Journal of Pediatric Surgery (2011) 46, 1913–1917

www.elsevier.com/locate/jpedsurg

A contemporary analysis of parenteral nutrition–associated

liver disease in surgical infants
Patrick J. Javid a,⁎, Frances R. Malone b , André A.S. Dick b , Evelyn Hsu c ,
Maria Sunseri a , Patrick Healey a,b , Simon P. Horslen b,c
a
Division of Pediatric General and Thoracic Surgery, Seattle Children's Hospital, University of Washington, Seattle,
WA 98105, USA
b
Division of Transplantation, Seattle Children's Hospital, University of Washington, Seattle, WA 98105, USA
c
Division of Gastroenterology, Seattle Children's Hospital, University of Washington, Seattle, WA 98105, USA

Received 11 July 2010; revised 4 May 2011; accepted 1 June 2011

Key words:
Abstract
Parenteral nutrition;
Background/purpose: Despite advances in pediatric nutritional support and a renewed focus on
Cholestasis;
management of intestinal failure, there are limited recent data regarding the risk of parenteral nutrition
Liver disease;
(PN)–associated liver disease in surgical infants. This study investigated the incidence of cholestasis
Neonatal surgery;
from PN and risk factors for its development in this population.
Abdominal surgery
Methods: A retrospective review was performed of all neonates in our institution who underwent
abdominal surgery and required postoperative PN from 2001 to 2006. Cholestasis was defined as 2
conjugated bilirubin levels greater than 2 mg/dL over 14 days. Nonparametric univariate analyses and
multivariate logistic regression were used to model the likelihood of developing cholestasis. Median
values with range are presented.
Results: One hundred seventy-six infants met inclusion criteria, and patients received PN for 28 days
(range, 2-256 days). The incidence of cholestasis was 24%. Cholestatic infants were born at an earlier
gestational age (34 vs 36 weeks; P b .01), required a 3-fold longer PN duration (76 vs 21 days; P b
.001), had longer inpatient stays (86 vs 29 days; P b .001), and were more likely to be discharged on
PN. The median time to cholestasis was 23 days. Cholestasis was an early development; 77% of
cholestatic infants developed cholestasis by 5 weeks of PN exposure. On multivariate regression, only
prematurity was significantly associated with development of cholestasis (P b .05).
Conclusion: In this analysis, the development of PN-associated liver disease occurred early in the course
of exposure to PN. These data help to define the time course and prognosis for PN-associated cholestasis
in surgical infants.
© 2011 Elsevier Inc. All rights reserved.

Liver disease associated with the administration of

parenteral nutrition (PN) remains a challenge in the care of
⁎ Corresponding author. Tel.: +1 206 987 6129; fax: +1 206 987 3925. pediatric surgical patients. Among infants who receive
E-mail address: patrick.javid@seattlechildrens.org (P.J. Javid). prolonged courses of PN, the incidence of liver injury ranges

0022-3468/$ – see front matter © 2011 Elsevier Inc. All rights reserved.
doi:10.1016/j.jpedsurg.2011.06.002
1914
from 18% to 67% and is characterized histologically by bile
duct proliferation, portal inflammation, bridging fibrosis, and
ultimately cirrhosis [1-5]. Clinically, elevation in serum
conjugated bilirubin is observed at an early stage.
Several promising advances in the nutritional support of
surgical infants have evolved over the past decade. These
include an early emphasis on enteral nutrition, revised
guidelines for caloric allotment in the critically ill infant, new
elemental formulas for children with feeding intolerance, and
the administration of ursodiol for hyperbilirubinemia [6-11].
Despite this progress in pediatric nutrition, new data on PN-
associated cholestasis are lacking, and recent data have
focused primarily on specific subgroups of neonates [12-14].
In this study, we sought to measure the contemporary
incidence of PN-associated cholestasis in a population of
surgical infants. We chose to study all neonates who
underwent abdominal surgery at our institution over a
multiyear period to better understand the occurrence of PN-
associated cholestasis as well as the time course and risk
factors for its development in a broad surgical population.
We hypothesized that the rate of PN-associated cholestasis in
surgical infants would be lower than prior estimates, given
the recent renewed clinical emphasis on nutritional support
and enteral nutrition in these infants.
1. Methods
After receiving approval from the Seattle Children's
Hospital Institutional Review Board, we conducted a
retrospective review of all infants who underwent abdominal
surgery at less than 30 days of age and received
postoperative PN at our institution from January 2001
through June 2006. Variables collected included gestational
age, surgical diagnosis, length of stay, mortality, duration of
PN, and serum bilirubin values. Patient data and outcomes
were measured to the date of initial discharge from the
hospital. Patients who died at less than 14 days of age were
excluded from this analysis.
Cholestasis was defined as 2 consecutive conjugated
bilirubin levels greater than 2 mg/dL over a minimum of 14
days. Per institutional protocol, infants on PN underwent serum
sampling for liver function on a weekly basis. All patients less
than 10 days of age who met the definition for cholestasis were
reviewed to rule out indirect hyperbilirubinemia as the primary
etiology; subjects were deemed to have cholestasis only when
the unconjugated bilirubin had normalized.
Our institutional protocol for PN used estimated energy
requirements of 90 to 120 and 85 to 105 kcal/kg per day for
preterm and term infants, respectively. Macronutrient goals
for full PN support included a dextrose infusion rate of 10 to
14 mg/kg per minute, protein allotment of 2 to 4 g/kg per
day, and lipid provision of 2.5 to 3 g/kg per day. Parenteral
nutrition was ordered by the pediatric surgical team for all
patients except for extremely low-birth-weight neonates who
P.J. Javid et al.
primarily resided on the neonatal intensive care service; for
these subjects, the composition of PN was managed by the
neonatal intensive care service with input from the attending
pediatric surgeon. In general, ursodiol (30 mg/kg per day
divided in 3 doses) was administered to patients who met the
definition for cholestasis. Lipid modification, including
lower lipid allotment and ω-3 fatty acid supplementation,
was not used during this study period. Initiation and
advancement of enteral nutrition were encouraged, but the
ultimate enteral nutrition regimen was provider dependent.
Basic descriptive statistics were calculated for the
population. Nonparametric univariate analyses (eg, 2-sample
Wilcoxon's test for continuous variables and Fisher's Exact
test for categorical variables) were used to compare the
characteristics of patients who developed cholestasis with
that of those who did not develop cholestasis. A Kruskal-
Wallis test was used to test differences in the distribution of
continuous measures by diagnosis. Multivariate exact
logistic regression was used to evaluate the association
between diagnosis, prematurity (defined as b36 weeks
gestation), and PN duration before onset of cholestasis
with the risk of developing cholestasis. P b .05 was
considered to be statistically significant. SAS 9.1.3 (SAS,
Cary, NC) was used for all analyses. Median values with
range are presented.
2. Results
One hundred seventy-six infants met study criteria, and
Table 1 lists the surgical diagnoses in this cohort. The overall
mortality rate was 5%. Infants were exposed to PN for a
median duration of 28 days (range, 2-256 days; Fig. 1). The
overall incidence of cholestasis was 24%, and the develop-
ment of cholestasis was an early phenomenon. The median
time to cholestasis was 23 days, and 77% of infants who
developed cholestasis did so within 5 weeks of PN exposure.
Fig. 2 illustrates the time to development of cholestasis as
well as the prevalence of cholestasis at each week of PN
exposure. Although all cholestatic infants met criteria for
cholestatic liver disease by 9 weeks of PN exposure, 40% of
infants on PN at this time never developed cholestasis.
Table 1 Surgical diagnoses in the study population
Surgical diagnosis No. of subjects %
Necrotizing enterocolitis 48 27
Gastroschisis 43 24
Jejunoileal atresia 27 15
Duodenal atresia 25 14
Omphalocele 15 9
Volvulus 5 3
Meconium disease 6 3
Other 7 4
PN - associated liver disease in surgical infants 1915
100

90

80

% Subjects on PN
70

60

50

40

30

20

10

0
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38
Weeks

Fig. 1 The duration of PN administration is illustrated. The median duration of PN exposure was 28 days.

Differences in the incidence of cholestasis by surgical 3. Discussion

diagnosis were not statistically significant (Fig. 3).
On univariate analysis, infants who developed cholestasis Cholestatic liver disease remains a frequent complication
were born at an earlier gestational age (34 vs 36 weeks; of PN administration. Although multiple theories have been
P b .01), required a 3-fold longer duration of PN (76 vs 21 proposed to explain its pathogenesis, the precise mechanism
days; P b .001), had longer inpatient lengths of stay (86 vs remains unknown and is thought to be multifactorial [5,15].
29 days; P b .001), and were more likely to be discharged on Recently, alternative lipid solutions rich in ω-3 fatty acids
PN (17% vs 6%; P = .05). Inpatient mortality was significant have shown promise in reversing the biochemical changes
higher in infants with cholestasis (17% vs 2%; P b .05). associated with PN-associated liver disease [16,17].
Using a multivariate regression model adjusted for diagnosis In the past decade, various advances in the nutritional
and duration of PN before the onset of cholestasis, only support of surgical neonates have been reported. For
prematurity was significantly associated with the develop- example, it has been demonstrated that critically ill and
ment of cholestasis. As listed in Table 2, full-term infants postoperative infants do not increase their energy expendi-
were associated with one-third the risk of developing ture, and revised caloric allotments have been suggested
cholestasis as compared with premature infants (odds ratio, [11]. There has been a renewed focus on the provision of
0.34; 95% confidence interval, 0.12-0.87; P b .05). adequate protein to optimize wound healing, minimize
whole-body protein breakdown, and improve overall devel-
100 opmental outcome [18-20]. Techniques for continuous and
90
postpyloric feedings as well as elemental and hypoallergenic
formulas have helped to promote the use of enteral nutrition
80
[9,18]. Finally, ursodiol has been studied for its effects on
Percentage of Subjects

70
reducing hyperbilirubinemia associated with PN, although
60
50
% of patients with Cholestasis

40 z
40
30 35
20 30
Time to development of cholestasis 25
10 20
Prevalence of cholestasis
0 15
1 2 3 4 5 6 7 8 9 10
Duration of PN (Weeks) 5
0
EC

se

us

er

Fig. 2 The development of cholestasis in this series was an early

si

si

si

el

th
ea

ul
hi

re

tre

oc
N

O
lv
sc

At

s
al
la

Vo
di

phenomenon. The median time to cholestasis was 23 days, and 77%

tro

ph
na

m
as

iu
de

of cholestatic subjects met the definition for liver injury by 5 weeks

O
G

on
uo

ec
D

of PN exposure. Conversely, the prevalence of cholestasis was

M

lower than expected at all time-points. Even at 9 weeks of PN

exposure, when all cholestatic subjects had developed liver injury, Fig. 3 There were no differences in the incidence of cholestasis
40% of subjects remained free of cholestasis. when compared by diagnostic cohort.
1916
Table 2 Multivariate regression analysis
Exact odds ratio for developing P
cholestasis (95% CI)
Diagnosis (reference, NEC)
Atresia 0.59 (0.12-2.39)
Duodenal atresia 0.77 (0.19-2.87)
Gastroschisis 0.81 (0.24-2.66)
Meconium disease 1.55 (0.11-14.78)
Volvulus 0.71 (0.01-9.82)
Omphalocele 0.85 (0.11-4.96)
Other 0.34 (0-2.75)
Time on PN (before 0.99 (0.96-1.01)
cholestasis)
Gestational age (≥36 0.34 (0.12-0.87)
vs b36 weeks)
CI indicates confidence interval; NEC, necrotizing enterocolitis.
recent prospective data question its efficacy in preventing
hyperbilirubinemia [10,21].
This study sought to review a single institution's
contemporary experience with PN-associated cholestasis
over a multiyear period in all infants who underwent an
abdominal operation within the first 30 days of life. Hence, our
study population was broad and included specific surgical
diagnoses that are not classically associated with prolonged PN
administration. Unlike other recently published data, however,
this study did not exclude subjects based on gestational age,
birth weight, surgical diagnosis, or duration of PN [12-14]. The
data were collected before the implementation of new
parenteral lipid minimization regimens and the use of ω-3
fatty acid supplementation to determine our baseline incidence
of cholestasis. We felt that these data would be helpful in
counseling families of postsurgical newborns as to the risks of
developing PN-associated liver disease.
Overall, the incidence of PN-associated liver injury was
lower than expected. This is explained in part by the fact that
the study population included subgroups of children not
expected to have high rates of PN-associated cholestasis,
such as infants with omphalocele or duodenal atresia.
Nonetheless, the incidence of cholestasis remained lower
than other published data when stratified by specific surgical
diagnosis (Fig. 3). For example, infants with gastroschisis
had an incidence of cholestasis of only 21% despite
consistent and prolonged exposure to PN.
More importantly, we found that the development of
cholestasis from PN was an early phenomenon. The median
time to cholestasis was just over 3 weeks, and most infants
with cholestasis had developed biochemical liver disease by
5 weeks of PN exposure. By 5 months of PN exposure, all
infants still dependent on PN had developed cholestasis.
However, all of these infants on long-term PN had already
developed cholestasis at a much earlier time-point. In fact,
every subject in this long-term PN subgroup had developed
cholestasis by 7 weeks of PN exposure. These data regarding
the early time course for PN-associated liver disease are
P.J. Javid et al.
corroborated by a recent analysis of cholestasis in the
neonatal population at a separate institution [22].
Taken together, these data suggest that factors in addition
to prolonged PN exposure may play a role in the development
of cholestatic liver disease. For example, it is possible that an
.6172
injurious element in PN may act as a source for PN-associated
.8994
.9139 cholestasis, rather than a chronic deficiency of trophic factors
1 in enteral nutrition. Another possibility is a more recent
1 hypothesis that standard parenteral lipid solutions may create
1 a proinflammatory state, which directly results in liver injury
.3457 [23]. It is also plausible that an early external variable, such as
.2002 sepsis, hemodynamic instability, or administration of hepa-
totoxic medications, predisposed the surgical neonate to the
.0222 development of cholestasis. The collection of these variables
was beyond the scope of this retrospective study. Regardless
of the mechanism, the early development of cholestasis
suggests that early initiation of enteral nutrition may be
beneficial when possible.
Clearly, there are deficiencies in this study that must be
addressed. The data collection was retrospective in nature and
encompassed several years, thereby increasing the potential
for error. In addition, data on intestinal length and specific
allotments of enteral nutrition were not consistently available.
The absence of the latter data is important because increasing
amounts of enteral nutrition have been shown to ameliorate
the progression of cholestasis in PN-dependent children [24].
Hence, the degree of PN support (ie, the percent caloric
allotment) was not factored into the multivariate analysis, and
it is possible that a prolonged duration of high levels of PN
support would be associated with cholestasis. Finally, since
the accrual of these data, there are now a number of
innovative nutritional modalities to treat cholestasis in high-
risk groups, such as ω-3 fatty acid supplementation and lipid
minimization strategies. Nonetheless, the typical surgical
neonate still receives similar PN support today.
The data presented here have direct relevance to the current
care of surgical infants. This study has helped to define the
prognosis for the development of PN-associated cholestasis in
young surgical infants. These data may be used to compare
rates of cholestasis with more recent and innovative nutritional
therapies designed to prevent PN-associated liver disease.
From these data, it is likely that the development of cholestatic
liver injury occurs early in the postoperative course of surgical
infants. Postoperative infants who are dependent on PN but do
not develop cholestasis within the first 2 months of PN
exposure may have a reasonable chance of avoiding further
liver injury. To better define the risk factors associated with
PN-associated cholestasis, large prospective multicenter
analyses will be required.
References
[1] Moss RL, Das JB, Raffensperger JG. Total parenteral nutrition–
associated cholestasis: clinical and histopathologic classification.
J Pediatr Surg 1993;28:1270-4.
PN - associated liver disease in surgical infants
[2] Sondheimer JM, Asturias E, Cadnapaphornchai M. Infection and
cholestasis in neonates with intestinal resection and long-term
parenteral nutrition. J Pediatr Gastroenterol Nutr 1998;27:131-7.
[3] Suita S, Masumoto K, Yamanouchi T, et al. Complications in neonates
with short bowel syndrome and long-term parenteral nutrition. JPEN J
Parenteral Enteral Nutr 1999;23(5 Suppl):S106-9.
[4] Albers MJ, de Gast-Bakker DA, van Dam NA, et al. Male sex
predisposes the newborn surgical patient to parenteral nutrition–
associated cholestasis and to sepsis. Arch Surg 2002;137:789-93.
[5] Teitelbaum DH, Trace T. Parenteral nutrition–associated cholestasis.
Sem Pediatr Surg 2001;10:72-80.
[6] Duro D, Kamin D, Duggan C. Overview of pediatric short bowel
syndrome. J Pediatr Gastroenterol Nutr 2008;47:S33-6.
[7] Krawinkel MB. Parenteral nutrition–associated cholestasis—what do
we know, what can we do? Eur J Pediatr Surg 2004;14:230-4.
[8] Suita S, Yamanouchi T, Masumoto K, et al. Changing profile of
parenteral nutrition in pediatric surgery; a 30-year experience at one
institution. Surgery 2002;131(1 Suppl):S275-82.
[9] de Lucas C, Moreno M, Lopez-Herce J, et al. Transpyloric enteral
nutrition reduces the complication rate and cost in the critically ill
child. J Pediatr Gastroenterol Nutr 2000;30:175-80.
[10] San Luis VA, Btaiche IF. Ursodiol in patients with parenteral
nutrition–associated cholestasis. Ann Pharmacother 2007;41:1867-72.
[11] Agus MS, Jaksic T. Nutritional support of the critically ill child. Curr
Op Pediatr 2002;14:470-81.
[12] Aspirot A, Su W, Flageole H, et al. Cholestasis associated with small bowel
atresia: do we always need to investigate? J Pediatr Surg 2007;42:873-7.
[13] Robinson DT, Ehrenkranz RA. Parenteral nutrition–associated cholestasis
in small for gestational age infants. J Pediat 2008;152:59-62.
[14] Hsieh MH, Pai W, Tseng HI, et al. Parenteral nutrition–associated
cholestasis in premature babies: risk factors and predictors. Pediatr
Neonatol 2009;50:202-7.
1917
[15] Carter BA, Shulman RJ. Mechanisms of disease: update on the
molecular etiology and fundamentals of parenteral nutrition-
associated cholestasis. Nat Clin Practice Gastroenterol Hepatol
2007;4:277-87.
[16] Puder M, Valim C, Meisel JA, et al. Parenteral fish oil improves
outcomes in patients with parenteral nutrition-associated liver injury.
Ann Surg 2009;250:395-402.
[17] de Meijer VE, Gura KM, Le HD, et al. Fish oil-based lipid
emulsions prevent and reverse parenteral nutrition-associated liver
disease: the Boston experience. JPEN J Parenteral Enteral Nutr
2009;33:541-7.
[18] Vlaardingerbroek H, van Goudoever JB, van den Akker CH. Initial
nutritional management of the preterm infant. Early Hum Devel
2009;85:691-5.
[19] Burrin DG, Davis TA. Proteins and amino acids in enteral nutrition.
Curr Op Clin Nutr Metab Care 2004;7:79-87.
[20] Premji S, Fenton T, Sauve R. Does amount of protein in formula matter
for low-birthweight infants? A Cochrane systematic review. JPEN J
Parenteral Enteral Nutr 2006;30:507-14.
[21] Arslanoglu S, Moro GE, Tauschel HD, et al. Ursodeoxycholic acid
treatment in preterm infants: a pilot study for the prevention of
cholestasis associated with total parenteral nutrition. J Pediatr
Gastroenterol Nutr 2008;46:228-31.
[22] Nasr A, Diamond IR, de Silva NT, et al. Is the use of omega-3 lipid
emulsions justified in surgical neonates with mild parenteral nutrition-
associated liver dysfunction? J Pediatr Surg 2010;45:980-6.
[23] Diamonr ID, Sterescu A, Pencharz PB, et al. Changing the paradigm:
omegaven for the treatment of liver failure in pediatric short bowel
syndrome. J Pediatr Gastroenterol Nutr 2009;48:209-15.
[24] Javid PJ, Collier S, Richardson D, et al. The role of enteral nutrition in
the reversal of parenteral nutrition-associated liver dysfunction in
infants. J Pediatr Surg 2005;40:1015-8.