Division of Gastroenterology and Hepatology, Department of Internal Medicine III, Medical University of Vienna, Vienna,
Austria; 2Vienna Hepatic Hemodynamic Lab, Vienna, Austria; 3Section for Medical Statistics, Center for Medical Statistics,
Informatics, and Intelligent Systems, Medical University of Vienna, Vienna, Austria; and 4Clinical Institute of Hospital Hygiene,
Vienna General Hospital, Vienna, Austria
This article has an accompanying continuing medical education activity on page e14. Learning Objective: Upon completion of this
exam, successful learners will be able to identify patients with cirrhosis and ascites that no longer benet from non-selective
b blocker treatment since the development of spontaneous bacterial peritonitis indicates an event when non-selective betablockers can compromise the circulatory reserve.
Abbreviations used in this paper: AKI, acute kidney injury; CI, condence
interval; CPS, Child-Pugh score; DAP, diastolic arterial pressure; HR,
hazard ratio; HRS, hepatorenal syndrome; IQR, interquartile range; MAP,
mean arterial pressure; NSBB, nonselective b blocker; PMN, polymorphonuclear neutrophil; SAP, systemic arterial pressure; SBP, spontaneous bacterial peritonitis.
2014 by the AGA Institute
0016-5085/$36.00
http://dx.doi.org/10.1053/j.gastro.2014.03.005
b1 adrenoreceptors that are down-regulated and desensitized in patients with advanced cirrhosis.20 Additional
pharmaceutical blockade induces chronotropic incompetence,20 decreases cardiac output,21 and blood pressure and
can increase the rate of HRS development after an SBP
episode, resulting in worse survival.
The aim of this study was to assess the inuence of NSBB
treatment on SBP incidence and the impact of SBP development on the effect of NSBB treatment on hospitalization
and transplant-free survival. In addition, rates of HRS and
acute kidney injury (AKI) development after the rst SBP
diagnosis were assessed in patients with NSBB treatment
and without.
Assessed Parameters
Epidemiological characteristics, etiology of cirrhosis, presence of varices, and information on previous variceal bleeding,
as well as follow-up variceal bleeding and liver transplantation,
were assessed from patient medical history. In addition, information on NSBB and rifaximin treatment, as well as systemic
hemodynamics, was obtained from patient medical history. The
following laboratory parameters were assessed at the time
of the rst paracenteses and the rst diagnosis of SBP: platelet
count, albumin, bilirubin, international normalized ratio,
creatinine, and ascitic uid neutrophil count. Hepatic venous
portal pressure gradient measurements were performed as
described previously.22 The model for end-stage liver disease23
and Child-Pugh score (CPS)24 were calculated based on laboratory parameters and patients medical history.
CLINICAL LIVER
June 2014
1682 Mandorfer et al
90 days after the rst SBP diagnosis. Patients with a history of
chronic kidney disease other than HRS or evidence for parenchymal kidney disease, such as proteinuria, microhematuria, or
abnormal renal ultrasonography were excluded from this
analysis.
AKI was diagnosed if a patient fullled the criteria of group
C of the modied AKI classication proposed by Fagundes
et al,28 comprising both stages 2 and 3 of the Acute Kidney
Injury Network criteria.29 AKI was dened as an increase of
serum creatinine >2-fold from baseline or a serum creatinine
4.0 mg/dL with an acute increase >0.5 mg/dL. In addition,
urine output <0.5 mL/kg per hour for >12 hours was
considered as AKI.
Statistics
CLINICAL LIVER
Statistical analyses were conducted using IBM SPSS Statistics 21 (SPSS Inc., Armonk, NY) and R.3.0.1 (R Development
Core Team 2008, The R Foundation for Statistical Computing,
Vienna, Austria). Continuous variables were reported as mean
standard deviation (SD) or median (interquartile range
[IQR]), and categorical variables were reported as number (n)
of patients with the certain characteristic (proportion of patients with the certain characteristic [%]). Student t test was
used for group comparisons of continuous variables when
applicable. Otherwise, Mann-Whitney U test was applied. Group
comparisons of categorical variables were performed using
either Pearsons c2 or Fishers exact test. The impact of NSBB
treatment on SBP incidence and transplant-free survival was
analyzed using semi-parametric proportional hazard Cox
models. The proportional hazard assumption was veried by
inspecting the cumulative hazard plot for parallel curves.
Patients entered the SBP incidence model (Model 1) with
their rst paracentesis and were followed until their last paracentesis. In liver transplant recipients and in case of death, the
last paracentesis before the liver transplantation or death was
the end of follow-up. Patients who had only one paracentesis,
as well as patients with SBP at their rst paracentesis, were
excluded from this analysis. NSBB treatment and CPS stage
were considered as covariates. In addition, SBP incidence rates
among patients with NSBB treatment, and without, as well as
the SBP incidence rate ratio were calculated.
The rates of follow-up paracentesis and variceal bleeding
after the rst paracentesis were calculated for NSBB and noNSBB patients. Similarly, follow-up paracentesis and bleeding
rates after the rst development of SBP were calculated.
Patients entered the rst transplant-free survival model
(Model 2) with their rst paracentesis and were followed until
2013. Transplant-free survival time was dened as the time to
liver transplantation, death, or end of follow-up. Patients who
received a liver transplantation were censored at the day of
surgery. Besides NSBB treatment, the model included presence
of varices, CPS stage, and SBP status as covariates. The latter
variable was treated as a time-dependent covariate in the
following fashion: The rst development of SBP in patients
without SBP at the rst paracentesis separated the follow-up
into 2 adjoining time intervals. In the rst (right censored)
interval the SBP status is negative, and it is positive in the
second (left or interval censored) time period. To assess the
modulating effect of SBP development on the impact of NSBB
treatment on transplant-free survival, the interaction term
NSBB SBP was incorporated into the model.
Ethics
This study was conducted in accordance with the Declaration of Helsinki and approved by the local ethics committee of
the Medical University of Vienna (EK Nr. 1008/2011).
Results
Patient Characteristics and Systemic
Hemodynamics at the First Paracentesis
Except for a higher proportion of female subjects (NSBB:
34% vs no-NSSB: 27%; P .048) and patients with varices
(NSBB: 90% vs no-NSBB: 62%; P < .001) in the NSBB group,
no statistically signicant differences in patient characteristics between patients with NSBB treatment, and without,
were observed (Table 1).
Although the mean heart rate (NSBB: 77.5 16.5 beats/
min vs no-NSBB: 83.9 15.5 beats/min; P < .001) and systolic arterial pressure (SAP) (NSBB: 114 18 vs no-NSBB:
117 18 mm Hg; P .05) (Figure 1) were lower among
patients with NSBB treatment than in patients without, no
differences in diastolic arterial pressure (DAP) and mean
arterial pressure (MAP) (Figure 1B) were observed. The
proportion of hemodynamically compromised patients, as
dened by a SAP <100 mm Hg (NSBB: 18% vs no-NSBB: 15%;
P .391) (Figure 1C) or an MAP <82 mm Hg (NSBB: 39% vs
no-NSBB: 38%; P .757) (Figure 1D), was comparable between the NSBB treatment groups at the rst paracentesis.
In the NSBB group, daily propranolol doses ranged between 20 mg and 120 mg (20 mg: 10%, 30 mg: 4%, 40 mg:
25%, 50 mg: 1%; 60 mg: 12%; 80 mg: 17%, 100 mg: 1%,
120 mg: 4%), and the administered carvedilol doses ranged
between 6.25 mg and 25 mg (6.25 mg: 7%, 12.5 mg: 18%,
25 mg: 3%).
Age, y, mean SD
Sex, n (%)
Male
Female
Etiology, n (%)
ALD
Viral
ALD and viral
Other
HCC, n (%)
Previous variceal bleeding, n (%)
Varices, n (%)
HVPG, mm Hg, mean SD
MELD, median (IQR)/mean SD
CPS stage, n (%)
A
B
C
Platelet count, 109/L, median (IQR)
Albumin, g/L, mean SD
Bilirubin, mg/dL, median (IQR)
INR, median (IQR)
Creatinine, mg/dL, median (IQR)
Rifaximin treatment, n (%)
Any
Before rst paracentesis
After the rst paracentesis
Heart rate, beats/min, mean SD
SAP, mm Hg, mean SD
SAP <100 mm Hg, n (%)
DAP, mm Hg, mean SD
MAP, mm Hg, mean SD
MAP <82 mm Hg, n (%)
57.5 11.8
426 (70)
181 (30)
No NSBB (n 362)
57.2 12.1
NSBB (n 245)
58 11.2
265 (73)
97 (27)
161 (66)
84 (34)
336 (55)
113 (19)
49 (8)
109 (18)
129 (21)
98 (16)
443 (73)
18.7 6.5a
17.5 (10.6)
192 (53)
70 (19)
27 (7)
73 (20)
74 (20)
54 (15)
223 (62)
18.3 7.1
17.8 (11)
144 (59)
43 (18)
22 (9)
36 (15)
55 (22)
44 (18)
220 (90)
19 5.8
17.2 (9.9)
22
281
304
117
27.2
3.2
1.38
1.14
13 (4)
166 (46)
183 (51)
117 (108)
27 5.4
3.05 (6.98)
1.41 (1.19)
1.14 (0.83)
9
115
121
120
27.6
3.05
1.41
1.14
36 (10)
8 (2)
34 (9)
83.9 15.5
117 18
55 (15)
71 12.5
86.2 13.2
135 (38)
33 (14)
6 (2)
29 (12)
77.5 16.5
114 18
43 (18)
69.8 12.4
84.4 13.2
93 (39)
P value
.438
.048
All (n 182)
No NSBB (n 96)
NSBB (n 86)
57.7 11.1
58.1 10.5
57.3 11.8
76 (79)
20 (21)
58 (67)
28 (33)
134 (74)
48 (26)
.264
(4)
(46)
(50)
(107)
5.7
(6.02)
(0.58)
(0.78)
69 (11)
14 (2)
63 (10)
81.2 16.2c
115 18e
98 (16)e
70.5 12.5
85.4 13.3
228 (38)
(4)
(47)
(49)
(107)
6.1
(5.98)
(0.58)
(0.83)
.553
.318
<.001
.398
.422
.961
.912
.231
.623
.262
.288
.18
.847
.333
<.001
.05
.391
.263
.112
.757
P value
.655
.073
.07
94 (52)
43 (24)
14 (8)
31 (17)
46 (25)
37 (20)
139 (76)
19.3 6.6b
20.8 9
42 (44)
25 (26)
11 (11)
18 (19)
25 (26)
17 (18)
58 (60)
19.1 7.3
20 8.7
52
18
3
13
21
20
81
19.5
21.6
(60)
(21)
(4)
(15)
(24)
(23)
(94)
5.8
9.3
6 (3)
67 (37)
109 (60)
118 (97)
27.1 5.9
3.61 (5.8)
1.41 (0.71)
1.39 (1.26)
3
42
51
119
27.5
3
1.41
1.4
3
25
58
118
26.6
5.11
1.39
1.34
(4)
(29)
(67)
(111)
6.3
(9.32)
(0.69)
(1.31)
23 (13)
9 (5)
21 (12)
79.4 15.7d
109 18
50 (27)
65.7 12.5
80.1 13
97 (53%)
14 (15)
4 (4)
13 (14)
82.9 15.7
113 18
17 (18)
67.3 12.5
82.6 12.5
42 (44)
(3)
(44)
(53)
(86)
5.6
(5)
(0.74)
(1.2)
9 (11)
5 (6)
8 (9)
76.2 15.2
104 17
33 (38)
63.8 12.3
77.2 13.1
55 (64)
.801
.353
<.001
.806
.241
.107
.36
.315
.024
.534
.876
.404
.609
.371
.027
<.001
.002
.06
.005
.006
ALD, alcoholic liver disease; CPS, Child-Pugh score; DAP, diastolic arterial pressure; HCC, hepatocellular carcinoma; HVPG, hepatic venous pressure gradient; INR,
international normalized ratio; MAP, mean arterial pressure; MELD model for end-stage liver disease; SAP, systolic arterial pressure.
a
Available in 220 patients.
b
Available in 82 patients.
c
Available in 395 patients.
d
Available in 109 patients.
e
Available in 595 patients.
CLINICAL LIVER
Patient characteristics
June 2014
Table 1.Patient Characteristics at the First Paracentesis and at the First Development of Spontaneous Bacterial Peritonitis and Comparison of Patients With and Without
Nonselective b Blocker Treatment
1684 Mandorfer et al
CLINICAL LIVER
June 2014
Table 2.Impact of Spontaneous Bacterial Peritonitis Development on the Effect of Nonselective b Blocker Treatment on
Transplant-Free Survival (Model 2): Univariate and Multivariate Analyses Including All patients and Multivariate
Analysis According to Spontaneous Bacterial Peritonitis Status
Patient characteristics
HR
Varices, yes
CPS stage B
CPS stage C
SBP, yes
NSBB, yes
NSBB SBP
1.047
1.076
2.203
1.893
0.945
Varices, yes
CPS stage B
CPS stage C
NSBB, yes
1.112
0.934
1.778
0.75
Lower
Upper
Univariate (n 607)
0.841
1.303
0.584
1.983
1.203
4.035
1.538
2.328
0.775
1.152
Multivariate, no-SBP (n 493)
0.846
1.46
0.455
1.916
0.872
3.625
0.581
0.968
95% CI
P value
HR
.682
.814
.011
<.001
.578
0.983
1.134
2.24
1.401
0.771
1.917
.447
.852
.113
.027
0.703
1.71
3.717
1.58
Lower
Upper
Multivariate (n 607)
0.783
1.235
0.615
2.09
1.222
4.107
1.056
1.857
0.598
0.993
1.256
2.924
Multivariate, SBP (n 182)
0.462
1.068
0.527
5.552
1.157
11.946
1.098
2.274
P value
.886
.687
.009
.019
.044
.003
.098
.372
.028
.014
CLINICAL LIVER
95% CI
1686 Mandorfer et al
CLINICAL LIVER
Table 3.Inuence of Nonselective b Blocker Treatment on Transplant-Free Survival After the First Spontaneous Bacterial
Peritonitis Diagnosis (Model 3) and Survival Rates After Spontaneous Bacterial Peritonitis Development
95% CI
HR
Patient characteristics
Varices, yes
CPS stage B
CPS stage C
NSBB, yes
0.97
1.478
2.954
1.539
Lower
Upper
Univariate (n 182)
0.657
1.432
0.459
4.757
0.932
9.365
1.095
2.164
95% CI
P value
.88
.512
.066
.013
HR
0.695
1.642
3.318
1.644
no-NSBB (n 96)
Estimate
Time point
1 month
6 month
12 month
0.713
0.495
0.358
95% CI
Lower
0.627
0.403
0.271
Upper
0.81
0.608
0.472
Lower
Upper
Multivariate (n 182)
0.457
1.057
0.507
5.314
1.036
10.627
1.145
2.361
NSBB (n 86)
Estimate
0.566
0.279
0.227
95% CI
Lower
0.47
0.198
0.152
Upper
0.682
0.393
0.338
P value
.089
.408
.043
.007
Discussion
A recent landmark study by Serste et al12 demonstrated
reduced survival in patients with refractory ascites who were
treated with NSBBs and has initiated a lively debate among
gastroenterologists on the appropriate use of NSBBs in patients with advanced cirrhosis. In summary, the editorial30
and various other responses to this study brought up concerns about the unequal distribution of esophageal varices
between the NSBB and no-NSBB group. In addition, the
trends toward a higher prevalence of CPS stage C and HCC in
patients with NSBB treatment31,32 and the lack of consecutive
patient enrollment33 were highlighted as factors signicantly
limiting the conclusions drawn within the study. We aimed to
address these limitations in our study by adjusting for the
presence of varices in all analyses of transplant-free survival.
In addition, all analyses were adjusted for CPS stage, as at the
rst SBP diagnosis, there was a trend toward a lower prevalence of CPS stage B and a higher prevalence of CPS stage C
among patients with NSBB treatment. Except for a higher
proportion of females in the NSBB group at the rst paracentesis and higher bilirubin levels in the NSBB group at the
rst SBP diagnosis, no other statistically signicant differences in patient characteristics were observed. In order to
avoid multicollinearity, bilirubin was not considered as a
covariate, as it is a component of the CPS. As our study
population comprises consecutive patients meeting the inclusion and exclusion criteria of this study, our results are
based on a large, representative cohort treated at a single
academic center.
Another response by Senzolo et al11 addressed a potential benecial effect of NSBB treatment on SBP development, highlighting a positive meta-analysis by the same
author.11 In our study, no statistically signicant effect of
NSBB treatment on SBP incidence was observed. As patients who had only one paracentesis and patients with
SBP at their rst paracentesis were excluded and the
CLINICAL LIVER
June 2014
1688 Mandorfer et al
CLINICAL LIVER
Supplementary Material
Note: To access the supplementary material accompanying
this article, visit the online version of Gastroenterology at
www.gastrojournal.org, and at http://dx.doi.org/10.1053/
j.gastro.2014.03.005.
References
1. Blachier M, Leleu H, Peck-Radosavljevic M, et al. The
burden of liver disease in Europe: a review of available
epidemiological data. J Hepatol 2013;58:593608.
2. Asrani SK, Larson JJ, Yawn B, et al. Underestimation of
liver-related mortality in the United States. Gastroenterology 2013;145:375382, e12.
3. DAmico G, Garcia-Tsao G, Pagliaro L. Natural history
and prognostic indicators of survival in cirrhosis: a systematic review of 118 studies. J Hepatol 2006;44:
217231.
4. Ge PS, Runyon BA. The changing role of beta-blocker
therapy in patients with cirrhosis. J Hepatol 2014;60:
643653.
5. Peck-Radosavljevic M, Angermayr B, Datz C, et al.
Austrian consensus on the denition and treatment of
portal hypertension and its complications (Billroth II).
Wien Klin Wochenschr 2013;125:200219.
6. de Franchis R. Revising consensus in portal hypertension: report of the Baveno V consensus workshop on
methodology of diagnosis and therapy in portal hypertension. J Hepatol 2010;53:762768.
CLINICAL LIVER
June 2014
1690 Mandorfer et al
41. Krag A, Bendtsen F, Moller S. Does cardiac dysfunction explain deleterious effects of beta-blockers in
cirrhosis and refractory ascites? Hepatology 2011;53:
370371.
CLINICAL LIVER
June 2014