indices, citrulline levels, and Respiratory rate, breaths/min
postprandial GLP-2 levels were repeated at 1, 6, and 12 months. 0 43 6 42 11 40 13
15 42 1 41 4 45 1
60 39 3 38 4 36 2
Statistics and Data Analysis Mean arterial pressure,b mm Hg
0 61 6 65 13 67 22
Results were collected during the patient visit and recorded in a 15 66 3 66 2 61 4
secure database. Descriptive statistics with mean SEM were cal- 60 61 9 59 6 62 10
culated to provide information about the primary end points: phar-
a
macokinetic and safety results related to the therapy. A repeated For each determination, n = 7. Values are shown for 060 minutes for
days 03 of the study. No further changes were seen in the values out to
measures 1-way analysis of variance was used to compare the val- 180 minutes after injection and in the repeated assessments to the 1-year
ues of the dependent variables after GLP-2 therapy and then again follow-up. Values are presented as mean SEM.
b
after the 1-month washout phase. Post hoc comparisons based on Mean arterial pressure = [(2 diastolic) + systolic] / 3.
Bonferronis test were performed. An alpha level of 0.05 for sig-
nificance and 0.1 for a trend to significance were used for all analy- intestinal failure from birth; the majority were on home PN
ses. The test for homogeneity of variance was completed to test the and being treated as outpatients (Table 1). The requirement
assumptions underlying the application of analysis of variance. for PN had been stable for at least 3 months in all patients
prior to enrollment. The results show that GLP-2 therapy was
well tolerated. There were no changes in physiologic param-
Results eters, including heart rate, blood pressure, or temperature
A total of 7 subjects were enrolled over a 13-month period. with injection in any patient throughout the study (Table 2).
As anticipated, the patient population was one of severe The data for the first 3 days of treatment are shown; there

Downloaded from pen.sagepub.com at UNIVERSITAET OSNABRUECK on February 3, 2016

Sigalet et al 5

Table 3. Liver and Renal Function.a

Study Day Months Poststudy

Parameter Reference16 0 14 28 42 1 6 12
Endogenous GLP-2,
pM
Fasting 17 5 17 8 N/A N/A 11 7 N/A N/A N/A
Postprandial 72 8 33 5 N/A N/A 31 10 36 8 35 7 34 5
Hgb, g/L 106225 116 9 100 8 110 10 100 7.1 122 5 110 14 120 16
Hct, L/L 0.310.55 0.32 0.06 0.30 0.06 0.32 0.04 0.30 0.04 0.36 0.07 0.34 0.05 0.39 0.04
WBC, 109/L 5.019.0 8.1 3.3 7.3 1.2 6.9 1.1 5.7 0.9 7.8 1.2 8.8 4.9 7.1 2.3
Platelets, 109/L 150400 249 24 218 24 240 22 201 24 199 25 183 83 203 42
Creatinine, mol/L 2060 29 3.4 25 1.8 26 2 24 3 32 3 31 3 28 3
AST, U/L 1065 46 8 55 9 66 12 57 8 55 8 58 11 64 10
ALT, U/L 135 69 19 53 11 72 18 63 16 63 12 68 14 74 18
Bilirubin, total, 023 3.3 2.2 7.2 1.5 3.9 1.3 7.5 1.7 10.0 2.2 6.7 1.2 19 7
mol/L
GGT, U/L 863 62 31 56 32 56 21 65 20 61 31 64 28 56 16
Total protein, g/L 5479 65 8 60.2 2.2 62 3 57 2 65 2 53 9 61 0
Albumin, g/L 3050 38 3 37 3 39 3 36 2 37 3 38 5 34 6
Citrulline, mol/L 635 17.1 3.0 N/A N/A 19.7 2.8 15.2 2.3 18.3 2.5 17.3 2.5

ALT, alanine transaminase; AST, aspartate transaminase; GGT, gamma-glutamyl transferase; Hct, hematocrit; Hgb, hemoglobin; N/A, not applicable;
WBC, white blood cell.
a
Values are presented as either range (for reference values) or mean SEM (for study data).

were no changes over the remaining year of the study period.
No patients had a stoma or an operative procedure while in
the study; thus, no mucosal sampling was possible during the
phase of GLP-2 therapy.
The injections were well tolerated, and the Insuflon injec-
tion system worked well. No child developed sensitivity to the
GLP-2 preparation or other problems at the injection sites.
The parents did not report any consistent changes in mood or
appetite with the treatment; however, several families reported
an increased interest in oral feeding during the study period,
which reverted to baseline once the study was completed. This
information was spontaneously reported by parents but was
not quantified, nor was this a primary outcome measure of the
study.
Baseline fasting GLP-2 levels were within normal limits,
but endogenous postprandial levels were low at the beginning
of the study; neither fasting nor postprandial levels were
changed from day 0 to week 6 (Table 3). GLP-2 levels were
significantly increased by the treatment (P < .001); the peak
levels at day 3 and day 42 were similar (380 76 vs 295 57
pM/L, P = .34 by Students t test for paired data), and the
apparent half-life of 80 minutes was unchanged over the course
of the study (Figure 2).
There was a numeric but nonsignificant (P = .14) increase
in the Z scores for weight and a decrease in the number of
stools per day (P = .47), which was maintained once the ther-
apy was discontinued (Figure 3). There were also trends
toward improvements in the citrulline levels with treatment,
which reverted to baseline at 1-month posttherapy (P = .56).
There was no apparent change in the proportion of calories
taken enterally or parenterally (Figure 4); similarly, there were
no changes in electrolytes (not shown), creatinine, liver func-
tion tests, serum protein, or albumin levels (Table 3).
Three patients experienced adverse events during the study,
which were judged as serious; 1 child received twice the
desired dose for 2 days due to a medication administration
error while in hospital. The drug was stopped for 2 days and
then resumed, with no detectable changes in in clinical status
or PN bloodwork. Two other children had undergone a barium
study just prior to initiating the study; both were admitted on
day 3 of the study after the preliminary bloodwork due to vom-
iting of residual barium. This settled after 24 hours, and the
drug was continued without any further episodes of vomiting.
No event was judged to be directly caused by the treatment
with GLP-2. There were no other serious adverse events.
Discussion
This study shows, for the first time in a pediatric population,
that treatment with GLP-2 over a 6-week period is safe, with
a pharmacokinetic profile similar to that seen in adults.
While the trial was not designed or powered to show changes
in nutrition status, there were encouraging suggestions that
GLP-2 may have a clinical effect, as demonstrated by Z
scores, and a possible trophic effect on the gastrointestinal
mucosa, as evidenced by the increase in citrulline levels.
There were no measurable changes in vital signs at any
point throughout the trial, demonstrating that the therapy does

Downloaded from pen.sagepub.com at UNIVERSITAET OSNABRUECK on February 3, 2016

6 Journal of Parenteral and Enteral Nutrition XX(X)
Figure 2. Pharmacokinetic (PK) studies done at (a) 3 days
and (b) 42 days of therapy. Mean of plasma concentrations of
intact glucagon-like peptide 2 (GLP-2 [1-33]) after injection
of subcutaneous GLP-2 (10 g/kg), measured at the following
times: 0 (preinjection), 60, 90, and 120 minutes after injection.
Plasma GLP-2 levels in picomolar (pM) as determined by
radioimmunoassay.18 Mean SEM, n = 7 at each time point.
not have cardiovascular or central nervous system effects in
this population of children. Additionally, there were no
adverse events attributed to the GLP-2 therapy. Direct obser-
vations of the effects on the mucosa were not possible, because
no patients in this study had a stoma. This population would
be expected to have multiple intra-abdominal adhesions,
which could cause pain if motility patterns or the physical size
of the bowel changed quickly as a result of the GLP-2 therapy.
Acute volume changes in the intestine can be due to the acute
increase in blood flow that GLP-2 induces or to a change in
motility that might affect production or passage of luminal
gas. Chronic changes may be due to an actual increase in
mucosal mass.15 No complaints of cramping or vomiting
(aside from the initial barium-related episodes in 2 patients)
Downloaded from pen.sagepub.com at UNIVERSITAET OSNABRUECK on February 3, 2016
were noted in any patient. Although this is difficult to quan-
tify, this suggests that there were no major changes in intesti-
nal volume due to either acute or chronic actions of GLP-2.
There was no evidence of any patient having peripheral edema
or any other signs of fluid retention. It is interesting to note
that, in contrast to the adult patients treated with GLP-2
ligands, the baseline creatinine levels in these pediatric
patients were normal, and there was no change during therapy
with GLP-2.15 There were also no changes in any of the hema-
tologic or liver function values.
There are important observations related to the pharmacoki-
netics of GLP-2. These results show that the distribution and
metabolism of GLP-2 in these pediatric patients are similar to
those reported by Hartmann et al in healthy adults.20 However,
the peak levels may be lower in infants: the average peak level
in the Hartmann study was approximately 1400 pM, after an
injection of 8 g/kg, while in the current study, peak levels
were in the order of 400 pM after receiving 10 g/kg.
Furthermore, the apparent half-life after injection in the adult
studies was ~120 minutes, while in the current study, it was
~80 minutes.20 Although the findings of the current study are
limited by the restricted sampling that was possible in these
small patients, the dose response suggests more rapid clear-
ance of GLP-2 delivered by subcutaneous injection in pediatric
recipients. These differences may be due to variations in sub-
cutaneous blood flow or the difference in the formulation and
the carrier solvent for the GLP-2 used in these 2 studies. It is
unlikely that the differences were due to the assay; the system
was the same for both studies.
Results related to the pharmacokinetics of GLP-2 suggest a
difference in the volume of distribution and metabolism of the
GLP-2 between adults and children; however, this will require
further direct study. These findings also emphasize our limited
understanding of the activity of the dipeptidyl peptidase IV
enzymatic system in the pediatric population.21 As more drugs
and hormones are developed that are metabolized via this
pathway, there will be a need to study this pathway in greater
detail in the pediatric population with a more rigorous research
design.
An additional end point examined was the interaction of
exogenous GLP-2 administration on the production of endog-
enous GLP-2. There was no effect on either fasting or post-
prandial GLP-2 levels over the course of the study; this
suggests that there is not a significant downregulation of
endogenous GLP-2 production by supraphysiologic levels
given exogenously. It is also interesting that none of these
patients produced >43 pM of endogenous GLP-2 (with the
average being 33 pM) following full meal stimulation. This is
similar to the level previously reported as a discriminating
level for PN independence in infants; it is notably less than the
postprandial levels seen in either normal children or children in
the postresection period of adaptation.16 This adds further sup-
port to the association between the ability to produce GLP-2
Sigalet et al 7

Figure 3. Nutrition indices vs time: (a) weight, kg; (b) Z score, World Health Organization methodology; (c) stool frequency, per day;
(d) plasma citrulline, mol/L.

and the process of spontaneous adaptation. However, the small
number of patients in any 1 anatomic subtype precludes any
correlation between the efficacy of exogenous GLP-2 (or the
production of endogenous GLP-2) and the specific segment of
intestine remaining. As well, these children were all at least 1
year old so that the influence of gestational age at time of birth
and resection on GLP-2 responsiveness could not be assessed.
However, all of these factors may be relevant and should be
included as variables in planning studies of efficacy and as a
consideration of a crossover study design, to control for spon-
taneous adaptation that may occur during the study period.
A trophic effect on the intestinal mucosa is suggested by
the trend of an increase in citrulline levels, which then
reversed after cessation of therapy; note that the citrulline
levels at baseline in this pediatric population were very sim-
ilar to those in adults requiring long-term PN.13 Citrulline
levels have been shown to be well correlated with mucosal
mass and less well correlated with adaptation.22 No patients
had procedures or stomas available for biopsy in this cohort;
in future studies, this would be a valuable end point to fol-
low. In terms of future studies, careful evaluations of nutri-
tion requirements, absorption, and growth over a longer
time course, with protocol-driven reductions in PN if enteral
nutrition tolerance improves, will be required to understand
the nutrition effects in this population. Furthermore,
although this was a stable population and remained so out to
1 year after the study, in pediatric SBS patients there is an
expected spontaneous but gradual improvement in nutrient
absorption over time. Thus, to show a specific effect of
GLP-2 or any proadaptive therapy will require an appropri-
ate study design. This optimally should include a crossover
from a treatment to an observation phase and sufficient sub-
jects to control for spontaneous adaptation.23 The current
study was also limited by the modest supply of GLP-2 avail-
able and the lack of long-term stability data. This dictated
both the number of patients that could be enrolled and the
number of dosing regimens studied, only allowing 1.
Nonetheless, it does provide a baseline for future pediatric
studies, with native GLP-2 or longer-lasting ligands, as well
as other hormones that are metabolized by the dipeptidyl
peptidase IV pathway.
The fact that all patients in the present study remained on
an almost identical proportion of calories via PN at the 1-year
follow-up as they were on at the beginning of the study speaks

Downloaded from pen.sagepub.com at UNIVERSITAET OSNABRUECK on February 3, 2016

8 Journal of Parenteral and Enteral Nutrition XX(X)

12 days per week, with a concomitant effect on quality-of-

life measures.
This study design was limited by the relatively short dosing
period of 6 weeks. The short dosing period was, in turn, a regu-
latory limitation imposed by the length of the supporting large
animal dosing study.18
The emergence of this population of severe SBS pediatric
patients who are stable but very dependent on home PN ther-
apy is a tribute to the improvements in nutrition care of this
population; in the past, many died due to PN complications.24
The challenge for the pediatric community is to develop thera-
pies that will improve the adaptation of the remnant intestine.
The findings of this study suggest that GLP-2 therapy is safe in
the pediatric population >1 year of age. Future studies are indi-
cated to determine if there is a dose-response relationship.
Having established an appropriate dosing range, long-term
studies with appropriate randomization and blinding are sug-
gested, to establish the potential utility in this most vulnerable
population.

Acknowledgments
The support of the staff and families of the intestinal failure units
of the respective institutions is gratefully acknowledged. Our
inspiration is the spirit and resolve of the families who helped us
design, execute, and complete all aspects of the study.

Statement of Authorship
D. L. Sigalet, M. Brindle, D. Boctor, V. Lam, B. Hartmann, and
J. Holst contributed to the design of the research; D. L. Sigalet,
M. Brindle, D. Boctor, L. Casey, B. Dicken, S. Butterworth, V.
Lam, B. Hartmann, and J. Holst contributed to the conception of
the research; D. L. Sigalet, M. Brindle, J. Holst, and E. de Heuval
contributed to the interpretation of the data; and D. L. Sigalet, M.
Brindle, D. Boctor, L. Casey, B. Dicken, S. Butterworth, V. Lam,
V. Karnik, E. de Heuval, B. Hartman, and J. Holst, contributed to
acquisition and analysis of the data. All authors drafted the manu-
script, critically revised the manuscript, agree to be fully account-
able for ensuring the integrity and accuracy of the work, and read
and approved the final manuscript.

References
Figure 4. Nutrient intake: (a) total calories administered, kcal/
kg/d; (b) enteral nutrition (EN) as percentage total calories; (c)
parenteral nutrition (PN) as percentage total calories.
to the stability of this cohort. It also reinforces that the primary
indication for PN in this population is nutrition and not fluid
related. Thus, if GLP-2 is not able to improve the functional
absorption of nutrients (fat, protein, carbohydrate), then it will
not likely have much benefit in the pediatric population. This
is in contrast to the findings in adults, where a significant
improvement in fluid balance can reduce PN dependency by
1. Dudrick SJ, Wilmore DW, Vars HM, Rhoads JE. Long-term total paren-
teral nutrition with growth, development, and positive nitrogen balance.
Surgery. 1968;64(1):134-142.
2. Kocoshis SA. Medical management of pediatric intestinal failure. Semin
Pediatr Surg. 2010;19(1):20-26.
3. Squires RH, Duggan C, Teitelbaum DH, et al. Natural history of pedi-
atric intestinal failure: initial report from the Pediatric Intestinal Failure
Consortium. J Pediatr. 2012;161(4):723-728.
4. Martin GR, Beck PL, Sigalet DL. Gut hormones, and short bowel syn-
drome: the enigmatic role of glucagon-like peptide-2 in the regulation of
intestinal adaptation. World J Gastroenterol. 2006;12(26):4117-4129.
5. Orskov C, Holst JJ, Knuhtsen S, Baldissera FG, Poulsen SS, Nielsen
OV. Glucagon-like peptides GLP-1 and GLP-2, predicted products of the
glucagon gene, are secreted separately from pig small intestine but not
pancreas. Endocrinology. 1986;119(4):1467-1475.

Downloaded from pen.sagepub.com at UNIVERSITAET OSNABRUECK on February 3, 2016

Sigalet et al 9
6. Brubaker PL, Anini Y. Direct and indirect mechanisms regulating secre-
tion of glucagon-like peptide-1 and glucagon-like peptide-2. Can J
Physiol Pharmacol. 2003;81(11):1005-1012.
7. Drucker DJ, Yusta B. Physiology and pharmacology of the enteroendocrine
hormone glucagon-like peptide-2. Annu Rev Physiol. 2014;76:561-583.
8. Meier JJ, Nauck MA, Pott A, et al. Glucagon-like peptide 2 stimulates
glucagon secretion, enhances lipid absorption, and inhibits gastric acid
secretion in humans. Gastroenterology. 2006;130(1):44-54.
9. Martin GR, Wallace LE, Hartmann B, et al. Nutrient-stimulated GLP-2
release and crypt cell proliferation in experimental short bowel syndrome.
Am J Physiol Gastrointest Liver Physiol. 2005;288(3):G431-G438.
10. Martin GR, Wallace LE, Sigalet DL. Glucagon-like peptide-2 induces
intestinal adaptation in parenterally fed rats with short bowel syndrome.
Am J Physiol Gastrointest Liver Physiol. 2004;286(6):G964-G972.
11. Sangild PT, Tappenden KA, Malo C, et al. Glucagon-like peptide 2
stimulates intestinal nutrient absorption in parenterally fed newborn pigs.
J Pediatr Gastroenterol Nutr. 2006;43(2):160-167.
12. Suri M, Turner JM, Sigalet DL, et al. Exogenous glucagon-like pep-
tide-2 improves outcomes of intestinal adaptation in a distal-intestinal
resection neonatal piglet model of short bowel syndrome. Pediatr Res.
2014;76(4):370-377.
13. Jeppesen PB, Sanguinetti EL, Buchman A, et al. Teduglutide (ALX-
0600), a dipeptidyl peptidase IV resistant glucagon-like peptide 2 ana-
logue, improves intestinal function in short bowel syndrome patients. Gut.
2005;54(9):1224-1231.
14. Jeppesen PB, Pertkiewicz M, Messing B, et al. Teduglutide reduces need
for parenteral support among patients with short bowel syndrome with
intestinal failure. Gastroenterology. 2012;143(6):1473-1481.
Downloaded from pen.sagepub.com at UNIVERSITAET OSNABRUECK on February 3, 2016
15. Jeppesen PB, Lund P, Gottschalck IB, et al. Short bowel patients treated
for two years with glucagon-like peptide 2: effects on intestinal morphol-
ogy and absorption, renal function, bone and body composition, and mus-
cle function. Gastroenterol Res Pract. 2009;2009:616054.
16. Amin H, Holst JJ, Hartmann B, Wallace L, Wright J, Sigalet DL.
Functional ontogeny of the proglucagon-derived peptide axis in the pre-
mature human neonate. Pediatrics. 2008;121(1):e180-e186.
17. Sigalet DL, Martin G, Meddings J, Hartman B, Holst JJ. GLP-2 levels in
infants with intestinal dysfunction. Pediatr Res. 2004;56(3):371-376.
18. Sigalet DL, de Heuvel E, Wallace L, et al. Effects of chronic glucagon-
like peptide-2 therapy during weaning in neonatal pigs. Regul Pept.
2014;188:70-80.
19. Hartmann B, Johnsen AH, Orskov C, Adelhorst K, Thim L, Holst JJ.
Structure, measurement, and secretion of human glucagon-like peptide-2.
Peptides. 2000;21(1):73-80.
20. Hartmann B, Harr MB, Jeppesen PB, et al. In vivo and in vitro deg-
radation of glucagon-like peptide-2 in humans. J Clin Endo Metab.
2000;86:2884-2888.
21. Holst JJ. Glucagon and glucagon-like peptides 1 and 2. Results Probl Cell
Differ. 2010;50:121-135.
22. Bailly-Botuha C, Colomb V, Thioulouse E, et al. Plasma citrulline con-
centration reflects enterocyte mass in children with short bowel syndrome.
Pediatr Res. 2009;65(5):559-563.
23. Peretti N, Loras-Duclaux I, Kassai B, et al. Growth hormone to improve
short bowel syndrome intestinal autonomy: a pediatric randomized open-
label clinical trial. JPEN J Parenter Enteral Nutr. 2011;35(6):723-731.
24. Sigalet D, Boctor D, Brindle M, Lam V, Robertson M. Elements of suc-
cessful intestinal rehabilitation. J Pediatr Surg. 2011;46(1):150-156.