Alimentary Pharmacology and Therapeutics

Randomised clinical trial: enteral nutrition does not improve

the long-term outcome of alcoholic cirrhotic patients with
jaundice
B. Dupont*, T. Dao*, C. Joubert*, C. Dupont-Lucas†, R. Gloro*, E. Nguyen-Khac‡, E. Beaujard§, P. Mathurin¶,
E. Vastel*, M. Musikas*, I. Ollivier* & M.-A. Piquet*

*Department of Hepato- SUMMARY

Gastroenterology and Nutrition, Caen
University Hospital, Caen, France.
†
Department of Statistics, Caen
University Hospital, Caen, France.
‡
Department of Hepato-
Gastroenterology, Amiens University
Hospital, Amiens, France.
§
Department of Hepato-
Gastroenterology, Alençon
Community Hospital, Alençon, France.
¶
Department of Hepato-
Gastroenterology, Lille University
Hospital, Lille, France.
Correspondence to:
Dr B. Dupont, Service d’Hépato-
gastroenterologie, Avenue Côte de
Nacre, CHU Côte de Nacre, 14033
Caen Cedex 9, France.
E-mail: dupont-be@chu-caen.fr
Publication data
Submitted 17 January 2012
First decision 8 February 2012
Resubmitted 4 March 2012
Accepted 5 March 2012
Background
Malnutrition and jaundice are independent prognostic factors in cirrhosis.
Aim
To assess the impact of enteral nutrition on the survival of alcoholic
cirrhotic patients with jaundice but without acute alcoholic hepatitis.
Methods
The study was a multicentre prospective randomised controlled trial compar-
ing effects of enteral nutrition vs. a symptomatic support in patients with alco-
holic cirrhosis and jaundice (bilirubin  51 µmol/L) but without severe acute
alcoholic hepatitis. A total of 99 patients were randomised to receive either the
conventional symptomatic treatment (55 patients) or the symptomatic sup-
port associated with 35 kcal/Kg/day of enteral nutrition during 4 weeks fol-
lowed by an oral nutritional support during 2 months (44 patients).
Randomisation was stratified on nutritional status. One-year survival curves
were compared using the Kaplan–Meier method and Logrank test.
Results
Populations in both arms were similar. One-year survival was similar in the over-
all population (27/44 patients (61.4%) in the enteral nutrition arm vs. 36/55
(65.5%) in the control arm; Logrank P = 0.60) and in the subgroup suffering
from malnutrition [18/29 patients (62.1%) in the enteral nutrition arm vs. 20/32
(62.5%) in the control arm; Logrank P = 0.99]. There was no statistical differ-
ence for bilirubin, prothrombin rate, Child-Pugh score, albumin or nutritional
assessment. Complications during treatment (bleeding, encephalopathy, infec-
tion) occurred in 23% of patients in the enteral nutrition group (10/44) vs. 16%
(9/55) of the control patients (P = 0.59).

Conclusion
Enteral nutrition does not improve the survival and hepatic or nutritional
parameters of cirrhotic patients with jaundice.

Aliment Pharmacol Ther

ª 2012 Blackwell Publishing Ltd 1

doi:10.1111/j.1365-2036.2012.05075.x
B. Dupont et al.
INTRODUCTION
Cirrhosis is one of the major chronic diseases and causes
of death. Mortality from cirrhosis reached age-standar-
dised (world population) rates of 9.8/100 000 in men
and 4.4/100 000 in women in the United States and
15.3/100 000 in men and 5.9/100 000 in women in the
European Union in 2000–2002.1 Malnutrition is com-
monly associated with cirrhosis with prevalence ranging
from 34% to 65%.2–5 The severity of malnutrition is cor-
related with the severity of the liver disease.6 Figueiredo
et al. demonstrated, using a gold standard method of
nutritional assessment (combination of absorptiometry
and isotopic dilution) that the percentage of malnutrition
could, respectively, be 34.4%, 69%, and 94.4% for the
Child-Pugh classes A, B, and C.7 Furthermore, malnutri-
tion increases morbidity from cirrhosis. Many authors
reported an increase in the rate of infections, hepatic
encephalopathy episodes, variceal bleeding, and refrac-
tory ascites in cirrhotic patients with malnutrition.4, 8, 9
Even though the respective roles of malnutrition and cir-
rhosis severity are difficult to distinguish, many authors
have demonstrated that malnutrition in cirrhosis is an
independent prognostic factor for survival in multivariate
analysis.8, 10–13
For this reason, several teams have studied the impact
of nutritional support in cirrhosis. Although, studies on
the impact of oral supplementation show a beneficial
effect on nutritional and liver parameters,14–16 none have
demonstrated an improvement in survival. Out of four
randomised studies dealing with the impact of enteral
nutrition (EN) in cirrhotic patients,17–20 three showed an
improvement in nutritional parameters.17, 19, 20 Cabré
et al. showed a benefit of EN on survival, in a study
bearing on a small population and focusing only on the
short-term impact of EN.17 The long-term impact of EN
in patients with cirrhosis remains unknown.
Jaundice is an independent prognostic factor in cir-
rhosis. Poynard et al. specify that the 1 year overall sur-
vival of cirrhotic patients with jaundice (Bilirubin
 51 µmol/L) is 28% vs. 68% in patients without jaun-
dice.21 Jaundice may reflect either Acute Alcohol Hepati-
tis (AAH) or a terminal phase of cirrhosis. In the case of
AAH, the gold standard treatment of severe forms
defined by the Maddrey score over 32, is corticoster-
oids.22 EN could be an effective alternative treatment in
this indication.23 However, in the case of jaundice with-
out AAH, no gold standard treatment is defined except
for symptomatic treatment. As malnutrition is often
associated with advanced cirrhosis, and due to its prog-
2 Aliment Pharmacol Ther
nostic value, EN seems to be a logical treatment in this
situation. The impact of EN on cirrhotic patients with
jaundice but without AAH has never been studied.
The main goal of this study was to assess the long-
term impact of EN on the survival of alcoholic cirrhotic
patients with jaundice but without severe AAH. We also
studied the impact of this therapy on the hepatic and
nutritional parameters and on the occurrence of compli-
cations of the cirrhosis.
MATERIALS AND METHODS
Population
We conducted a multicentre prospective randomised
controlled trial (with the participation of nine study cen-
tres: the University Hospitals of Caen, Amiens, and Lille;
and the community hospitals of Saint-Lô, Cherbourg,
Bayeux, Argentan and Alençon, France) from June 2004
to December 2008. All patients with liver cirrhosis and
jaundice (bilirubin  51 µmol/L) but without severe
AAH were eligible. The patients were inpatients,
recruited during a hospitalisation for a complication, and
outpatients recruited via consultation at a hospital partic-
ipating in the study.
Inclusion criteria were as follows: patients had to be
over 18 years old, with alcoholic liver cirrhosis (histolog-
ically proven or suspected on clinical and biological
arguments) and jaundice (total bilirubin  51 µmol/L).
All patients provided written informed consent. Exclu-
sion criteria were: the presence of severe AAH assessed
by transjugular liver biopsy when the Maddrey score was
over 32, a non-alcoholic aetiology to cirrhosis, any other
life-threatening disease, a noncontrolled complication of
the cirrhosis (variceal bleeding, infection, etc.), a hepato-
cellular carcinoma and an ongoing corticosteroid treat-
ment.
Study overview
At inclusion, clinical, anthropological and biological data
were collected. A liver biopsy was performed when the
Maddrey score was over 32. The patients were rando-
mised into two groups: a control arm (Ctl) which
received the standard treatment and EN arm which
received EN in addition to the standard treatment. The
standard treatment of cirrhosis consisted of withdrawal
from alcohol intoxication associated with symptomatic
treatment prescribed by the patient’s referring doctor
(B1 vitamins, diuretics, salt restriction, lactulose, beta-
blockers, and antibiotics after bacteriological samples if
ª 2012 Blackwell Publishing Ltd
 Randomised clinical trial: impact of enteral nutrition in cirrhosis
required.). All patients received endoscopic surveillance
for portal hypertension. An adequate prophylaxis by liga-
tion or betablockers was performed respecting Baveno’s
consensus.24, 25 The patients with a past history of spon-
taneous bacterial peritonitis received an antibiotic pro-
phylaxis. This attitude was similar in each hospital.
Oral diet supplied 1800 Kcal per day, 60 g of protein
and 3 g of NaCl. In the EN arm, as recommended by
the European Society for Parenteral and EN, the patients
received 30–35 kcal/kg/day of a polymeric solution (Iso-
source Energy, 1.6 kcal/mL, Novartis Nutrition, France)
for a period of 3–4 weeks, through a nasogastric feeding
tube (Kangaroo, Covidien, France).26 Subsequently they
received three oral nutritional supplements (Clinutren,
Nestlé, France) per day for 2 months.
Patients were followed up at 1 week, 2 weeks,
3 weeks, 1 month, 2 months, 3 months, 6 months, and
1 year after inclusion. Clinical, anthropological and bio-
logical data were collected at each visit. Anthropological
data collected were the weight, the Triceps Skinfold
Thickness (TSF), the Mid-Arm Circumference (MAC)
and the Mid-Arm Muscle Circumference (MAMC)
(MAMC = MAC p 9 TSF). Randomisation was cen-
tralised at the Clinical Research Center of Caen and
equilibrated every four subjects. Randomisation was
stratified by centre and by nutritional status at inclusion.
Malnutrition was defined by a mid-arm muscle circum-
ference lower than the tenth percentile which is a usual
definition for malnutrition in cirrhosis.3
Study outcome
The primary outcome was 1 year overall survival.
The secondary outcomes were the progression of the
liver and nutritional parameters, incidence of complica-
tions (bleeding, infection, hepatic encephalopathy, and
hepatorenal syndrome) and the total number of hospital
days during the year of follow-up. In the case of clinical
hepatic encephalopathy, EN was stopped and precipitat-
ing factors were sought.
Statistical analysis
A sample size calculation determined the number of
patients required to test for superiority of EN over sup-
port treatment alone on 1 year survival. We based our
calculations on an estimated 1 year survival rate for cir-
rhotic patients with jaundice of 70%,21 and on a minimal
survival improvement under EN of 14% as reported by
Kearns et al.19 In the hypothesis of a 1 year survival rate
of 85% in the EN Arm and 70% in the Ctl arm, the
number of subjects was estimated at 67 patients in each
Aliment Pharmacol Ther
ª 2012 Blackwell Publishing Ltd
arm for a total of 134 subjects (a = 0.05 and b = 0.20).
This objective was not reached due to slow inclusion and
financial difficulties, and the study had to be stopped
prematurely in December 2008.
Data are presented as mean (±standard deviation),
median (range) or n (%) as appropriate. Baseline charac-
teristics were compared across groups: for intergroup
comparisons of continuous variables Student’s t-test was
used, and differences between categorical variables were
tested by Pearson chi-square test.
Overall survival was the primary study endpoint,
defined as the time from randomisation until either death
from any cause or liver transplantation. The 1 year sur-
vival curves of the EN group and the control group were
constructed using the Kaplan–Meier method and statisti-
cal differences between curves assessed by log-rank test.
As a secondary outcome, the progression of biological
and nutritional parameters during the year of follow-up
was compared between the intervention and control
groups using linear mixed modelling. The mixed model
was chosen to take into account the intra-patient correla-
tion of these parameters, as repeated measures close in
time can be better correlated than distant measures.
Treatment alone and in combination with time were
considered as fixed effects, with baseline measurement as
a covariate, time (1, 2, 3, 6, and 12 months) as a
repeated factor, and patients as a random factor. The lin-
earity hypothesis was verified for each factor. An
unstructured covariance structure was chosen.
All analyses were conducted in intention to treat. A P
value of 0.05 was taken to be statistically significant. Sta-
tistical analyses were performed using SAS v.9.2 (SAS
Institute Inc., Cary, NC, USA).
Ethics
The medical ethics committee of Basse Normandie
approved the protocol, which had been submitted
previously.
RESULTS
One hundred and one patients were randomised between
June 2004 and December 2008 in the nine study centres
(Figure 1). Two patients were excluded: one for second-
ary withdrawal of consent and another who was trans-
ferred for a severe haemorrhagic complication just after
randomisation to a centre that did not participate in the
study. Thus, total 99 patients took part in the study: 55
in the Ctl arm and 44 in the EN arm. In the Ctl arm, 23
patients were not malnourished vs. 32 malnourished. In
the EN arm, 15 were not malnourished and 29 malnour-
3
B. Dupont et al.
101 Randomised
55 Assigned to receive
standard treatment
55 received standard
treatment as assigned
55 included in analysis
5 lost of follow-up *
Figure 1 | Patient enrolment and randomisation. * For all other randomised patients, 1 year survival data were
collected. Amongst the surviving patients at 1 year (n = 59), seven patients were lost during the follow-up, so data
concerning nutritional and liver parameters were exhaustively collected in only 92.9% (92/99) of the cases. EN,
enteral nutrition.
ished. Three patients had a liver transplant. For statistical
purposes, these patients were censored at the time of
transplantation. For all other randomised patients, 1 year
survival data were collected. Amongst the surviving
patients at 1 year (n = 59), seven patients were lost dur-
ing the follow-up, so data concerning nutritional and
liver parameters were exhaustively collected in only
92.9% (92/99) of the cases. Patients’ characteristics are
described in Table 1. The included patients were mostly
men of mean age 55 years, with active consumption of
alcohol (66/99, 66.7%), advanced cirrhosis stage Child C
(81/99, 81.8%), and ascites (77/99, 77.8%). Sixty-one of
the 99 patients (61.6%) were initially malnourished. A
liver biopsy was performed on 73 of the 99 studied
patients (73.7%): 40 of the 55 patients (72.7%) in the Ctl
arm and 33 of the 44 patients (75%) in the EN arm.
Both groups were comparable in terms of general char-
acteristics at inclusion (Table 1). However, serum albu-
min level was significantly lower and Child-Pugh score
significantly higher in the EN arm. Characteristics of
patients in subgroups of malnourished or well-nourished
were also studied. In the subpopulation suffering from
malnutrition, there was no significant difference between
EN and Ctl groups (data not shown). In the population
with a normal initial nutritional status, a difference was
observed in the Child-Pugh score (9.8 ± 1.6 in the EN
arm vs. 10.8 ± 1.1 in the Ctl arm, P = 0.03) and albumin
4 Aliment Pharmacol Ther
46 Assigned to receive EN
2 Excluded
• 1 withdrawal of consent
• 1 tranfered to a center
that did not participate in
the study
44 received EN
as assigned
44 included in analysis
• 2 lost of follow-up *
• 15 discontinued EN
• 5 failure of tube placement
• 8 poor tolerance
• 2 adverse effect
(29.3 ± 5.3 in the EN arm vs. 25.4 ± 5.4 in the Ctl arm,
P = 0.04).
Overall survival
One-year overall mortality of patients included in the
study was 36.4% (36/99). It was similar in the malnour-
ished and well-nourished subpopulation. Nineteen of the
55 patients (34.5%) included in the Ctl arm died within
1 year of their inclusion vs. 17 of the 44 patients (38.6%)
in the EN arm. One-year overall survival was thus simi-
lar in both arms (P = 0.60) [Figure 2(a)]. The results
were similar as well for the malnourished subpopulation
(P = 0.96) and the well-nourished subpopulation
(P = 0.37) [Figure 2(b)].
Progression of liver parameters
During the year of follow-up, an improvement of all the
liver function tests was noted. Total bilirubin decreased
from 107 µmol/L ± 77 at inclusion to 48 µmol/L ± 39
in the EN arm and from 97 µmol/L ± 39 at inclusion to
41 µmol/L ± 29 in the Ctl arm [Figure 3(a)]. The Child-
Pugh Score and prothrombin rate also improved in both
groups during the year of follow-up (Figures 3(b) and
3(c)]. For these three parameters, the improvement was
statistically significant in time (P < 0.0001 for bilirubin,
P = 0.0001 for prothrombin rate and P < 0.0001 for
Child-Pugh score) but not statistically different between the
ª 2012 Blackwell Publishing Ltd
 Randomised clinical trial: impact of enteral nutrition in cirrhosis

year after inclusion [Figures 4(c) and 4(d)]. Oral caloric

Table 1 | Baseline characteristics of the patients
and protein intakes during the first month were studied
Enteral in a subgroup of 29 patients: 13 in the Ctl arm and 16
Control nutrition in the EN arm. It was an arbitrary sample of the study
Characteristics N = 55 N = 44
population depending on the availability of our dietician
Age (year) 54.6 ± 9.6 56.1 ± 9.6 to calculate an exhaustive evaluation of the oral caloric
Male gender 35 (63.6%) 30 (68.2%)
intake of these patients. Daily oral caloric intakes did not
Patient with active consumption 38 (69.1%) 28 (63.6%)
of alcohol differ between arms of treatment: 1409 kcal/day ± 596 in
Biological parameters the EN arm vs. 1790 kcal/day ± 510 in the Ctl arm
Na (mmol/L) 134 ± 4.7 133 ± 4.3 (P = 0.07). Daily oral protein intakes were also similar in
Creatinine (µmol/L) 77.8 ± 27.2 70.9 ± 23.2 the two groups: 49 g/day ± 24 in the EN arm vs.
Platelets (109/L) 117 ± 71 113 ± 54
Prothrombin rate (%) 49 ± 11.8 45 ± 12.5
60 g/day ± 17 in the Ctl arm (P = 0.17). During the 3–
Total bilirubin (µmol/L) 96.6 ± 39.1 106.8 ± 76.6 4 weeks of enteral feeding in the EN arm, the treatment
Albumin (g/L) 26.8 ± 5.5 24.3 ± 5.5 brought a supplement of 1883 kcal ± 446 and 67 g ± 16
Hepatic parameters of protein per day leading to a total intake of
Maddrey score 37.1 ± 15.4 44.7 ± 20.5
3292 kcal ± 781 per day in the EN group.
Child-Pugh score 10.5 ± 1.5 11.2 ± 1.3
Nutritional parameters
Weight 75.1 ± 18.8 73.6 ± 17.2 Tolerance and complications
MAMC (mm) 221 ± 49.7 214 ± 0.9 The EN was conducted according to the protocol (at
TSF (mm) 15 ± 8.5 14 ± 8.6 least 30 kcal/kg/day for a period of three weeks) in 29 of
Malnourished patients 32 (58.2%) 29 (65.9%)
the 44 patients randomised in the EN arm (65.9%). The
TSF, triceps skinfold; MAMC, mid-arm muscle circumference. mean duration of EN was 2.8 weeks ± 1.2. Five patients
Data except proportion of males are exposed in mean with (11%) had EN for less than one week due to tube place-
standard deviation. ment failure or to poor tolerance of the tube feeding.
During the enteral feeding period, 13 complications
two arms (P = 0.96 for bilirubin, P = 0.62 for prothrombin (bleeding, hepatic encephalopathy, infections or hepato-
rate and P = 0.72 for Child-Pugh score) [Figure 3]. renal syndrome) occurred in 10 of the 44 patients
Mean corpuscular volume, respectively, decreased (22.7%) in the EN arm vs. 13 complications in nine of
from 102.2 µ3 ± 8.9 at inclusion to 95.5 ± 9.3 at the the 55 patients in the Ctl arm (P = 0.59) (Table 2).
sixth month of follow-up in the Ctl arm and from There were seven episodes of hepatic encephalopathy. In
101.8 µ3 ± 7.6 at inclusion to 98.6 ± 7.3 at the sixth five cases, a precipitating factor was identified. In two
month of follow-up in the EN arm (P < 0.0001). This cases no other factor but EN was found. During the year
improvement was similar in both groups (P = 0.80). The after inclusion, 54.4% (30/55) of patients in the Ctl arm
Gamma Glutamyl Transpeptidase (GGT) decreased in vs. 52.3% (23/44) of the patients in the EN arm were
both groups in a similar pattern (P = 0.96). hospitalised (P = 0.982). There was no statistical differ-
ence in the average lengths of hospitalisation between
Progression of nutritional parameters the two arms: 17 days ± 31 in the EN arm vs.
Serum albumin levels increased from 24.3 g/L ± 5.5 at 11.5 days ± 25 in the Ctl arm (P = 0.36).
inclusion to 31.1 g/L ± 6.8 at 1 year in the EN arm. In
the Ctl arm, serum albumin levels increased from DISCUSSION
26.8 g/L ± 5.5 at inclusion to 32.2 g/L ± 6.9 at 1 year We proved the absence of benefit of nutritional support
[Figure 4(a)]. This improvement was statistically signifi- on overall survival in cirrhotic patients with jaundice but
cant in time (P < 0.0001 for albumin) but not statisti- without severe AAH. The absence of impact of EN was
cally different between the two groups (P = 0.31 for observed in the overall population and in the malnour-
albumin). Transthyretin increased in a similar way in ished population. The EN did not improve liver or nutri-
both groups [Figure 4(b)]. The progression was compa- tional parameters either. However, EN did not increase
rable in the malnourished and well-nourished sub-pop- the number of complications of cirrhosis.
ulations. As suggested in previous studies, this trial confirms
The MAMC and TSF were comparable between the the fact that the presence of jaundice in patients with
two arms at inclusion and remained stable during the stable cirrhosis is a very pejorative prognostic factor.

Aliment Pharmacol Ther 5

ª 2012 Blackwell Publishing Ltd
B. Dupont et al.

100
Overall population
Logrank P = 0.60
75

Survival (%)
50 EN arm
Ctl arm
25

(a) Days
0
0 50 100 150 200 250 300 350 400

100 Malnourished subpopulation

Logrank P = 0.96
75
Survival (%)

50

25

(b) Days
0
0 50 100 150 200 250 300 350 400

Figure 2 | One-year overall survival. (a) One-year overall survival in overall population. (b) One-year overall survival in
malnourished population. The 1 year overall survival was similar in EN arm and Ctl arm in overall population as in
malnourished sub-population. EN, enteral nutrition; Ctl, control.

(a) Bilirubin (b) Prothrombin rate

200 90
180 Time: P < 0.0001 80
160 70
140 Group*time: P = 0.96
(µmol/L)

(%)

60
120 50
100
80 40
60 30
Time: P < 0.0001
40 20
20 10 Group*time: P = 0.62
0 0
M0 M1 M2 M3 M6 M12 M0 M1 M2 M3 M6 M12

(c)
Child-Pugh score
14
12
10 EN arm
8
6 Ctl arm
4 Time: P < 0.0001
2 Group*time: P = 0.72
0
M0 M1 M2 M3 M6 M12

Figure 3 | Progression of liver parameters during the year after inclusion in overall population. (a) Progression of total
bilirubin. (b) Progression of prothrombin rate. (c) Progression of Child-Pugh score. Total bilirubin, prothrombin rate or
Child-Pugh score improve during the year of follow-up, but there is no statistical difference in the improvement of
these parameters between the two groups. Data are exposed in mean with standard deviation.

Indeed, our population was essentially composed of patients with cirrhosis and moderate AAH in Menden-
patients with Child-Pugh C cirrhosis and ascites. The hall et al.’s study.27 The prevalence of malnourished
1 year overall survival rate of 66% in our study is com- patients in our study population (61%) is also compara-
parable to the 65% 1 year overall survival rate of the ble to the published data.2, 6

6 Aliment Pharmacol Ther

ª 2012 Blackwell Publishing Ltd
 Randomised clinical trial: impact of enteral nutrition in cirrhosis

(a) Albumin (b) Transthyretin

45 0.25
40 Time: P = 0.01
35 0.2
30 Group*time: P = 0.30

(g/L)
(g/L)
0.15
25
20 0.1
15 Time: P < 0.0001
10 0.05
5 Group*time: P = 0.31
0 0
M0 M1 M2 M3 M6 M12 M0 M1 M2 M3 M6 M12
–0.05

(c) MAMC (d) TSF

300 30
Time: P = 0.18
250 25 Group*time: P = 0.56
(mm)

(mm)
200 20
150 15
Time: P = 0.27
100 10
Group*time: P = 0.25
50 5
0 0
M0 M1 M12 M0 M1 M12

EN arm Ctl arm

Figure 4 | Progression of nutritional parameters during the year after inclusion in overall population. (a) Progression
of albumin. (b) Progression of transthyretin. (c) Progression of MAMC. (d) Progression of TSF thickness. Albumin and
transthyretin score improve during the year of follow-up, but there is no statistical difference in the improvement of
these parameters between the two groups. The MAMC and TSF stay similar during the year of follow-up. Data are
exposed in mean with standard deviation. TSF, triceps skinfold; MAMC, mid-arm muscle circumference.

showed an increase in patient’s calorie intake, as

Table 2 | Complications occurring during the expected in patients benefiting from nutritional support,
treatment period but no improvement of the nutritional parameters
Enteral assessed by anthropometry. The trials of Cabré et al. and
Control Nutrition Kearns et al. showed an improvement of liver parame-
Complications N = 55 N = 44 ters17,19 and the Spanish study even showed an overall
Hepatic encephalopathy 2 5 survival benefit in the arm treated by EN.17 However,
Intestinal haemorrhage 2 2 these studies included less than 20 patients per arm.
Infection 6 4
Cabré et al. and Kearns et al. studies noted an improve-
Hepatorenal syndrome 3 1
Total complications 13 12 ment of hepatic parameters. Nevertheless, their study
Total concerned 9 (16.4%) 10 (22.7%) P = 0.59 populations included a significant number of patients
patients with AAH of unknown severity. Since EN has demon-
There is no statistical difference between the number of
strated its effectiveness in the treatment of AAH,23 the
patients suffering from a complication of liver disease during improvement of hepatic parameters could possibly be
the period of enteral nutrition in the Enteral Nutrition group due to the impact of EN on patients with AAH. It was
and in the control group. also possible that the impact of EN in our study was
minor because of the progress in management of cirrho-
sis and its complications and because of the important
Even though our study population seems to have the rate of effective withdrawal from alcohol intoxication in
expected profile, EN did not demonstrate any benefit. comparison to older trials.
Three prospective randomised trials demonstrated a ben- To our knowledge, our study is the largest published
eficial impact of this treatment in cirrhosis.17, 19, 20 They trial of EN in cirrhotic patients. It demonstrates the

Aliment Pharmacol Ther 7

ª 2012 Blackwell Publishing Ltd
B. Dupont et al.
absence of impact of EN on patients with advanced cir-
rhosis, be it in terms of survival, liver function or nutri-
tional parameters. However, EN could be expected to be
effective in malnourished patients, as malnutrition is a
prognostic factor in cirrhosis.11 In this study, malnutrition
failed to be corrected, despite correct compliance to the
enteral feeding protocol. Two-thirds of the patients in the
EN arm received the complete program of 30 kcal/kg/day
for a period of 3 weeks. Initial EN failures were excep-
tional: less than 10%. This rate of EN failure was not
higher than in Cabre’s study about impact of EN in severe
alcohol-induced hepatitis where it was 22.9%.23 The tech-
nique was generally well accepted.
The failure of EN could be due to three causes: suffi-
cient nutritional intake in the Ctl group, malabsorp-
tion28, 29 and insulin resistance. In our study, the
spontaneous caloric intakes of the patients in the Ctl
arm were near to 20–25 kcal/kg/day, which is less than
recommended26 but more than expected in this popula-
tion. The expected benefit of EN was therefore not
attained, because anorexia was not severe. Malnutrition
could be also due in a part to malabsorption in patients
with chronic jaundice, which could partly explain their
resistance to EN. Finally, patients with cirrhosis develop
insulin resistance, inducing an inappropriate use of car-
bohydrates and an early use of lipid reserves and
decreasing anabolism.30 Insulin resistance increases in
correlation with the severity of cirrhosis. Since our
patients were in the advanced stages of liver disease, the
insulin resistance factor could explain why the malnour-
ished patients in our study were resistant to EN. The
prognosis was so severe that it outweighed the benefit of
EN. This conclusion is supported by the fact that the
mortality was similar in malnourished and well-nour-
ished populations. The prognostic was not essentially
due to the nutritional status but more to the severity of
liver disease. Mean corpuscular volume and GGT
decreased in both arms in a similar pattern (respectively
P = 0.80 and P = 0.96). This data supports the idea that
the effective abstinence from alcohol was similar in the
both arms. If abstinence from alcohol was similar in
both arms, the impact of alcohol consumption on our
results must be estimated as null.
This study has two methodological limits. The two
arms are not equally balanced, with 55 patients in Ctl
arm and 44 in EN arm. This defect is due to the ran-
domisation method. The randomisation was specific to
each centre and balanced by blocks of four subjects. The
number of inclusions was very different from one centre
8 Aliment Pharmacol Ther
to another. However, both arms stayed globally compa-
rable, in particular in the malnourished population
which was our main interest population. The other limit
was the fact that we did not include the projected num-
ber of patients (101 vs. 134). However, even if we had
included the expected number of patients, it would still
be impossible to conclude to a positive impact of EN on
this population. We also encountered difficulties includ-
ing patients due to the reluctance on the part of physi-
cians to use EN in cirrhotic patients. But for us this
refusal is not justified. Because, even if EN did not dem-
onstrate any effectiveness in advanced cirrhosis with
jaundice, this study shows once more that this technique
can be used safely in cirrhosis.17, 19, 23 There were no
more complications in the EN arm than in the Ctl arm.
We described a few cases of hepatic encephalopathy due
to EN that were reversible on the discontinuation of
EN, but these cases were exceptions. The interest of EN
could be tested earlier in the progression of cirrhosis.
It could have an impact at an earlier stage of this
chronic disease when the patient could benefit from this
treatment.
CONCLUSIONS
The EN does not improve the survival of icteric cirrhotic
patients without severe AAH, whether malnourished or
not. Neither does it improve hepatic or nutritional
parameters. Physiopathological arguments lead us to
believe that these patients may be resistant to nutrition.
That is why we conclude that EN is not to be employed
in the treatment of patients with severe cirrhosis but
without AAH. The pertinence and effectiveness of EN
could be studied earlier in the progression of alcoholic
liver disease.
ACKNOWLEDGEMENTS
The authors would like to thank Dr Audrey Dugué for
her help with the statistical analysis. Declaration of per-
sonal interests: Dr Mathurin has received lecture fees and
grant support from, and served as a board member of
Roche; received lecture fees from and served as a board
member of Schering-Plough, Bristol-Myers Squibb, Gi-
lead; received lecture fees from Bayer Healthcare;
received grant support from and served as a board mem-
ber of Janssen-Cilag; and served as a board member of
Norgine. Declaration of funding interests: This study was
funded in full by the Programme Hospitalier de Recher-
che Clinique régional: a financial assistance from the
Department of Health of the French government.
ª 2012 Blackwell Publishing Ltd

REFERENCES
1. Bosetti C, Levi F, Lucchini F, Zatonski
WA, Negri E, La Vecchia C. Worldwide
mortality from cirrhosis: an update to
2002. J Hepatol 2007; 46: 827–39.
2. Nutritional status in cirrhosis. Italian
Multicentre Cooperative Project on
nutrition in liver cirrhosis. J Hepatol
1994; 21: 317–25.
3. Caregaro L, Alberino F, Amodio P,
et al. Malnutrition in alcoholic and
virus-related cirrhosis. Am J Clin Nutr
1996; 63: 602–9.
4. Lautz HU, Selberg O, Korber J, Burger
M, Muller MJ. Protein-calorie
malnutrition in liver cirrhosis. Clin
Investig 1992; 70: 478–86.
5. Thuluvath PJ, Triger DR. Evaluation of
nutritional status by using
anthropometry in adults with alcoholic
and nonalcoholic liver disease. Am J
Clin Nutr 1994; 60: 269–73.
6. Merli M, Riggio O, Dally L. Does
malnutrition affect survival in cirrhosis?
PINC (Policentrica Italiana Nutrizione
Cirrosi). Hepatology 1996; 23: 1041–6.
7. Figueiredo FA, Perez RM, Freitas MM,
Kondo M. Comparison of three
methods of nutritional assessment in
liver cirrhosis: subjective global
assessment, traditional nutritional
parameters, and body composition
analysis. J Gastroenterol 2006; 41: 476–
82.
8. Moller S, Bendtsen F, Christensen E,
Henriksen JH. Prognostic variables in
patients with cirrhosis and oesophageal
varices without prior bleeding. J
Hepatol 1994; 21: 940–6.
9. Pikul J, Sharpe MD, Lowndes R, Ghent
CN. Degree of preoperative
malnutrition is predictive of
postoperative morbidity and mortality
in liver transplant recipients.
Transplantation 1994; 57: 469–72.
10. Abad-Lacruz A, Cabre E, Gonzalez-
Huix F, et al. Routine tests of renal
function, alcoholism, and nutrition
improve the prognostic accuracy of
Child-Pugh score in nonbleeding
advanced cirrhotics. Am J Gastroenterol
1993; 88: 382–7.
11. Alberino F, Gatta A, Amodio P, et al.
Nutrition and survival in patients with
liver cirrhosis. Nutrition 2001; 17: 445–
50.
12. Llach J, Gines P, Arroyo V, et al.
Prognostic value of arterial pressure,
endogenous vasoactive systems, and
Aliment Pharmacol Ther
ª 2012 Blackwell Publishing Ltd
Randomised clinical trial: impact of enteral nutrition in cirrhosis
renal function in cirrhotic patients
admitted to the hospital for the
treatment of ascites. Gastroenterology
1988; 94: 482–7.
13. Mendenhall CL, Moritz TE, Roselle GA,
et al. A study of oral nutritional
support with oxandrolone in
malnourished patients with alcoholic
hepatitis: results of a Department of
Veterans Affairs cooperative study.
Hepatology 1993; 17: 564–76.
14. Cunha L, Happi Nono M, Guibert AL,
Nidegger D, Beau P, Beauchant M.
Effects of prolonged oral nutritional
support in malnourished cirrhotic
patients: results of a pilot study.
Gastroenterol Clin Biol 2004; 28: 36–9.
15. Hirsch S, de la Maza MP, Gattas V,
et al. Nutritional support in alcoholic
cirrhotic patients improves host
defenses. J Am Coll Nutr 1999; 18: 434–
41.
16. Marchesini G, Bianchi G, Merli M,
et al. Nutritional supplementation with
branched-chain amino acids in
advanced cirrhosis: a double-blind,
randomized trial. Gastroenterology 2003;
124: 1792–801.
17. Cabre E, Gonzalez-Huix F, Abad-
Lacruz A, et al. Effect of total enteral
nutrition on the short-term outcome of
severely malnourished cirrhotics. A
randomized controlled trial.
Gastroenterology 1990; 98: 715–20.
18. de Ledinghen V, Beau P, Mannant PR,
et al. Early feeding or enteral nutrition
in patients with cirrhosis after bleeding
from esophageal varices? A randomized
controlled study. Dig Dis Sci 1997; 42:
536–41.
19. Kearns PJ, Young H, Garcia G, et al.
Accelerated improvement of alcoholic
liver disease with enteral nutrition.
Gastroenterology 1992; 102: 200–5.
20. Mendenhall C, Bongiovanni G,
Goldberg S, et al. VA Cooperative
Study on Alcoholic Hepatitis. III:
changes in protein-calorie malnutrition
associated with 30 days of
hospitalization with and without enteral
nutritional therapy. JPEN J Parenter
Enteral Nutr 1985; 9: 590–6.
21. Poynard T, Zourabichvili O, Hilpert G,
et al. Prognostic value of total serum
bilirubin/gamma-glutamyl
transpeptidase ratio in cirrhotic
patients. Hepatology 1984; 4: 324–7.
22. Mathurin P, Mendenhall CL, Carithers
RL Jr, et al. Corticosteroids improve
short-term survival in patients with
severe alcoholic hepatitis (AH):
individual data analysis of the last three
randomized placebo controlled double
blind trials of corticosteroids in severe
AH. J Hepatol 2002; 36: 480–7.
23. Cabre E, Rodriguez-Iglesias P,
Caballeria J, et al. Short- and long-term
outcome of severe alcohol-induced
hepatitis treated with steroids or enteral
nutrition: a multicenter randomized
trial. Hepatology 2000; 32: 36–42.
24. de Franchis R. Updating consensus in
portal hypertension: report of the
Baveno III Consensus Workshop on
definitions, methodology and
therapeutic strategies in portal
hypertension. J Hepatol 2000; 33: 846–
52.
25. de Franchis R. Evolving consensus in
portal hypertension. Report of the
Baveno IV consensus workshop on
methodology of diagnosis and therapy
in portal hypertension. J Hepatol 2005;
43: 167–76.
26. Plauth M, Cabre E, Riggio O, et al.
ESPEN Guidelines on Enteral Nutrition:
liver disease. Clin Nutr 2006; 25: 285–
94.
27. Mendenhall CL, Anderson S, Garcia-
Pont P, et al. Short-term and long-term
survival in patients with alcoholic
hepatitis treated with oxandrolone and
prednisolone. N Engl J Med 1984; 311:
1464–70.
28. Romiti A, Merli M, Martorano M, et al.
Malabsorption and nutritional
abnormalities in patients with liver
cirrhosis. Ital J Gastroenterol 1990; 22:
118–23.
29. Vlahcevic ZR, Yoshida T, Juttijudata P,
Bell CC Jr, Swell L. Bile acid
metabolism in cirrhosis. 3. Biliary lipid
secretion in patients with cirrhosis and
its relevance to gallstone formation.
Gastroenterology 1973; 64: 298–303.
30. Chang WK, Chao YC, Tang HS, Lang
HF, Hsu CT. Effects of extra-
carbohydrate supplementation in the
late evening on energy expenditure and
substrate oxidation in patients with
liver cirrhosis. JPEN J Parenter Enteral
Nutr 1997; 21: 96–9.
9