Sunteți pe pagina 1din 12


PENXXX10.1177/0148607115626921Journal of Parenteral and Enteral NutritionDiamond et al

Original Communication
Journal of Parenteral and Enteral
Preventing the Progression of Intestinal FailureAssociated Volume 41 Number 5
July 2017 866877
Liver Disease in Infants Using a Composite Lipid Emulsion: 2016 American Society

A Pilot Randomized Controlled Trial of SMOFlipid for Parenteral and Enteral Nutrition
DOI: 10.1177/0148607115626921

Ivan R. Diamond, MD1; Robert C. Grant, MD1; Paul B. Pencharz, MD1;

Nicole de Silva, RN1; Brian M. Feldman, MD1; Peter Fitzgerald, MD2;
David Sigalet, MD3; Bryan Dicken, MD4; Justine Turner, MD4;
Valerie Marchand, MD5; Simon C. Ling, MD1; Aideen M. Moore, MD1;
Yaron Avitzur, MD1; and Paul W. Wales, MD1

Background: To examine whether SMOFlipid prevents progression of intestinal failureassociated liver disease (IFALD) in parenteral
nutrition (PN)dependent infants with early IFALD (conjugated bilirubin 1750 mol/L, 1-3 mg/dL). Study Design: Pilot multicenter
blinded randomized controlled trial comparing SMOFlipid with Intralipid. Patients received the trial lipid for up to 12 weeks, unless they
achieved full enteral tolerance sooner. The primary clinical outcome was the serum conjugated bilirubin. Results: Twenty-four infants
(mean age, 6 weeks) participated in the trial (13 Intralipid and 11 SMOFlipid). At the time of trial enrollment, patients in both groups were
receiving 90% of their calories by PN. Mean duration on trial was 8 weeks and did not differ according to treatment (P = .99). At trial
conclusion, patients who received SMOFlipid had a lower conjugated bilirubin than those who received Intralipid (mean difference, 59
mol/L; P = .03). Patients receiving SMOFlipid were also more likely to have a decrease in serum conjugated bilirubin to 0 mol/L than
those in the Intralipid group over the entire observation period (hazard ratio, 10.6; 95%; P = .03). The time to achievement of full enteral
tolerance did not differ statistically (hazard ratio, 1.3; P = .59) between the groups. There was no significant difference in safety outcomes
between the groups. Conclusions: Compared with Intralipid, SMOFlipid reduces the risk of progressive IFALD in children with intestinal
failure. This trial was registered at as NCT00793195. (JPEN J Parenter Enteral Nutr. 2017;41:866-877)

intestinal failure; intestinal failure associated liver disease; -3 fatty acids; intravenous lipid emulsions; clinical trials

Clinical Relevancy Statement complication.14 While the development of IFALD is related to

both components of the PN, as well as host factors such as sepsis
-3 Lipid emulsions are thought to improve intestinal failure and prematurity, much recent focus has been on the role that
associated liver disease (IFALD), but the true benefit is biased intravenous lipid emulsions (ILEs) play.58
by literature that has been dominated by level 3 evidence and Conventional ILEs that are composed of plant-based lip-
confounded by enteral feeding practices, lipid dose, and sepsis. ids, primarily soy, are thought to be contributory to the
This multicenter randomized controlled trial (RCT) is the first development of IFALD through a number of mechanisms,
to assess liver function as the primary outcome using SMOFlipid including accumulation of phospholipids,9 excess phytoster-
in early onset IFALD. The study was designed to control for ols,1013 predominance of -6 fatty acids (-6FAs) leading to
lipid dose, enteral feeding, and sepsis that have confounded pre- a proinflammatory state,1416 and also antioxidant imbalance
vious literature. Given the current paucity of level 1 evidence, related to inadequate provision of -tocopherol.1719 Given
this pilot RCT supporting the use of a composite lipid emulsion, the role that ILEs are believed to play in the pathophysiology
containing -3 lipids, provides information and data that have of IFALD, there is much interest in alternate approaches to
utility for designing further trials and improving current access lipid provision, namely, either restriction of the dose of con-
to third-generation lipid emulsions across North America. ventional lipid or provision of alternate lipid emulsions
based on fish oil.8 While clinical data support either
approach, limited randomized data exist on which to base
Introduction treatment recommendations.7,8,20
Intestinal failureassociated liver disease (IFALD) is a choles- SMOFlipid (Fresenius Kabi, Bad Homburg, Germany) is a
tatic disorder that develops in up to 66% of children receiving composite emulsion lipid that contains soybean oil (30%),
prolonged parenteral nutrition (PN), and historically, 10%50% medium-chain triglycerides (MCTs) (30%), olive oil (25%), and
of children receiving prolonged PN have died of this fish oil (15%). By reducing exposure to conventional soy-based
Diamond et al 867

lipids as well as by adding -3 lipids, SMOFlipid may be benefi- patients (ie, those who did not end early due to achievement
cial for the prevention of IFALD.21 We tested this hypothesis in of full enteral tolerance) had been enrolled in the trial.
a pilot RCT that compared SMOFlipid with a conventional ILE In June 2010, with 15 patients enrolled, an amendment was
Intralipid (Fresenius Kabi). Although the study was designed as made to continue the trial beyond this period. Although the trial
a pilot to examine the feasibility of conducting a definitive trial, had not met its objective of 16 enrollments within 18 months, it
this article presents the efficacy data from the trial. was felt that such a target could still be achieved within a rea-
sonable time period. The amendment included a provision to
analyze limited data, using blinded group assignments on
Method patients who were enrolled and had completed their participa-
Trial Design tion in the study by June 30, 2010. This blinded analysis was
presented to the Data Safety Monitoring Board (DSMB) in
We performed a multicenter parallel-group randomized con- April 2011, which agreed to allow the trial to continue until the
trolled trial with 1:1 assignment comparing SMOFlipid with entire target of 24 patients was met. The trial closed to enroll-
Intralipid for the prevention of progression of IFALD in ment in September 2011, having met the target of 24 patients.
infants. Given the limited experience with clinical trials in
neonates with IFALD, we designed the trial as a pilot study,
with the goal of examining the feasibility of performing a
clinical trial in this population and to obtain an estimate of Recruitment and enrollment. Potential patients were identified
effect size for a definitive study. Planned enrollment was for by examining the list of patients receiving PN weekly at each
24 patients. Since the trial was designed as a pilot study, the institution for infants with hepatic dysfunction who obtained
sample size was not based on an estimate of treatment effect >40% of their calories parenterally. Potential patients were
but rather to ensure good feasibility data. At the proposed screened weekly until they became eligible or were clearly
sample size, the trial had 80% power to detect a 60-point dif- ineligible. The clinical team caring for the patient made initial
ference in serum conjugated bilirubin (CB). This difference contact, at which point a member of the research team met with
was felt to be a reasonable treatment effect for SMOFlipid the family. Once consent was obtained, enrollment was com-
given our experience with Omegaven (Fresenius Kabi, Bad pleted by a member of the research team (physician or regis-
Homburg, Germany) in those with advanced IFALD.8,22 tered nurse), who notified the pharmacy of the enrollment by
The trial opened to enrollment at the Hospital for Sick submitting a PN order to the pharmacy for trial lipid.
Children in January 2009. Enrollment was opened at the
Hamilton subsite in July 2009, Calgary subsite in September Eligibility criteria. Detailed eligibility criteria are included in
2009, Edmonton subsite in November 2009, and Montreal Figure 1. The primary inclusion criteria were an infant (aged
subsite in October 2010. Enrollment was to close 24 months <24 months) with short bowel syndrome or intestinal failure
after the first site was opened or after 18 months if <16 valid who received substantial PN support (>40% total calories) and

From 1Hospital for Sick Children, Toronto, Ontario, Canada; 2McMaster Childrens Hospital, Hamilton, Ontario, Canada; 3Alberta Childrens
Hospital, Calgary, Alberta, Canada; 4Stollery Childrens Hospital, Edmonton, Alberta, Canada; and 5CHU Sainte-Justine, Montreal, Quebec,

Financial disclosure: The trial was funded by Fresenius Kabi, which also provided the SMOFlipid that was used in the study. The company
played no other role in the study and was not involved in the design of the study, data collection or analysis, or the preparation of the manuscript.
I. R. Diamond was supported by a fellowship award from the Canadian Institutes of Health Research, with additional support from the Surgeon
Scientist Training Program, Department of Surgery, University of Toronto. I. R. Diamond is also the recipient of a Graduate Studentship Award
from the Canadian Liver Foundation and the Chisholm Memorial Fellowship, Post Graduate Medical Education Office, the University of Toronto.
B. M. Feldman is supported by the Ho Family Chair in Autoimmune Disease. The remaining authors have no financial relationships relevant to
this article to disclose.

Conflicts of interest: P. W. Wales and J. Turner have received research funding support from Fresenius Kabi. The other authors have no conflicts of
interest to disclose.

Received for publication July 10, 2015; accepted for publication November 24, 2015.

This article originally appeared online on February 2, 2016.

Corresponding Author:
Paul W. Wales, MD, Division of General and Thoracic Surgery, The Hospital for Sick Children, 555 University Ave, Toronto, Ontario M5G 1X8, Canada.
868 Journal of Parenteral and Enteral Nutrition 41(5)

Inclusion Criteria

1. 24 months of age at enrolment

2. Evidence of early hepatic dysfunction
Serum conjugated bilirubin 17 umol/L on 2 consecutive readings 7 days apart
i. No evidence of sepsis
ii. No prior administration of Omegaven**
3. 40% of total calories administered by PN
4. Meet one of the following diagnostic categories
a. Short Bowel Syndrome
i. Abdominal surgical procedure including gastroschisis/omphalocele closure by any means and percutaneous drainage
procedures within the past 6 months and has been receiving PN since surgery
b. Intestinal Failure
i. One of the following diagnoses for which the child is dependent on PN
1. Gastrointestinal Motility Disorder
2. Mucosal Enteropathy
5. Expectation of the treating physician that the patient will require PN for at least 3 weeks following enrollment**
6. Parents willing to participate including randomization

**Criteria 2-ii and 5 were added to the inclusion criteria 12 months after the trial opened.

Exclusion Criteria

1. Sepsis or Hemodynamic Instability of any cause.

2. Coagulopathy (Platelets 150 000, or INR 1.4)
3. Hypersensitivity to fish-, egg- or soy protein or to any of the active substances or excipients
4. Current enrolment in another clinical trial involving a surgical or pharmacologic intervention
5. Serum conjugated bilirubin > 50 umol/L
6. Hyperlipidemia
7. Treatment with intravenous N-Acetylcysteine or oral Ursodeoxycholic acid
8. Renal insufficiency
9. Disorders of Fluid Balance
10. Unstable medical conditions

Figure 1. Eligibility criteria. INR, international normalized ratio; PN, parenteral nutrition.

was demonstrating early hepatic dysfunction (CB: 1750 support pharmacist at the Hospital for Sick Children and the dis-
mol/L [13 mg/dL]) in the absence of sepsis. pensing pharmacist at the patients institution were aware of the
group assignment. Unblinding of participants and investigators
Intervention occurred only at the conclusion of the entire trial.

Randomization and allocation concealment.Randomization Duration of treatment. Participants received trial lipid for up
was done centrally by the research support pharmacy at the to 12 weeks. Patients also ended the trial if they achieved full
Hospital for Sick Children. The randomization sequence was enteral tolerance (autonomy from PN) prior to this time point
developed by the research support pharmacy using a random- or if they developed progressive liver disease defined by a
number table prior to enrollment of the first patient. The serum CB exceeding 100 mol/L for >14 days. Since duration
sequence was developed in blocks of variable size without of PN is challenging to predict, we included a provision for
investigator input or knowledge. replacement of patients who discontinued PN prior to the sec-
Allocation concealment was achieved by the randomization ond week on the study due to achievement of full enteral toler-
sequence only being known to the research support pharmacy ance but not for reasons of safety or efficacy. By design, these
at the Hospital for Sick Children. The group assignment was patients were not to be included in the analysis of trial out-
relayed to the dispensing pharmacist at the patients institution comes or count toward recruitment targets.
only after enrollment had occurred. This was done immedi-
ately prior to dispensing the first dose of study medication. Lipid dosing. The PN solution was formulated according to a
nomogram (Figure 2) that took into account the proportion of
Blinding. The study lipid was administered in a blinded manner, caloric intake received parenterally. The feasibility of formu-
by dispensing the trial lipids in identical containers (syringe or lating PN according to a nomogram was one of the objectives
infusion bag). All participants, treating clinicians, and investiga- of the pilot study. Based on initial experience, the nomogram
tors were blinded to the treatment assignment. Only the research was modified 12 months into the study as indicated in Figure 2.
Diamond et al 869

PN Dosing Nomogram Initial Protocol

% of calories Lipid (g/kg/day) Protein (g/kg/day) Carbohydrate (g/ GIR* (mg/kg/min) PN Calorie
administered kg/day) Range** (kcal/kg/
by PN day)
80 100 % 2.5 3 34 15 20 10.4 13.9 88 114
60 80 % 2.5 3 2.5 3.5 9 15 6.3 10.4 66 95
50 60 % 23 2 2.5 7 11 4.9 7.6 52 77
40 50 % 1.5 2.5 1.5 2 69 4.2 6.3 41 69
30 40 % 1.0 2.5 1 1.5 57 3.5 4.9 31 55
20 30 % 0.7 2 0.7 1 34 2.1 2.8 20 38
10 20 % 0.3 1.5 0.5 1 12 0.7 1.4 8 28
< 10 % 01 0 1 01 0 0.7 0 17

PN Dosing Nomogram Revised January 2010

% of calories Lipid (g/kg/day) Protein (g/kg/day) Carbohydrate (g/ GIR* (mg/kg/min) PN Calorie
administered kg/day) Range** (kcal/kg/
by PN day)
80 100 % 23 34 10 20 6.9 13.9 64 111
60 80 % 23 2.5 3.5 8 16 5.6 11.1 55 95
50 60 % 23 2 2.5 6 12 4.2 8.3 46 78
40 50 % 1.5 2.5 1.5 2 5 10 3.5 6.9 37 65
30 40 % 1.0 2.5 1 1.5 48 2.8 5.6 27 56
20 30 % 0.7 2 0.7 1 26 1.4 4.2 16 42
10 20 % 0.3 1.5 0.5 1 14 0.7 2.8 8 31

< 10 % 01 0 1 02 0 1.4 0 20

Figure 2. Nomogram for dosing of parenteral nutrition (PN). *This value reflects the GIR for a 24-hour period based on the total
daily glucose intake. It does not reflect a maximum GIR in the setting of a PN cycle of < 24 hours. **Provided for reference only
with rounding to whole numbers. Lower limit calculated using the minimum macronutrient values and upper limit calculated using
the maximum values for each macronutrient. Caloric densities for the macronutrients used in the calculations were as follows: lipid =
9 kcal/g, protein = 4 kcal/g, carbohydrate = 3.4 kcal/g. GIR, glucose infusion rate.

Management of enteral and parenteral feeds.We antici- Outcome Assessment

pated that, over the course of the trial, adjustments would be
made to the proportion of feeds received enterally and par- Efficacy outcomes.Most outcomes were assessed weekly
enterally by each participant according to his or her clinical while the patient was on the trial. The primary end point for
needs. Choice of feeds, timing of the adjustment of the analysis was CB the week that the child received his or her
enteral feeds, and ratio of enteral to parenteral feeds were to last dose of trial lipid (at 12 weeks, at full enteral tolerance,
be at the discretion of the treating physician. Initially, we or on the development of progressive liver disease). In addi-
had provided a specific list of enteral feeds to choose from. tion to the primary outcome, other data collected at baseline,
However, due to poor buy-in from the treating physicians, weekly while on the trial, and 4 weeks following trial com-
this provision was removed from the protocol 12 months pletion (referred to as posttrial) included serum CB, serum
into the study. The only enteral formula that was still not albumin, liver enzymes, parenteral intake, and enteral intake.
allowed was enteral fish oil solution. The following measurements were all obtained at the time
870 Journal of Parenteral and Enteral Nutrition 41(5)

the primary outcome was met and in addition as reported. Results

Weight, length, and head circumference were assessed at
baseline, week 6, and posttrial. A complete blood count was Recruitment and Participant Flow
done at weeks 0, 4, and 8 and posttrial. An international nor- Figure 3 lists the outcomes of all patients who were screened.
malized ratio, C-reactive protein, immunologic markers Of the 394 potential patients screened, 350 were not eligible.
(interleukins 1, 6, 8, 10, and 12; tumor necrosis factor ), Most of these (284; 81%) were excluded since their bilirubin
nephelometry, serum cholesterol, and serum triglycerides was not 1750 mol/L for at least 7 days while they were
were assessed at baseline, week 6, and posttrial. Red blood receiving >40% of their calories parenterally. Twenty-two
cell phospholipid composition was assessed at baseline, potential patients were excluded because they were receiving
week 6, and posttrial. Omegaven, N-acetylcysteine, or ursodeoxycholic acid. Forty-
four potential patients were excluded because of a medical
Safety outcomes. Adverse events were handled according to
contraindication; most of these potential patients were prema-
the principles of Good Clinical Practice (GCP) as adopted
ture neonates who were excluded due to thrombocytopenia.
by Health Canada.23 Adverse events were recorded irrespec-
Of the 26 patients who were randomized, 17 were from the
tive of whether the adverse event was believed to be related
Hospital for Sick Children, 5 from the Calgary subsite, 3 from
to the patients participation in the trial. All unexpected
the Hamilton subsite, and 1 from the Edmonton subsite. No
adverse events were reported to the local Research Ethics
patients were enrolled in Montreal. A total of 26 patients were
Board as well as to the DSMB.
randomized to meet the final sample size of 24 as 2 patients
The DSMB included a pediatric surgeon, a pediatrician
achieved full tolerance of enteral feeds prior to the second
with expertise in nutrition and gastroenterology, and a bio-
week on trial. As per protocol, these patients were not consid-
statistician. All members of the DSMB were independent
ered valid participants, and their data were not included in the
from the trial to the extent that they were not investigators,
nor will they be included in any publications arising from
A total of 24 patients completed the trial: 11 in the
the trial. By protocol, the DSMB met to review the records
SMOFlipid group and 13 in the Intralipid group. All of these
of the first 4 patients within 6 weeks of the 4th patients
patients were analyzed in the group to which they were
posttrial assessment and again within 6 weeks of the 14th
assigned. The mean duration on trial did not differ statistically
patients posttrial assessment. No further scheduled meet-
between those assigned to SMOFlipid (8 weeks; 95% confi-
ings were held, although a meeting could be called by any
dence interval [CI], 5.510.5 weeks) vs Intralipid (8 weeks;
DSMB member or one of the investigators. The DSMB
95% CI, 5.710.7 weeks), P = .99.
members were provided regularly with a listing of unex-
pected adverse events and serious adverse events.
Statistical Analysis Table 1 provides the demographic characteristics of the partici-
The primary end point for analysis was CB the week that the pants at the time of recruitment. The mean age was approxi-
child received his or her last dose of trial lipid (at 12 weeks, mately 6 weeks in both groups, and most participants had been
at full enteral tolerance, or on the development of progressive receiving PN for their entire life. At the time of trial enroll-
liver disease). The values for the SMOFlipid patients were ment, patients in both groups were receiving 90% of their calo-
compared with the Intralipid patients at the time the primary ries by PN. All but 1 of the patients had a surgical diagnosis
end point was met and at the completion of the posttrial leading to their intestinal failure.
period (4 weeks following completion of trial lipid). Data are
presented as means and standard deviation (SD) to meet our Outcomes
goal of obtaining an estimate of effect size. For a similar rea-
son, the a priori analysis plan included parametric analysis; t Hepatic and nutrition outcomes. Hepatic and nutrition outcomes
tests were used for continuous data outcomes between groups at baseline and at the time the primary outcome was achieved (the
and Fisher exact tests were used for categorical data. Given primary endpoint) are shown in Table 2. The primary end point
that the primary end point could occur in an individual patient was CB the week that the child received his or her last dose of
at a different time, a time-to-event analysis until 4 weeks trial lipid (at 12 weeks, at full enteral tolerance, or on the develop-
posttrial is also included, using the log-rank test, and Cox ment of progressive liver disease). In the SMOFlipid group, 4
proportional hazards regressions generated hazard ratios patients were studied to 12 weeks, 6 achieved full enteral toler-
(HRs). A P value of .05 was deemed statistically significant, ance, and 1 developed progressive liver disease. In the Intralipid
and a value of .1 was considered a trend. Analyses were per- group, 6 patients were studied to 12 weeks, 6 achieved full enteral
formed using R (The R Project for Statistical Computing, tolerance, and 0 developed progressive liver disease. The 1
Vienna, Austria). patient in the SMOFlipid group who met the definition for
Diamond et al 871

Figure 3. Patients screened and participant flow.

Table 1. Characteristics of Patients at Trial Entry.

Characteristic SMOFlipid (n = 11) Intralipid (n = 13)

Age, mean (range), wk 6.5 (4.38.7) 5.3 (3.57.2)
Gestational age, mean (range), wk 34.5 (32.436.7) 35.2 (33.237.1)
Birth weight, mean (range), kg 2.39 (1.942.84) 2.55 (2.132.98)
Weight at enrollment, mean (range), kg 3.24 (2.853.62) 3.21 (2.823.59)
Male sex, No. (%) 6 (55) 7 (54)
Etiology of intestinal failure, No. (%)
Small bowel atresia 5 (46) 3 (23)
Necrotizing enterocolitis 2 (18) 1 (8)
Gastroschisis 2 (18) 4 (31)
Motility disorder 1 (9) 2 (15)
Diaphragmatic hernia 1 (9) 0 (0)
Volvulus 0 (0) 1 (8)
Meconium ileus 0 (0) 1 (8)
Mucosal enteropathy 0 (0) 1 (8)
Weeks of PN prior to trial, mean (range) 5.5 (3.77.2) 4.4 (2.95.8)
No. (%) with ileocecal valve in situ 8 (73) 11 (85)
No. (%) with stoma 5 (46) 4 (31)
No. (%) with STEP procedure 1 (9) 1 (8)
Comorbidity, No. (%)
Neurological 4 (36) 4 (31)
Cardiac 2 (18) 3 (23)
Respiratory 2 (18) 1 (8)
Genitourinary 3 (27) 1 (8)
Hematologic 5 (46) 5 (39)
Musculoskeletal 0 (0) 1 (8)

PN, parenteral nutrition; STEP, serial transverse enteroplasty.

progressive liver disease experienced a septic episode at 6 weeks a t test was performed. In this analysis, there was no statistical
with associated hepatic dysfunction, and the serum CB rose difference between the groups (P = .42). However, since it was
acutely to 288 mol/L. This posed a challenge statistically, as this recognized that there was a statistical outlier influencing the
value was a statistical outlier (Figure 4). Therefore, the analysis results, a nonparametric analysis was also performed using the
was performed a number of ways. First, as per the analysis plan, Wilcoxon rank sum test that demonstrated that the serum CB at
872 Journal of Parenteral and Enteral Nutrition 41(5)

Table 2. Hepatic and Nutrition Parameters at Trial Start and End.

SMOFlipid (n = 11) Intralipid (n = 13) Difference Between
SMOFlipid and P
Parameter Start End Start End Intralipid Value

Conjugated bilirubin,a mol/L 35 (2050) 22 (440) 36 (3240) 69 (4890) 47 (73 to 21) .001
Conjugated bilirubin,b mol/L 34 (2148) 46 (10 to 102) 36 (3240) 69 (4889) 23 (81 to 36) .42c
Unconjugated bilirubin, mol/L 23 (1134) 47.5 (32 to 127) 23 (1532) 38 (2155) 9 (71 to 89) .80
AST, U/L 58 (3582) 138 (92184) 66 (29102) 271 (98444) 133 (309 to 44) .13
ALT, U/L 51 (2378) 156 (71241) 51 (1192) 192 (88295) 36 (162 to 90) .56
ALP, U/L 304 (229380) 394 (290499) 339 (263414) 513 (426601) 119 (248 to 10) .07
GGT, U/L 196 (162231) 295 (190399) 151 (101200) 180 (131230) 114 (3226) .04
Serum albumin, g/L 27 (2429) 30 (2734) 28 (2530) 34 (3136) 3 (7 to 1) .12
INR, s 1.1 (11.2) 1.2 (11.3) 1.1 (11.2) 1.1 (11.2) 0.1 (0.1 to 0.2) .38
Proportion of calories parenterally 0.92 (0.811.03) 0.45 (0.140.77) 0.89 (0.790.98) 0.40 (0.20.59) 0.05 (0.3 to 0.41) .74
Weight, kg 3.2 (2.93.6) 4.5 (4.05.0) 3.2 (2.83.6) 4.3 (3.74.9) 0.2 (0.5 to 0.9) .51
Height, cm 50 (4952) 56 (5359) 50 (4952) 55 (5358) 0.5 (3 to 4) .78
Head circumference, cm 35 (3336) 38 (3639) 35 (3436) 38 (3740) 0.4 (2 to 2) .65

ALP, alkaline phosphatase; ALT, alanine transaminase; AST, aspartate aminotransferase; GGT, -glutamyl transferase; INR, international normalized ratio.
Does not include statistical outlier in the SMOFlipid group.
Includes statistical outlier in the SMOFlipid group.
The P value listed is for the t test; the corresponding nonparametric P value is .002.

trial completion remained significantly lower in the SMOFlipid

group (P = .02). Finally, to obtain an estimate of effect size, the
analysis was performed with the septic patient excluded. In this
analysis, the CB in the SMOFlipid group was 47 mol/L lower
than in the Intralipid group (P = .001). At 4 weeks, the patients
CB had declined to 90 mol/L. While this value was the highest
in the SMOFlipid group, it was no longer an outlier. At this time
point, the patients in the SMOFlipid group continued to exhibit a
lower CB than those in the Intralipid group, including all patients
with no exclusions (59 mol/L; P = .03). Patients who received
SMOFlipid were more likely to have a decrease in serum CB to 0
mol/L than those in the Intralipid group over the entire observa-
tion period (HR, 10.6; 95% CI, 1.386.9; P = .006).
At the time the primary end point was achieved, 3 (27%)
patients in the SMOFlipid group had a serum CB exceeding 50
mol/L, while 9 (69%) patients in the Intralipid group had a
bilirubin above this level (P = .04). Three (27%) patients in the
SMOFlipid group had a CB = 0, and 1 (9%) had a CB >100
mol/L (the patient with sepsis described above). There were 0
patients in the Intralipid group with a CB = 0 mol/L and 2
Figure 4. Box plots of the conjugated bilirubin in the final trial
(15%) with a CB >100 mol/L; these proportions were not sig-
week, demonstrating that 1 patient in the SMOFlipid group was a
nificantly different between groups. Posttrial, 1 (9%) patient in statistical outlier.
the SMOFlipid group and 6 (46%) patients in the Intralipid
group had a CB >50 mol/L (P = .08). At this time point, there
were 3 (23%) patients in the Intralipid group and no patients in statistically significant (P = .07). However, -glutamyl trans-
the SMOFlipid with a CB >100 mol/L (P = .22). ferase in the SMOFlipid group was significantly higher at
There was no significant difference in terms of unconju- trial completion (P = .04). There was no difference in these
gated bilirubin and transaminases between the groups at the cholestatic enzymes between the groups posttrial, but the
time the primary end point was achieved. Although alkaline patients in the SMOFlipid group had lower aspartate amino-
phosphatase was lower in the SMOFlipid group, it was not transferase levels (P = .02).
Diamond et al 873



Figure 5. (A) Conjugated bilirubin, (B) total caloric intake, (C) lipid intake, and (D) parenteral caloric intake throughout the trial.

At no time was there a statistical difference in terms of of -6:-3 fatty acids in the serum was lower in the SMOFli-
nutrition parameters. Six patients in both the SMOFlipid and pid group (8:1) compared with the Intralipid group (14:1) at
Intralipid groups achieved full enteral tolerance at the primary trial completion (P = .05). However, this difference was no
end point. Posttrial, there were 7 patients in the SMOFlipid longer evident posttrial. The patients who received SMOFlipid
group and 6 in the Intralipid group with full enteral tolerance. also had statistically lower levels of linoleic acid and linoleic
The time to achievement of full enteral tolerance did not differ acid and statistically higher levels of eicosapentaenoic acid,
statistically between SMOFlipid and Intralipid in a time-to- docosahexaenoic acid, and oleic acid in their red blood cell
event analysis (HR, 1.3; 95% CI, 0.54.0; P = .59). membranes while on the trial lipid, although none of these dif-
Figure 5 examines the relationship between serum CB, total ferences persisted posttrial. There was no impact of the inter-
caloric intake, proportion of caloric intake parenterally, and vention on triglycerides, cholesterol, or lipid clearance. There
total lipid intake throughout the trial. This figure demonstrates was no statistical difference in terms of immunologic parame-
that total caloric intake, proportion of caloric intake parenter- ters between the groups during or posttrial.
ally, and total lipid intake were similar throughout the trial for
those assigned to SMOFlipid and Intralipid, suggesting that
differences in these parameters do not account for the observa-
Adverse Events
tion that patients who received SMOFlipid had a decline in Adverse events are provided in Supplementary Table S2. All
serum CB over the course of the study, whereas those receiving patients experienced at least 1 adverse event. The mean num-
Intralipid had an increase in serum CB. ber of adverse events per patient was 12.4 in the SMOFlipid
group and 9.2 in Intralipid group (P = .30). There were no seri-
Immunologic and lipid outcomes. Immunologic and lipid out- ous unexpected adverse events or any unexpected adverse drug
comes at baseline and trial completion are included as supple- reactions as assessed by the DSMB. There were no deaths due
mentary material (see Supplementary Table S1). The proportion to liver disease during the trial, but 1 child who had received
874 Journal of Parenteral and Enteral Nutrition 41(5)

Intralipid died of progressive liver disease shortly after the trial of conventional -6containing lipids can both prevent and
was completed. This death was reviewed by the DSMB as well treat IFALD,28,35 but when used for prevention, there may be
as the Research Ethics Boards at the Hospital for Sick Children potential for negative impact on normal growth and develop-
as well as the patients institution. ment.28 Therefore, we have argued previously that there is
insufficient evidence to recommend these novel therapies in
children with early or no IFALD to prevent the progression to
Discussion advanced IFALD due to unproven efficacy in clinical trials, the
IFALD is the greatest contributor to the morbidity experienced unknown long-term safety profile (coagulopathy, essential fatty
by children with intestinal failure (IF).1,2,24 Historically, IFALD acid deficiency, neurocognitive outcome), and added cost.25,26
was the primary cause of death of children with IF. While the Intralipid, a soy-based lipid emulsion, has been the standard
pathophysiology of IFALD is multifactorial, recently much lipid emulsion for parenterally supported patients for over 40
attention has been devoted to the critical role that -6FAs con- years. It provides energy and is a source of essential fatty acids.
taining ILEs play in the development of IFALD.3 Understanding Intralipid comprises predominantly -6 long-chain PUFAs and
the role that ILEs play in the development of IFALD may allow is believed to contribute to IFALD due to its more inflamma-
the use of targeted therapies.25,26 The 2 approaches that have tory cytokine profile, significant phytosterol content (that
been advanced are that of lipid minimization and alternative inhibits bile acid excretion), and relatively low antioxidant
ILEs, including those containing -3 fatty acids (-3FAs). content (-tocopherol). It has an -6:-3 ratio of 8:1. It does
Retrospective human studies and animal research have revealed not provide any significant docosahexaenoic acid (DHA),
that decreasing or discontinuing parenteral soybean-based lipid which is important especially in preterm infants for neurocog-
emulsions will reverse cholestasis and improve bile flow.10,2733 nitive development. The essential -3 PUFA in Intralipid,
-3 Lipid emulsions are believed to improve hepatic function linolenic acid, cannot be converted into DHA that is required
via several mechanisms, including improved bile flow. Clearance by neonates; therefore, the conventional lipid emulsion pro-
of lipid is improved and uptake by extrahepatic tissues is vided to neonates is DHA deficient.
increased through enhanced -oxidation. Added -3 polyunsat- Omegaven has been used in Canada since 2006, after Gura
urated fatty acids (PUFAs) are metabolized into eicosanoids etal36 first reported reversal of cholestasis in 2 patients. It is
with a less inflammatory profile and therefore may modulate the available through compassionate release from Health Canada.
immune system, with reduced hepatic inflammation and scar- It is made of fish oil predominantly (-6:-3 = 1:7). It was
ring ultimately reducing the risk of cirrhosis and liver failure.7 never designed as a source of sole parenteral lipid support. It
An -6:-3 ratio <4:1 is believed to reduce inflammation, and gained popularity for rescue therapy in children with advanced
<2.5:1 has been shown to improve bile flow in animal models. IFALD who would otherwise die without liver transplantation.
Current third-generation, -3containing lipid emulsions all With the reported success of Omegaven in several retrospec-
contain more -tocopherol than conventional soy-based lipids. tive cohort studies, its use expanded, and it began to be used in
Preliminary studies in preterm piglets suggest this may be the patients with lower risk disease. The literature unfortunately is
key factor in improving bile metabolism.19 The third-generation of poor quality due to the retrospective nature of study design
lipids also contain less phytosterols than conventional lipids, and the confounding of results by the effect of lipid dose
which can accumulate in the liver and also, based on animal restriction, phytosterol restriction, enteral nutrition advance-
studies, appear to affect bile acid metabolism.13 ment, and enhanced delivery of -tocopherol. SMOFlipid, a
Regardless of our incomplete understanding of the exact third-generation product, is a composite lipid emulsion com-
mechanisms by which -3containing lipids can improve posed of soybean oil, MCTs, olive oil, and fish oil. It has an
IFALD, these therapies have been studied in children predomi- -6:-3 ratio of 2.5. It is designed as a balanced emulsion for
nantly in uncontrolled settings and primarily in those having any patient requiring parenteral lipid support. It provides
advanced IFALD. A small trial has been conducted in IFALD essential fatty acids and also has a desirable immune profile
comparing 2 lipid emulsions, -3 (n = 9) and -6 (n = 7), at the due to the added -3 PUFAs. SMOFlipid has been available in
same dose but was confounded by enteral advancement, and Europe for several years and has been licensed for pediatric use
unfortunately, while there were differences in CB that favored since 2009. It has been licensed in Canada since June 2013.
the -3 treatment, there was a high rate of resolution of IFALD There have been 4 randomized trials, 3 in premature babies
in both groups.34 A systematic review of the available literature that were short term (24 weeks) and evaluated safety as the
concluded that there was evidence that -3containing lipid primary outcome. Most also demonstrated improved liver
emulsions could reverse severe IFALD but that there were lim- function as a secondary outcome.3740
ited data for prevention.20 The review identified only 3 random- While the randomized controlled trial provides the best evi-
ized trials and only 2 of good quality. The findings are also dence for the efficacy and safety of an intervention,41 these tri-
supported by 1 recent RCT, which found equivalent efficacy als may be challenging to perform in children with IF. First, IF
between -3 treatment and -6 dose restriction in preventing is a rare disease, and therefore trials will need to be performed
early onset IFALD.29 In contrast, it appears that dose restriction across multiple centers. Second, children with IF are a
Diamond et al 875

heterogeneous group of patients who typically have multiple have been other randomized studies exploring SMOFlipid in
additional comorbidities, and the standardization of manage- infants,3740 these were all short-term trials with safety as the
ment required in a clinical trial may be difficult to achieve. primary outcome. It also represents the longest intervention
Also, it is unclear if randomized studies would be acceptable to studied with a treatment period of up to 12 weeks. The random-
the caregivers of these complex children. There is also little ized design addresses many of the concerns with respect to the
evidence on which to base estimates of treatment efficacy with literature in an area that has been patient to confounding related
the novel lipid-based approaches for use in sample size calcu- to indication, as well as dosing, of lipid. Second, lipid dosing
lations. Finally, there is limited experience with clinical trials was standardized for the trial. We believe that such standard-
in the IF population.42 ization is necessary to ensure that the treatment groups receive
Given the potential for challenges in performing such ran- equivalent nutrition intake allowing for unbiased assessment
domized studies, we felt it prudent to evaluate the feasibility of of the treatment response. Finally, the range of lipid intake
the randomized controlled design for the evaluation of a novel used in the trial was within the range traditionally considered
lipid-based strategy for the prevention of advanced IFALD. appropriate for developing infants, and therefore, the strategy
While recruitment was slower than anticipated, primarily due employed allays our concerns with respect to impact of treat-
to the delay in launching external sites and a number of refus- ment on growth and development. Finally, it is important to
als early in the trial, the overall design was assessed to be fea- recognize that this trial was designed as a pilot study, and while
sible. One of the specific issues that we felt needed to be this article primarily focused on the efficacy outcomes from
addressed was the feasibility of standardizing the administra- the pilot trial, it is important to note that our feasibility objec-
tion of PN for patients enrolled in the trial. We believed this tive was met, which should facilitate future trials.
aspect of the design to be important to allow the groups to be Despite the strengths of the trial, there are a number of limi-
balanced from a nutrition perspective and facilitate optimal tations. First is the fact that this was a pilot study, and while the
assessment of efficacy of the therapy. We do acknowledge this effect size demonstrated by the study is consistent with what
remains a limitation of conducting these trials and that our ini- we hypothesized, the results need to be tempered by this fact.
tial nomogram was adjusted during this trial with feedback The trial was never meant to be the definitive randomized trial,
from investigational centers. The PN dosing nomogram was rather to establish the feasibility of randomized trials in this
the most common reason for deviations in the pilot study. The patient population and to provide data to facilitate the design of
outcome of the adjustments made was to substantially reduce these trials such as estimates of effect size. The impact of sam-
the incidence of such deviations over the course of the trial. ple size was highlighted by the 1 patient who was a statistical
This will be critical in the design of future trials, and we con- outlier, due to an episode of severe sepsis. In a larger trial, a
clude that it is feasible to standardize dosing of PN in the con- single patient would not likely have such an influential impact
text of a randomized controlled trial. on the results, and while we believe that we addressed the
The objective of this study was to explore whether impact of this patient statistically, this patient does weaken our
SMOFlipid, a composite lipid emulsion, would reduce the risk conclusion. Second, we adopted a very comprehensive
of IFALD progression in children receiving PN who were approach to screening patients for the trial and hence the large
exhibiting early hepatic dysfunction. Our results confirmed our number of patients whose bilirubin was not in range for the
hypothesis that SMOFlipid, provided at conventional dosing, study (1750 mol/L for at least 7 days). However, there were
lessens the risk of IFALD progression as quantified by serum 44 patients who had another exclusion, typically sepsis or
CB, our primary clinical outcome. Children who received thrombocytopenia, at the time that their CB was within range
SMOFlipid had a higher level of the cholestatic marker (1750 mol/L) for the study. Most of the infants who were
-glutamyltransferase (GGT) but a trend toward a lower alka- excluded for this reason were preterm or low-birth-weight
line phosphatase (ALP). We do not have an explanation for this infants. In designing future studies, it would be ideal to include
dichotomy. We do not have histological samples to assess for these infants, since they are known to be at significant risk of
alterations in intrahepatic bile ducts. Another potential expla- progressive IFALD.2,4345 Thrombocytopenia should be recon-
nation is a direct activation of GGT gene expression, without sidered as an exclusion criterion as coagulopathy is largely a
underlying liver pathology. This issue should be explored in consideration with an ILE that is exclusively composed of
future studies. We did not demonstrate a meaningful difference -3FAs and not particularly germane to a composite emulsion
in terms of immunologic outcomes. This may be related to the such as SMOFlipid.46 Finally, it is important to recognize that
fact that we measured serum cytokines, whereas tissue markers while the results are consistent with the experience with -3
may have been more sensitive, yet unfeasible to examine. lipids in advanced IFALD, SMOFlipid is a composite emul-
There were no differences in terms of nutrition or safety param- sion, which has both reduced -6FAs and additional -3FAs,
eters between the groups. and a positive result does not necessarily mean that the addi-
There are a number of strengths to this study, the first ran- tion of -3FA is the reason for a beneficial effect. Therefore, in
domized blinded controlled trial of SMOFlipid for the preven- concert with clinical trials, animal models are required to fur-
tion of progression of early IFALD in children. While there ther investigate ILEs of varying composition and their role in
876 Journal of Parenteral and Enteral Nutrition 41(5)

the management of IFALD. These studies will also allow References

investigators to further delineate the mechanism of action of 1. Wales PW, De Silva N, Kim J, Lecce L, To T, Moore AM. Population-
these therapies. based estimates of incidence and mortality rates. J Pediatr Surg.
Several important questions remain to be answered about 2004;39:690-695.
the use of lipid in children with PN-dependent intestinal fail- 2. Wales PW, de Silva N, Kim JH, Lecce L, Sandhu A, Moore AM. Neonatal
short bowel syndrome: a cohort study. J Pediatr Surg. 2005;40:755-762.
ure.4749 Will a long-term reduction in soybean oil dose to 1 g/ 3. Andorsky DJ, Lund DP, Lillehei CW, etal. Nutritional and other post-
kg/d result in adequate growth and neurological development, operative management of neonates with short bowel syndrome correlates
and will essential fatty acid deficiency (EFAD) be prevented? with clinical outcomes. J Pediatr. 2001;139(1):27-33.
Is a fish oil emulsion more effective than an equivalently dosed 4. Warner BW, Ziegler MM. Management of the short bowel syndrome in
soybean oil emulsion at preventing IFALD, promoting neuro- the pediatric population. Pediatr Clin North Am. 1993;40(6):1335-1350.
5. Kaufman SS. Prevention of parenteral nutritionassociated liver disease in
logical development, and what is the incidence of EFAD if the children. Pediatr Transplant. 2002;6(1):37-42.
low dose is given over a long duration? Is SMOFlipid, deliv- 6. Goulet O, Joly F, Corriol O, Colomb-Jung V. Some new insights in intes-
ered at conventional lipid doses, effective at preventing the tinal failureassociated liver disease. Current Opin Organ Transplant.
development of IFALD while optimizing growth and develop- 2009;14(3):256-261.
ment over the long term? At this stage, it is unethical to design 7. Diamond IR, Sterescu A, Pencharz PB, Wales PW. The rationale for the
use of parenteral omega-3 lipids in children with short bowel syndrome
a trial evaluating novel lipid strategies (dose restriction or fish and liver disease. Pediatr Surg Int. 2008;24:773-778.
oil emulsions) in the setting of rescue therapy for children with 8. Diamond IR, Pencharz PB, Feldman BM, Ling SC, Moore AM, Wales
advanced IFALD. These children traditionally have a high PW. Novel lipid-based approaches to pediatric intestinal failureassoci-
mortality and will die without transplantation. The focus of ated liver disease. Arch Pediatr Adolesc Med. 2012;166(5):473-478.
future trials, therefore, should be on IFALD prevention with 9. Cavicchi M, Beau P, Crenn P, Degott C, Messing B. Prevalence of liver
disease and contributing factors in patients receiving home parenteral nutri-
short-term hepatic and longer term growth and developmental
tion for permanent intestinal failure. Ann Intern Med. 2000;132(7):525-
outcomes. As evidence from clinical trials accumulates on the 532.
efficacy of either novel lipid-based approach as a preventative 10. Van Aerde JE, Duerksen DR, Gramlich L, etal. Intravenous fish oil emul-
strategy, trials directly comparing these strategies should be sion attenuates total parenteral nutritioninduced cholestasis in newborn
performed. In our recently published expert belief elicitation piglets. Pediatr Res. 1999;45(2):202-208.
11. Llop JM, Virgili N, Moreno-Villares JM, etal. Phytosterolemia in par-
study,50 there already appears to be evidence of equipoise in
enteral nutrition patients: implications for liver disease development.
terms of expert opinion on the relative efficacy of the alternate Nutrition. 2008;24(11-12):1145-1152.
approaches. This may provide justification for a trial that 12. Clayton PT, Bowron A, Mills KA, Massoud A, Casteels M, Milla PJ.
would directly compare the -3FA approach to lipid minimiza- Phytosterolemia in children with parenteral nutritionassociated choles-
tion. Obstacles to progress include lack of consensus on the tatic liver disease. Gastroenterology. 1993;105(6):1806-1813.
13. El Kasmi KC, Anderson AL, Devereaux MW, etal. Phytosterols promote
definition of cholestasis, determination of the most appropriate
liver injury and Kupffer cell activation in parenteral nutritionassociated
study clinical end point,51 individual clinician bias and percep- liver disease. Sci Transl Med. 2013;5(206):1-15.
tion of advanced IFALD, and access to novel lipid products 14. Heller A, Koch T, Schmeck J, van Ackern K. Lipid mediators in inflam-
and strategies outside a clinical trial (the backdoor). matory disorders. Drugs. 1998;55(4):487-496.
15. Broughton KS, Wade JW. Total fat and (n-3):(n-6) fat ratios influence
eicosanoid production in mice. J Nutr. 2002;132(1):88-94.
Acknowledgments 16. Grimm H, Tibell A, Norrlind B, etal. Immunoregulation by parenteral
We thank Mark Bedford (research pharmacy), Janice Bowers lipids: impact of the n-3 to n-6 fatty acid ratio. JPEN J Parenter Enteral
(administrative support), and Eveline Lapidus-Kroll, Julia Nutr. 1994;18(5):417-421.
Pemberton, Viona Lam, and Sandra Young (research coordinators). 17. Antebi H, Mansoor O, Ferrier C, etal. Liver function and plasma anti-
oxidant status in intensive care unit patients requiring total parenteral
nutrition: comparison of 2 fat emulsions. JPEN J Parenter Enteral Nutr.
Statement of Authorship 2004;28(3):142-148.
I. R. Diamond, P. B. Pencharz, B. M. Feldman, S. C. Ling, A. M. 18. Grimm H, Mertes N, Goeters C, etal. Improved fatty acid and leukotri-
Moore, Y. Avitzur, and P. W. Wales contributed to conception or ene pattern with a novel lipid emulsion in surgical patients. Eur J Nutr.
design of the research; I. R. Diamond, R. C. Grant, N. de Silva, P.
19. Ng K, Stoll B, Chacko S, etal. Vitamin E in new-generated lipid emul-
Fitzgerald, D. Sigalet, B. Dicken, J. Turner, V. Marchand, and P.
sions protects against parenteral nutritionassociated liver disease in par-
W. Wales contributed to acquisition, analysis, or interpretation of enteral nutritionfed preterm pigs [published online January 16, 2015].
the data; and I. R. Diamond, R. C. Grant, and P. W. Wales drafted JPEN J Parenter Enteral Nutr.
the manuscript. All authors critically revised the manuscript, gave 20. Park HW, Lee NM, Kim JH, Kim KS, Kim SN. Parenteral fish oil
final approval, and agree to be accountable for all aspects of work containing lipid emulsions may reverse parenteral nutritionassociated
ensuring integrity and accuracy. cholestasis in neonates: a systematic review and meta-analysis. J Nutr.
21. Muhammed R, Bremner R, Protheroe S, Johnson T, Holden C, Murphy
Supplementary Material
MS. Resolution of parenteral nutritionassociated jaundice on changing
Supplementary Tables S1 and S2 are available with the article from a soybean oil emulsion to a complex mixed-lipid emulsion. J Pediatr
online at Gastroenterol Nutr. 2012;54(6):797-802.
Diamond et al 877

22. Diamond IR, Sterescu A, Pencharz PB, Kim JH, Wales PW. Changing the 38. Skouroliakou M, Konstantinou D, Koutri K, etal. A double-blind, ran-
paradigm: Omegaven for the treatment of liver failure in pediatric short domized clinical trial of the effect of omega-3 fatty acids on the oxida-
bowel syndrome. J Pediatr Gastroenterol Nutr. 2009;48(2):209-215. tive stress of preterm neonates fed through parenteral nutrition. Eur J Clin
23. Scientific and Medical Advisory Branch Health Canada. Good Clinical Nutr. 2010;64(9):940-947.
Practice: consolidated guideline. 1997. 39. Tomsits E, Pataki M, Tolgyesi A, Fekete G, Rischak K, Szollar L. Safety
prodpharma/applic-demande/guide-ld/ich/efficac/e6-eng.php. Accessed and efficacy of a lipid emulsion containing a mixture of soybean oil,
March 29, 2011. medium-chain triglycerides, olive oil, and fish oil: a randomised, double-
24. Vanderhoof JA, Langnas AN. Short-bowel syndrome in children and blind clinical trial in premature infants requiring parenteral nutrition.
adults. Gastroenterology. 1997;113:1767-1778. J Pediatr Gastroenterol Nutr. 2010;51(4):514-521.
25. Diamond IR, Pencharz PB, Wales PW. Omega-3 lipids for intestinal fail- 40. Rayyan M, Devlieger H, Jochum F, Allegaert K. Short-term use of paren-
ure associated liver disease. Semin Pediatr Surg. 2009;18:239-245. teral nutrition with a lipid emulsion containing a mixture of soybean oil,
26. Diamond IR, Pencharz PB, Wales PW. What is the current role for paren- olive oil, medium-chain triglycerides, and fish oil: a randomized double-
teral lipid emulsions containing omega-3 fatty acids in infants with short blind study in preterm infants. JPEN J Parenter Enteral Nutr. 2012;36(1)
bowel syndrome? Minerva Pediatr. 2009;61:263-272. (suppl):81S-94S.
27. Josephson J, Turner JM, Field CJ, etal. Parenteral soy oil and fish oil emul- 41. Moss RL, Henry MC, Dimmitt RA, Rangel S, Geraghty N, Skarsgard ED.
sions: impact of dose restriction on bile flow and brain size of parenteral nutri- The role of prospective randomized clinical trials in pediatric surgery: state
tionfed neonatal piglets. JPEN J Parenter Enteral Nutr. 2015;39(6):677-687. of the art? J Pediatr Surg. 2001;36:1182-1186.
28. Rollins MD, Ward RM, Jackson WD, etal. Effect of decreased paren- 42. Hadaway LC. Best-practice interventions: keeping central line infection at
teral soybean lipid emulsion on hepatic function in infants at risk for par- bay. Nursing. 2006;36:58-64.
enteral nutritionassociated liver disease: a pilot study. J Pediatr Surg. 43. Beale EF, Nelson RM, Bucciarelli RL, Donnelly WH, Eitzman DV.
2013;48(6):1348-1356. Intrahepatic cholestasis associated with parenteral nutrition in premature
29. Nehra D, Fallon EM, Carlson SJ, etal. Provision of a soy-based intrave- infants. Pediatrics. 1979;64:342-347.
nous lipid emulsion at 1 g/kg/d does not prevent cholestasis in neonates. 44. Christensen RD, Henry E, Wiedmeier SE, Burnett J, Lambert DK.
JPEN J Parenter Enteral Nutr. 2013;37(4):498-505. Identifying patients, on the first day of life, at high-risk of devel-
30. Cober MP, Killu G, Brattain A, Welch KB, Kunisaki SM, Teitelbaum DH. oping parenteral nutritionassociated liver disease. J Perinatol.
Intravenous fat emulsions reduction for patients with parenteral nutrition 2007;27:284-290.
associated liver disease. J Pediatr. 2012;160(3):421-427. 45. Duro D, Kalish LA, Johnston P, etal. Risk factors for intestinal failure in
31. Rollins MD, Scaife ER, Jackson WD, Meyers RL, Mulroy CW, Book LS. infants with necrotizing enterocolitis: a Glaser Ped Res Network Study.
Elimination of soybean lipid emulsion in parenteral nutrition and supple- J Pediatr. 2010;157:203-208.e1.
mentation with enteral fish oil improve cholestasis in infants with short 46. Fischler L, Meredith DO, Reinhart WH. Influence of a parenteral fish-oil
bowel syndrome. Nutr Clin Pract. 2010;25(2):199-204. preparation (Omegaven) on erythrocyte morphology and blood viscosity
32. Shin JI, Namgung R, Park MS, Lee C. Could lipid infusion be a risk for in vitro. Clin Hemorheol Microcirc. 2003;28:79-88.
parenteral nutritionassociated cholestasis in low birth weight neonates? 47. Wales PW, Allen N, Worthington P, George D, Compher C; the American
Eur J Pediatr. 2008;167(2):197-202. Society for Parenteral and Enteral Nutrition, Teitelbaum D. A.S.P.E.N.
33. Colomb V, Jobert-Giraud A, Lacaille F, Goulet O, Fournet JC, Ricour C. Clinical Guidelines: support of pediatric patients with intestinal failure
Role of lipid emulsions in cholestasis associated with long-term parenteral at risk of parenteral nutritionassociated liver disease. JPEN J Parenter
nutrition in children. JPEN J Parenter Enteral Nutr. 2000;24(6):345-350. Enteral Nutr. 2014;38(5):538-557.
34. Lam HS, Tam YH, Poon TCW, etal. A double-blind randomized con- 48. Seida JC, Mager DR, Hartling L, Vandermeer B, Turner JM. Parenteral
trolled trial of fish oilbased versus soy-based lipid preparations in the -3 fatty acid lipid emulsions for children with intestinal failure and
treatment of infants with parenteral nutritionassociated cholestasis. other conditions: a systematic review. JPEN J Parenter Enteral Nutr.
Neonatology. 2014;105:290-296. 2013;37(1):44-55.
35. Calkins KL, Havranek T, Kelley-Quon L, etal. Low-dose parenteral soy- 49. Vanek VW, Seidner DL, Allen P, etal; Novel Nutrient Task Force,
bean oil for the prevention of parenteral nutritionassociated liver disease in Intravenous Fat Emulsions Workgroup; American Society for Parenteral
neonates with gastrointestinal disorders: a multicenter randomized controlled and Enteral Nutrition (A.S.P.E.N.) Board of Directors. A.S.P.E.N. posi-
pilot study [published online May 29, 2015]. JPEN J Parenter Enteral Nutr. tion paper: clinical role for alternative intravenous fat emulsions. Nutr
36. Gura KM, Duggan CP, Collier SB, etal. Reversal of parenteral nutri- Clin Pract. 2012;27(2):150-192.
tionassociated liver disease in two infants with short bowel syndrome 50. Diamond IR, Grant RC, Feldman BM, etal. Expert beliefs regarding
using parenteral fish oil: implications for future management. Pediatrics. novel lipid-based approaches to pediatric intestinal failureassociated
2006;118(1):e197-e201. liver disease. JPEN J Parenter Enteral Nutr. 2013;38(6):702-710.
37. Goulet O, Antebi H, Wolf C, etal. A new intravenous fat emulsion con- 51. Gastroenterology Regulatory Endpoints and the Advancement of
taining soybean oil, medium-chain triglycerides, olive oil, and fish oil: Therapeutics (GREAT): Parenteral Nutrition-Induced Liver Disease
a single-center, double-blind randomized study on efficacy and safety in (PNALD) workshop. Maryland, September 24, 2012. http://www.regu-
pediatric patients receiving home parenteral nutrition. JPEN J Parenter!documentDetail;D=FDA-2012-N-0001-0093. Accessed
Enteral Nutr. 2010;34(5):485-495. April 20, 2013.