Embed Size (px)
DESCRIPTION
journal
Citation preview
Gastroenterology 2014;146:1680–1690
CLINICALLIVER
Nonselective b Blockers Increase Risk for HepatorenalSyndrome and Death in Patients With Cirrhosisand Spontaneous Bacterial PeritonitisMattias Mandorfer,1,2 Simona Bota,1,2 Philipp Schwabl,1,2 Theresa Bucsics,1,2
Nikolaus Pfisterer,1,2 Matthias Kruzik,1,2 Michael Hagmann,3 Alexander Blacky,4
Arnulf Ferlitsch,1,2 Wolfgang Sieghart,1,2 Michael Trauner,1,2 Markus Peck-Radosavljevic,1,2
and Thomas Reiberger1,2
1Division 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 thisexam, successful learners will be able to identify patients with cirrhosis and ascites that no longer benefit from non-selectiveb blocker treatment since the development of spontaneous bacterial peritonitis indicates an event when non-selective beta-blockers can compromise the circulatory reserve.
Abbreviations used in this paper: AKI, acute kidney injury; CI, confidenceinterval; 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, poly-morphonuclear neutrophil; SAP, systemic arterial pressure; SBP, spon-taneous bacterial peritonitis.
© 2014 by the AGA Institute0016-5085/$36.00
http://dx.doi.org/10.1053/j.gastro.2014.03.005
See Covering the Cover synopsis on page 1583;see editorial on page 1597.
BACKGROUND & AIMS: Nonselective b blockers (NSBBs)reduce portal pressure and the risk for variceal hemorrhage inpatients with cirrhosis. However, development of spontaneousbacterial peritonitis (SBP) in these patients could precludetreatment with NSBBs because of their effects on the circu-latory reserve. We investigated the effects of NSBBs in pa-tients with cirrhosis and ascites with and without SBP.METHODS: We performed a retrospective analysis of datafrom 607 consecutive patients with cirrhosis who had theirfirst paracentesis at the Medical University of Vienna from2006 through 2011. Cox models were calculated to investi-gate the effect of NSBBs on transplant-free survival timeand adjusted for Child-Pugh stage and presence of varices.RESULTS: NSBBs increased transplant-free survival inpatients without SBP (hazard ratio ¼ 0.75; 95% confidenceinterval: 0.581–0.968; P ¼ .027) and reduced days of non-elective hospitalization (19.4 days/year for patients on NSBBsvs 23.9 days/year for patients not taking NSBBs). NSBBs hadonly moderate effects on systemic hemodynamics at patients’first paracentesis. However, at the first diagnosis of SBP, theproportion of hemodynamically compromised patients withsystolic arterial pressure <100 mm Hg was higher amongthose who received NSBBs (38% vs 18% of those not takingNSBBs; P ¼ .002), as was the proportion of patients witharterial pressure <82 mm Hg (64% of those taking NSBBs vs44% of those not taking NSBBs; P ¼ .006). Among patientswith SBP, NSBBs reduced transplant-free survival (hazardratio ¼ 1.58; 95% confidence interval: 1.098–2.274; P ¼ .014)and increased days of nonelective hospitalization (29.6 days/person-year in patients on NSBBs vs 23.7 days/person-yearin those not taking NSBBs). A higher proportion of patients
on NSBBs had hepatorenal syndrome (24% vs 11% in thosenot taking NSBBs; P ¼ .027) and grade C acute kidney injury(20% vs 8% for those not taking NSBBs; P ¼ .021). CON-CLUSIONS: Among patients with cirrhosis and SBP, NSBBsincrease the proportion who are hemodynamically compro-mised, time of hospitalization, and risks for hepatorenalsyndrome and acute kidney injury. They also reducetransplant-free survival. Patients with cirrhosis and SBPshould not receive NSBBs.
Keywords: Liver Fibrosis; Bacterial Infection; Mortality;b Blocker; Adrenergic Receptor.
irrhosis accounts for 1.8% of all deaths in Europe1th
Cand is the 12 leading cause of death in theUnited States, and a recent report suggests that even thismight be an underestimation.2
According to a prognostic model proposed by D’Amicoet al,3 the occurrence of varices initiates the second stage ofcirrhosis, and the third stage is defined by the developmentof ascites. Variceal hemorrhage, the most common lethalcomplication of cirrhosis, results directly from portal hy-pertension and triggers the fourth stage of cirrhosis.
June 2014 b Blockers in Spontaneous Bacterial Peritonitis 1681
CLINICAL
LIVE
R
Nonselective b blockers (NSBBs) have been shown toreduce the hepatic venous portal pressure gradient bylowering cardiac output and splanchnic vasoconstrictionthrough the inhibition of the binding of catecholamines to b1and b2 adrenoreceptors.4
Based on a broad body of evidence, current guidelines5–7
recommend NSBB treatment for the prevention of the firstvariceal hemorrhage and rebleeding. NSBBs are among themost widely used drugs in patients with cirrhosis and haveeven been referred to as “the aspirin of hepatologists.”4
However, the role of b blockers appears to have changedover the years.4 Although the efficacy of NSBB treatment inprevention of variceal hemorrhage was established as earlyas 1981,8 recent studies suggest additional beneficial effectsof NSBB treatment. In patients with compensated cirrhosis,hemodynamic response to NSBBs has been shown to reducethe risk of developing ascites, refractory ascites, and hep-atorenal syndrome (HRS).9 In addition, NSBB treatment hasbeen shown to decrease intestinal permeability indepen-dently of hemodynamic response10 and to prevent thedevelopment of spontaneous bacterial peritonitis (SBP).11 Incontrast, a recent study by Serste et al12 has demonstrated adetrimental effect of NSBB treatment in patients withcirrhosis and refractory ascites and has initiated an intensedebate on whether the use of NSBBs should be contra-indicated in this group of patients.
The window hypothesis,13 proposed by Krag et al, isconsistent with these results and implies the existence of atherapeutic window for NSBB treatment, which closes latein the natural course of cirrhosis. Asmeta-analysis has shownthat the reduction of the risk of variceal hemorrhage by NSBBtreatment is significantly lower among patients withascites,14 the end of this therapeutic window might be ex-pected at some point after the first development of ascites.
The occurrence of bacterial infections, even with recov-ery, can establish an additional fifth stage of cirrhosis, termedthe critically ill patient with cirrhosis.15 These bacterial in-fections predominately occur in advanced cirrhosis, which isassociated with profound changes in systemic hemody-namics, most importantly peripheral vasodilation with animpaired responsiveness to vasoconstrictors16 and a simul-taneous increase in cardiac output to maintain adequateorgan perfusion.13 However, parallel to the progression ofhyperdynamic circulation, the cardiac compensatory reservegradually decreases, resulting in an impaired adaptiveresponse to acute circulatory stress, such as SBP. Develop-ment of SBP has been shown to be associated with pro-nounced portal and systemic hemodynamic derangements,as bacteremia has been found to decrease systemic vascularresistance17 and patients with low cardiac output have beenshown to be at high risk for development of HRS,18 a majorcomplication of SBP that further reduces survival. In patientswith ascites, impaired cardiac output has been found to beassociated with substantially increased mortality.19
We therefore hypothesized that the development of SBPcloses the window of opportunity for NSBB treatment, asmaintaining the circulatory reserve is crucial in critically illpatients with cirrhosis. The main adaptive mechanism tocirculatory stress is the increase in heart rate mediated by
b1 adrenoreceptors that are down-regulated and desensi-tized in patients with advanced cirrhosis.20 Additionalpharmaceutical blockade induces chronotropic incompe-tence,20 decreases cardiac output,21 and blood pressure andcan increase the rate of HRS development after an SBPepisode, resulting in worse survival.
The aim of this study was to assess the influence of NSBBtreatment on SBP incidence and the impact of SBP devel-opment on the effect of NSBB treatment on hospitalizationand transplant-free survival. In addition, rates of HRS andacute kidney injury (AKI) development after the first SBPdiagnosis were assessed in patients with NSBB treatmentand without.
Patients and MethodsStudy Design
A total of 607 consecutive patients with cirrhosis who un-derwent their first paracentesis at the Medical University ofVienna between 2006 and 2011, had bacterial cultures fromascites, and did not display any exclusion criteria, wereincluded in this retrospective study. Patients with other causesof ascites, such as severe cardiovascular disease, renal insuffi-ciency, extrahepatic malignancies, and noncirrhotic portalhypertension, were excluded from the study.
Assessed ParametersEpidemiological characteristics, etiology of cirrhosis, pres-
ence 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, in-formation on NSBB and rifaximin treatment, as well as systemichemodynamics, was obtained from patient medical history. Thefollowing laboratory parameters were assessed at the timeof the first paracenteses and the first diagnosis of SBP: plateletcount, albumin, bilirubin, international normalized ratio,creatinine, and ascitic fluid neutrophil count. Hepatic venousportal pressure gradient measurements were performed asdescribed previously.22 The model for end-stage liver disease23
and Child-Pugh score (CPS)24 were calculated based on labo-ratory parameters and patients’ medical history.
Paracenteses and Diagnosis of SpontaneousBacterial Peritonitis
Paracenteses were performed either in a diagnostic ortherapeutic setting. In accordance with national guidelines,5,25
albumin was administered in all large-volume paracenteses.SBP was diagnosed if the ascitic fluid neutrophil count was>250 mL without an evident intra-abdominal source of infec-tion or another explanation for an elevated ascitic fluidneutrophil count.5,26,27 Patients with SBP were receiving albu-min in addition to antibiotic treatment, as recommended bynational guidelines during the full study period.5,25
Diagnosis of Hepatorenal Syndrome and AcuteKidney Injury
The diagnosis of HRS was established if creatinine levelsincreased to >1.5 mg/dL or doubled to a level of >2.5 mg/dL inthe absence of evidence for shock or hypovolemia5,26,27 within
1682 Mandorfer et al Gastroenterology Vol. 146, No. 7
CLINICALLIVER
90 days after the first SBP diagnosis. Patients with a history ofchronic kidney disease other than HRS or evidence for paren-chymal kidney disease, such as proteinuria, microhematuria, orabnormal renal ultrasonography were excluded from thisanalysis.
AKI was diagnosed if a patient fulfilled the criteria of groupC of the modified AKI classification proposed by Fagundeset al,28 comprising both stages 2 and 3 of the Acute KidneyInjury Network criteria.29 AKI was defined as an increase ofserum 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 wasconsidered as AKI.
StatisticsStatistical analyses were conducted using IBM SPSS Statis-
tics 21 (SPSS Inc., Armonk, NY) and R.3.0.1 (R DevelopmentCore 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 pa-tients with the certain characteristic [%]). Student t test wasused for group comparisons of continuous variables whenapplicable. Otherwise, Mann-Whitney U test was applied. Groupcomparisons of categorical variables were performed usingeither Pearson’s c2 or Fisher’s exact test. The impact of NSBBtreatment on SBP incidence and transplant-free survival wasanalyzed using semi-parametric proportional hazard Coxmodels. The proportional hazard assumption was verified byinspecting the cumulative hazard plot for parallel curves.
Patients entered the SBP incidence model (Model 1) withtheir first paracentesis and were followed until their last par-acentesis. In liver transplant recipients and in case of death, thelast paracentesis before the liver transplantation or death wasthe end of follow-up. Patients who had only one paracentesis,as well as patients with SBP at their first paracentesis, wereexcluded from this analysis. NSBB treatment and CPS stagewere considered as covariates. In addition, SBP incidence ratesamong patients with NSBB treatment, and without, as well asthe SBP incidence rate ratio were calculated.
The rates of follow-up paracentesis and variceal bleedingafter the first paracentesis were calculated for NSBB and no-NSBB patients. Similarly, follow-up paracentesis and bleedingrates after the first development of SBP were calculated.
Patients entered the first transplant-free survival model(Model 2) with their first paracentesis and were followed until2013. Transplant-free survival time was defined as the time toliver transplantation, death, or end of follow-up. Patients whoreceived a liver transplantation were censored at the day ofsurgery. Besides NSBB treatment, the model included presenceof varices, CPS stage, and SBP status as covariates. The lattervariable was treated as a time-dependent covariate in thefollowing fashion: The first development of SBP in patientswithout SBP at the first paracentesis separated the follow-upinto 2 adjoining time intervals. In the first (right censored)interval the SBP status is negative, and it is positive in thesecond (left or interval censored) time period. To assess themodulating effect of SBP development on the impact of NSBBtreatment on transplant-free survival, the interaction termNSBB � SBP was incorporated into the model.
A second transplant-free survival model (Model 3)restricted to SBP patients was calculated with patients enteringthe model with the first development of SBP and NSBB treat-ment, presence of varices and CPS stage at the time of SBPdiagnosis as covariates. Kaplan-Meier curves are shown for allmodels.
P values <.05 were considered as statistically significant.No adjustment for multiplicity was performed.
EthicsThis study was conducted in accordance with the Declara-
tion of Helsinki and approved by the local ethics committee ofthe Medical University of Vienna (EK Nr. 1008/2011).
ResultsPatient Characteristics and SystemicHemodynamics at the First Paracentesis
Except for a higher proportion of female subjects (NSBB:34% vs no-NSSB: 27%; P ¼ .048) and patients with varices(NSBB: 90% vs no-NSBB: 62%; P < .001) in the NSBB group,no statistically significant differences in patient character-istics 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 sys-tolic arterial pressure (SAP) (NSBB: 114 � 18 vs no-NSBB:117 � 18 mm Hg; P ¼ .05) (Figure 1) were lower amongpatients with NSBB treatment than in patients without, nodifferences in diastolic arterial pressure (DAP) and meanarterial pressure (MAP) (Figure 1B) were observed. Theproportion of hemodynamically compromised patients, asdefinedby a SAP<100mmHg (NSBB: 18%vsno-NSBB: 15%;P¼ .391) (Figure 1C) or an MAP<82 mm Hg (NSBB: 39% vsno-NSBB: 38%; P ¼ .757) (Figure 1D), was comparable be-tween the NSBB treatment groups at the first paracentesis.
In the NSBB group, daily propranolol doses ranged be-tween 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 rangedbetween 6.25 mg and 25 mg (6.25 mg: 7%, 12.5 mg: 18%,25 mg: 3%).
Influence of Nonselective b Blocker Treatment onSpontaneous Bacterial Peritonitis Incidence
NSBB treatment (hazard ratio [HR] ¼ 0.699; 95% con-fidence interval [CI]: 0.427–1.146; P ¼ .156) and CPS stage(stage B: HR ¼ 1.211; 95% CI: 0.283–5.192; P ¼ .796 andstage C: HR ¼ 2.033; 95% CI: 0.485–8.519; P ¼ .332) werenot associated with SBP incidence in univariate analysis(Model 1, Supplementary Figure 1).
In multivariate analysis, neither NSBB treatment (HR ¼0.728; 95% CI: 0.422–1.198; P ¼ .211), nor CPS stage (stageB: HR ¼ 1.082; 95% CI: 0.25–4.686; P ¼ .916 and stage C:HR¼ 1.792; 95% CI: 0.422–7.611; P¼ .429) were associatedwith SBP development (Model 1, Supplementary Figure 1).
SBP incidence rates were comparable between patientswith (0.107 per person-year; 95% CI: 0.071–0.15) and
Table 1.Patient Characteristics at the First Paracentesis and at the First Development of Spontaneous Bacterial Peritonitis and Comparison of Patients With and WithoutNonselective b Blocker Treatment
Patient characteristics
At the first paracentesis At the first development of SBP
All (n ¼ 607) No NSBB (n ¼ 362) NSBB (n ¼ 245) P value All (n ¼ 182) No NSBB (n ¼ 96) NSBB (n ¼ 86) P value
Age, y, mean � SD 57.5 � 11.8 57.2 � 12.1 58 � 11.2 .438 57.7 � 11.1 58.1 � 10.5 57.3 � 11.8 .655Sex, n (%) .048 .073
Male 426 (70) 265 (73) 161 (66) 134 (74) 76 (79) 58 (67)Female 181 (30) 97 (27) 84 (34) 48 (26) 20 (21) 28 (33)
Etiology, n (%) .264 .07ALD 336 (55) 192 (53) 144 (59) 94 (52) 42 (44) 52 (60)Viral 113 (19) 70 (19) 43 (18) 43 (24) 25 (26) 18 (21)ALD and viral 49 (8) 27 (7) 22 (9) 14 (8) 11 (11) 3 (4)Other 109 (18) 73 (20) 36 (15) 31 (17) 18 (19) 13 (15)
HCC, n (%) 129 (21) 74 (20) 55 (22) .553 46 (25) 25 (26) 21 (24) .801Previous variceal bleeding, n (%) 98 (16) 54 (15) 44 (18) .318 37 (20) 17 (18) 20 (23) .353Varices, n (%) 443 (73) 223 (62) 220 (90) <.001 139 (76) 58 (60) 81 (94) <.001HVPG, mm Hg, mean � SD 18.7 � 6.5a 18.3 � 7.1 19 � 5.8 .398 19.3 � 6.6b 19.1 � 7.3 19.5 � 5.8 .806MELD, median (IQR)/mean � SD 17.5 (10.6) 17.8 (11) 17.2 (9.9) .422 20.8 � 9 20 � 8.7 21.6 � 9.3 .241CPS stage, n (%) .961 .107
A 22 (4) 13 (4) 9 (4) 6 (3) 3 (3) 3 (4)B 281 (46) 166 (46) 115 (47) 67 (37) 42 (44) 25 (29)C 304 (50) 183 (51) 121 (49) 109 (60) 51 (53) 58 (67)
Platelet count, 109/L, median (IQR) 117 (107) 117 (108) 120 (107) .912 118 (97) 119 (86) 118 (111) .36Albumin, g/L, mean � SD 27.2 � 5.7 27 � 5.4 27.6 � 6.1 .231 27.1 � 5.9 27.5 � 5.6 26.6 � 6.3 .315Bilirubin, mg/dL, median (IQR) 3.2 (6.02) 3.05 (6.98) 3.05 (5.98) .623 3.61 (5.8) 3 (5) 5.11 (9.32) .024INR, median (IQR) 1.38 (0.58) 1.41 (1.19) 1.41 (0.58) .262 1.41 (0.71) 1.41 (0.74) 1.39 (0.69) .534Creatinine, mg/dL, median (IQR) 1.14 (0.78) 1.14 (0.83) 1.14 (0.83) .288 1.39 (1.26) 1.4 (1.2) 1.34 (1.31) .876Rifaximin treatment, n (%)
Any 69 (11) 36 (10) 33 (14) .18 23 (13) 14 (15) 9 (11) .404Before first paracentesis 14 (2) 8 (2) 6 (2) .847 9 (5) 4 (4) 5 (6) .609After the first paracentesis 63 (10) 34 (9) 29 (12) .333 21 (12) 13 (14) 8 (9) .371
Heart rate, beats/min, mean � SD 81.2 � 16.2c 83.9 � 15.5 77.5 � 16.5 <.001 79.4 � 15.7d 82.9 � 15.7 76.2 � 15.2 .027SAP, mm Hg, mean � SD 115 � 18e 117 � 18 114 � 18 .05 109 � 18 113 � 18 104 � 17 <.001SAP <100 mm Hg, n (%) 98 (16)e 55 (15) 43 (18) .391 50 (27) 17 (18) 33 (38) .002DAP, mm Hg, mean � SD 70.5 � 12.5 71 � 12.5 69.8 � 12.4 .263 65.7 � 12.5 67.3 � 12.5 63.8 � 12.3 .06MAP, mm Hg, mean � SD 85.4 � 13.3 86.2 � 13.2 84.4 � 13.2 .112 80.1 � 13 82.6 � 12.5 77.2 � 13.1 .005MAP <82 mm Hg, n (%) 228 (38) 135 (38) 93 (39) .757 97 (53%) 42 (44) 55 (64) .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.aAvailable in 220 patients.bAvailable in 82 patients.cAvailable in 395 patients.dAvailable in 109 patients.eAvailable in 595 patients.
June2014
bBlockers
inSpontaneous
BacterialPeritonitis1683
CLINICAL LIVER
Figure 1. Effect of NSBBtreatment on (A) systolicarterial pressure (SAP), (B)mean arterial pressure(MAP) and the proportionof hemodynamically com-prised patients with (C)SAP <100 mmHg and (D)MAP <82 mmHg at thetime of the first para-centesis and the first SBPdevelopment.
1684 Mandorfer et al Gastroenterology Vol. 146, No. 7
CLINICALLIVER
without NSBB treatment (0.117 per person-year; 95% CI:0.084–0.156), with an incidence rate ratio of 0.915 (95% CI:0.565–1.483).
Influence of Nonselective b Blocker Treatment onSpontaneous Bacterial Peritonitis Development,Paracentesis and Variceal Bleeding Rates,Hospitalization as Well as Transplant-FreeSurvival
A total of 607 patients were followed for 660 person-years after their first paracentesis. There was a trend to-ward a lower paracentesis rate among patients with (2.45per person-year; 95% CI: 2.27–2.63), when compared withpatients without NSBB treatment (2.76 per person-year;95% CI: 2.6–2.94).
Thirty-three episodes of variceal bleeding were observedduring follow-up. Bleeding rates were comparable betweenpatients with (0.059 per person-year; 95% CI: 0.034–0.068)and without NSBB treatment (0.043 per person-year; 95%CI: 0.025–0.067). The proportions of patients receiving aliver transplant were 12% and 10% in the NSBB and no-NSBB group, respectively (P ¼ .46).
After their first paracentesis, patients with NSBB treat-ment were hospitalized for 26.7 days per person-year (95%CI: 26.1–27.3), and patients without NSBB treatment werehospitalized for 30.9 days per person-year (95% CI:30.4–31.5). Although the duration of elective hospitaliza-tions was similar (NSBB: 7.37 days per person-year; 95% CI:7.07–7.69 vs no-NSBB: 6.98 days per person-year; 95% CI:
6.72–7.26), the duration of nonelective hospitalization waslower among patients with (19.4 days per person-year; 95%CI: 18.9–19.9) when compared with patients without NSBBtreatment (23.9 days per person-year; 95% CI: 23.4–24.4).
NSBB treatment was associated with higher transplant-free survival (HR ¼ 0.771; 95% CI: 0.598–0.993; P ¼ .044)in multivariate analysis, when adjusting for the presence ofvarices (HR ¼ 0.983; 95% CI: 0.783–1.235; P ¼ .886), CPSstage (stage B: HR¼ 1.134; 95% CI: 0.615–2.09; P¼ .687 andstage C: HR ¼ 2.24; 95% CI: 1.222–4.107; P ¼ .009), andSBP development (HR ¼ 1.401; 95% CI: 1.056–1.857;P¼ .019) (Model 2, Table 2). In addition, the interaction termNSBB � SBP (HR ¼ 1.917; 95% CI: 1.256–2.924; P ¼ .003)was statistically significantly associated with transplant-freesurvival. This suggests that the development of SBP modu-lates the effect of NSBB treatment on transplant-free survivalwarranting additional subgroup analysis.
Impact of Spontaneous Bacterial PeritonitisDevelopment on the Effect of Nonselective b
Blocker Treatment on Transplant-Free SurvivalUp to the development of SBP, NSBB treatment had a
beneficial effect on transplant-free survival (HR ¼ 0.75;95% CI: 0.581–0.968; P ¼ .027) when adjusting for thepresence of varices and CPS stage (Model 2, Table 2,Figure 2A) and was associated with a reduction in mortalityrisk of 25%. In contrast, once a patient developed SBP, NSBBtreatment was associated with lower transplant-free sur-vival (HR ¼ 1.58; 95% CI: 1.098–2.274; P ¼ .014) when
Table 2. Impact of Spontaneous Bacterial Peritonitis Development on the Effect of Nonselective b Blocker Treatment onTransplant-Free Survival (Model 2): Univariate and Multivariate Analyses Including All patients and MultivariateAnalysis According to Spontaneous Bacterial Peritonitis Status
Patient characteristics HR
95% CI
P value HR
95% CI
P valueLower Upper Lower Upper
Univariate (n ¼ 607) Multivariate (n ¼ 607)Varices, yes 1.047 0.841 1.303 .682 0.983 0.783 1.235 .886CPS stage B 1.076 0.584 1.983 .814 1.134 0.615 2.09 .687CPS stage C 2.203 1.203 4.035 .011 2.24 1.222 4.107 .009SBP, yes 1.893 1.538 2.328 <.001 1.401 1.056 1.857 .019NSBB, yes 0.945 0.775 1.152 .578 0.771 0.598 0.993 .044NSBB � SBP 1.917 1.256 2.924 .003
Multivariate, no-SBP (n ¼ 493) Multivariate, SBP (n ¼ 182)Varices, yes 1.112 0.846 1.46 .447 0.703 0.462 1.068 .098CPS stage B 0.934 0.455 1.916 .852 1.71 0.527 5.552 .372CPS stage C 1.778 0.872 3.625 .113 3.717 1.157 11.946 .028NSBB, yes 0.75 0.581 0.968 .027 1.58 1.098 2.274 .014
June 2014 b Blockers in Spontaneous Bacterial Peritonitis 1685
CLINICAL
LIVE
R
adjusting for the same variables (Model 2, Table 2,Figure 2B). NSBB treatment was associated with an increaseof 58% in mortality risk once a patient developed SBP.
Patient Characteristics and SystemicHemodynamics at First Development ofSpontaneous Bacterial Peritonitis
A total of 182 patients developed at least one SBPepisode. Comparing NSBB and no-NSBB patient character-istics revealed no statistically significant differences exceptfor a higher prevalence of varices (NSBB, 94% vs no-NSBB,60%; P < .001) and higher median bilirubin levels (NSBB:5.11 mg/dL [IQR, 9.32 mg/dL] vs no-NSBB: 3 mg/dL [IQR, 5mg/dL]; P ¼ .024) in the NSBB group. In addition, there wasa trend toward a higher proportion of female patients(NSBB: 33% vs no-NSBB: 21%; P ¼ .073), as well as patientswith alcoholic etiology (NSBB: 60% vs no-NSBB: 44%;P ¼ .07) and CPS stage C (NSBB: 67% vs no-NSBB: 53%;P ¼ .107) among patients with NSBB treatment.
Patients with NSBB treatment had a lower mean heartrate (NSBB: 76.2 � 15.2 beats/min vs no-NSBB: 82.9 � 15.7
Figure 2. Impact of NSBBtreatment on transplant-free survival according toSBP status. Patientswithout SBP at first diag-nosis started in (A) (no-SBP), while Patients withSBP at the first para-centesis started in (B)(SBP). At the first SBPdiagnosis, patients werecensored in (A) (no-SBP),entered (B) (SBP), andremained in this group.
beats/min; P ¼ .027), SAP (NSBB: 104 � 17 mm Hg vsno-NSBB: 113 � 18 mm Hg; P < .001) (Figure 1A), and MAP(NSBB: 77.2 � 13.1 mm Hg vs no-NSBB: 82.6 � 12.5 mm Hg;P ¼ .005) (Figure 1B). In addition, DAP tended to be loweramong patients with NSBB treatment (NSBB: 63.8� 12.3mmHg vs no-NSBB: 67.3 � 12.5 mm Hg; P ¼ .06). Similarly, theproportion of patients with SAP<100mmHg (NSBB: 38% vsno-NSBB:18%; P ¼ .002) (Figure 1C) or MAP <82 mm Hg(NSBB: 64% vs no-NSBB:44%; P ¼ .006) (Figure 1D) wassubstantially higher among patients with NSBB treatment.
Impact of Nonselective b Blocker Treatment onParacentesis and Variceal Bleeding Rates,Hospitalization as Well as Transplant-FreeSurvival After the First Spontaneous BacterialPeritonitis Diagnosis
After SBP development, 182 patients were followed for atotal of 147 person-years. The paracentesis rate was com-parable between no-NSBB (3.67 per person-year; 95% CI:3.29–4.08) and NSBB patients (3.36 per person-year; 95%CI: 2.9–3.84).
Figure 3. Influence of NSBB treatment on transplant-freesurvival after the first SBP diagnosis.
1686 Mandorfer et al Gastroenterology Vol. 146, No. 7
CLINICALLIVER
Thirteen episodes of variceal bleeding were observedafter the first SBP diagnosis. Bleeding rates were similaramong patients with (0.086 per person-year; 95% CI:0.028–0.175) and without NSBB treatment (0.09 per person-year; 95% CI: 0.039–0.162). In addition, the proportion ofpatients receiving a liver transplantation did not vary sta-tistically significantly throughout the NSBB treatment groups(no-NSBB: 9% vs NSBB: 13%; P ¼ .462).
After the first diagnosis of SBP, the duration of hospi-talization was higher in the NSBB group (NSBB: 33.4 daysper person-year; 95% CI: 31.9–34.9 vs no-NSBB: 28.8 daysper person-year; 95% CI: 27.6–29.9). The duration of elec-tive hospitalizations was comparable between both treat-ment groups (NSBB: 3.6 days per person-year; 95% CI:3.1–4.1 vs no-NSBB: 3.9 days per person-year; 95% CI:3.5–4.3). In contrast, a higher duration of nonelective hos-pitalizations was observed in the NSBB group (NSBB: 29.6days per person-year; 95% CI: 28.2–31 vs no-NSBB: 23.7days per person-year; 95% CI: 22.7–24.8).
In multivariate analysis, NSBB treatment was associatedwith impaired transplant-free survival after SBP diagnosis(HR¼ 1.644; 95% CI: 1.145–2.361; P¼ .007), after adjustingfor CPS stage (stage B: HR ¼ 1.642; 95% CI: 0.507–5.314;P ¼ .408; stage C: HR ¼ 3.318; 95% CI: 1.036–10.627;P ¼ .043) and the presence of varices (HR ¼ 0.695; 95% CI:0.457-1.057; P ¼ .089) at the time of SBP development(Model 3, Table 3, Figure 3). Rates of transplant-free survivalafter SBP development are shown in Table 3.
Influence of Nonselective b Blocker Treatment onHepatorenal Syndrome and Acute Kidney InjuryDevelopment After the First SpontaneousBacterial Peritonitis Diagnosis
A total of 11 patients were excluded from the analysisof HRS development because of evidence for shock, hypo-volemia, or chronic kidney disease other than HRS. Inaddition, 6 patients without HRS, liver transplantation, or
Table 3. Influence of Nonselective b Blocker Treatment on TranPeritonitis Diagnosis (Model 3) and Survival Rates Afte
HR
95% CI
Lower Upper
Patient characteristics Univariate (n ¼ 182)Varices, yes 0.97 0.657 1.432CPS stage B 1.478 0.459 4.757CPS stage C 2.954 0.932 9.365NSBB, yes 1.539 1.095 2.164
no-NSBB (n ¼ 96)
Estimate 95% CI
Time point Lower Upper1 month 0.713 0.627 0.816 month 0.495 0.403 0.60812 month 0.358 0.271 0.472
death were excluded because of a follow-up period of <90days after the first SBP diagnosis. Among the remaining165 patients, HRS was observed in 18% (29 of 165) ofpatients within 90 days after the SBP diagnosis. The rate ofHRS development was significantly higher in patients with(24% [20 of 83]) when compared with patients withoutNSBB treatment (11% [9 of 82]; P ¼ .027) (Figure 4).Eighty percent (16 of 20) of patients with NSBB treatmentdeveloping HRS died within 90 days after the HRSdiagnosis.
Six patients without HRS, liver transplantation, or deathwere excluded from the analysis of AKI development becauseof a follow-up period of <90 days after the first SBP diag-nosis. Grade C AKI was observed in 14% (24 of 176) of thepatients included in this analysis. Patients with NSBB treat-ment (20% [17 of 86]) showed a higher rate of grade C AKIdevelopment when compared with patients without NSBB
splant-Free Survival After the First Spontaneous Bacterialr Spontaneous Bacterial Peritonitis Development
P value HR
95% CI
P valueLower Upper
Multivariate (n ¼ 182).88 0.695 0.457 1.057 .089.512 1.642 0.507 5.314 .408.066 3.318 1.036 10.627 .043.013 1.644 1.145 2.361 .007
NSBB (n ¼ 86)
Estimate 95% CI
Lower Upper0.566 0.47 0.6820.279 0.198 0.3930.227 0.152 0.338
Figure 4. Influence of NSBB treatment on HRS and grade CAKI development within 90 days after the first SBP diagnosis.
June 2014 b Blockers in Spontaneous Bacterial Peritonitis 1687
CLINICAL
LIVE
R
treatment (8% [7 of 90]; P ¼ .021) (Figure 4). Eighty-eightpercent (15 of 17) of patients with NSBB treatment did notsurvive the 90-day period after grade C AKI development.
DiscussionA recent landmark study by Serste et al12 demonstrated
reduced survival in patients with refractory ascites whoweretreated with NSBBs and has initiated a lively debate amonggastroenterologists on the appropriate use of NSBBs in pa-tients with advanced cirrhosis. In summary, the editorial30
and various other responses to this study brought up con-cerns about the unequal distribution of esophageal varicesbetween the NSBB and no-NSBB group. In addition, thetrends toward a higher prevalence of CPS stage C and HCC inpatientswith NSBB treatment31,32 and the lack of consecutivepatient enrollment33 were highlighted as factors significantlylimiting the conclusions drawnwithin the study. We aimed toaddress these limitations in our study by adjusting for thepresence of varices in all analyses of transplant-free survival.In addition, all analyses were adjusted for CPS stage, as at thefirst SBP diagnosis, there was a trend toward a lower prev-alence of CPS stage B and a higher prevalence of CPS stage Camong patients with NSBB treatment. Except for a higherproportion of females in the NSBB group at the first para-centesis and higher bilirubin levels in the NSBB group at thefirst SBP diagnosis, no other statistically significant differ-ences in patient characteristics were observed. In order toavoid multicollinearity, bilirubin was not considered as acovariate, as it is a component of the CPS. As our studypopulation comprises consecutive patients meeting the in-clusion and exclusion criteria of this study, our results arebased on a large, representative cohort treated at a singleacademic center.
Another response by Senzolo et al11 addressed a po-tential beneficial effect of NSBB treatment on SBP devel-opment, highlighting a positive meta-analysis by the sameauthor.11 In our study, no statistically significant effect ofNSBB treatment on SBP incidence was observed. As pa-tients who had only one paracentesis and patients withSBP at their first paracentesis were excluded and the
follow-up duration was limited to the last paracentesis forstatistical reasons, a lack of statistical power of Model 1 todetect a clinically significant effect of NSBB treatment onSBP development cannot be excluded. However, theselimitations do not apply to the analyses of SBP incidencerates, which were comparable between patients with andwithout NSBB treatment. The proportion of CPS stage Cpatients in our study was substantially higher whencompared with previous studies designed for the assess-ment of the effect of NSBB treatment on varicealbleeding,13 and all patients included in our study hadpreviously developed ascites. As NSBB treatment has beenshown to reduce the risk of developing ascites in patientswith compensated cirrhosis,9 the overall effect of NSBBtreatment initiated before the first decompensation cannotbe assessed in our study. Our results provide a rationalefor additional meta-analyses assessing the effect of NSBBtreatment on SBP development in patients with moreadvanced stages of cirrhosis.
Nearly half of the patients included in the study bySerste et al12 received a comparatively high daily propran-olol dose of 160 mg. In a consecutive crossover study,propranolol treatment was found to be associated with ahigher risk for paracentesis-induced circulatory dysfunctionin patients with refractory ascites and cirrhosis.34 Seven of10 patients included in this study received 160 mg pro-pranolol. Because high NSBB doses have profound effects onsystemic circulation and might not be well tolerated byhemodynamically compromised patients, such as patientswith advanced cirrhosis, Robins et al35 suggested that theuse of lower NSBB doses is safe among patients with re-fractory ascites. In contrast, substantially lower NSBB doseswere administered in our study. However, carvedilol mightnot only be more efficient in decreasing portal pressure byadditional a1 adrenoreceptor inhibition,36,37 it might alsolead to a more pronounced decrease in MAP.36
The efficacy of NSBB treatment for the prevention ofvariceal bleeding has primarily been studied in patientswithout ascites, and meta-analysis suggests that the bene-ficial effect in the prevention of variceal bleeding might belower among patients with ascites.14 The characteristics orour study population might explain the low rates of varicealbleeding during follow-up observed in our study. There wasa significant proportion of patients without varices in theno-NSBB group, and the proportion of patients with previ-ous variceal bleeding included in this study was low. Whencomparing bleeding rates between patients with andwithout NSBB treatment, the unequal distribution of variceshas to be considered.
Patients with more advanced liver disease, such as pa-tients with cirrhosis and refractory ascites or bacterial in-fections have yet not been specifically addressed in clinicaltrials investigating the effectiveness of NSBB treatment.4
Our study assessed the effect of NSBB treatment in acohort of patients with cirrhosis undergoing their firstparacentesis and considered the incidence of SBP as a time-dependent covariate. Thus, our study comprised an internalcontrol group of patients with ascites who yet have notdeveloped SBP. This study design is in accordance with the
1688 Mandorfer et al Gastroenterology Vol. 146, No. 7
CLINICALLIVER
common understanding of cirrhosis as a dynamic processand facilitates the design of prospective studies to validateour observations. Although up to the development of SBP,NSBB treatment had a beneficial effect on transplant-freesurvival and was associated with a reduction in mortalityrisk of 25%, transplant-free survival was worse in SBP pa-tients receiving NSBB treatment with an increase in mor-tality risk of 58%. We also confirmed this observation in anadditional analysis with patients entering the model withthe first development of SBP. These observations suggestthat NSBB treatment should be discontinued at the firstdevelopment of SBP, raising the question of whether topermanently discontinue or to restart NSBB treatment afterresolving the SBP episode. Based on the slope of the Kaplan-Meier curves, the higher hazard in the NSBB patients waslimited to the first 6 months after SBP development, withcurves converging toward the end of follow-up. However,this observational data must be interpreted with caution, assignificant bias can arise from the high rate of early mor-tality observed in the NSBB group, leading to the selection ofpatients with better prognosis and therefore lower hazardor with improvement of hepatic function during follow-up.In addition, the development of additional SBP episodeswas not considered in this model. Arvaniti et al havedemonstrated that following the occurrence of bacterialinfections, 30% of patients with cirrhosis patients die within1 month, and another 30% die within 1 year.15 The devel-opment of bacterial infections might define a distinct groupof critically ill patients with cirrhosis with a highly restrictedprognosis in which maintaining the circulatory reserve iscrucial not only during the acute phase of infection but alsolater on. This hypothesis is supported by recent data onmortality in patients with cirrhosis admitted to the intensivecare unit for severe sepsis or septic shock in which NSBBtreatment was discontinued during the period of hemody-namic instability. Although intensive care unit mortality wascomparable, higher mortality rates at 3 and 6 months wereobserved among patients discharged from the intensive careunit with re-established NSBB treatment.38
Similar to our observations on transplant-free survival,NSBB treatment had a detrimental effect on nonelective andoverall hospitalization after SBP development. NSBB treat-ment might not only lead to increased mortality, but also toincreased morbidity and burden of disease in this subgroupof patients.
NSBB treatment had only a moderate effect on systemichemodynamics at the time of the first paracentesis andtherefore did not increase the proportion of hemodynami-cally compromised patients. In contrast, at the time of thefirst SBP diagnosis, there was a substantially increased pro-portion of hemodynamically compromised patients in theNSBB group, defined as SAP <100 mm Hg or MAP <82 mmHg. MAP <82 mm Hg39 or impaired cardiac output19 havebeen found to be associated with substantially increasedmortality in patients with cirrhosis or both cirrhosis andascites, respectively. Also, patients with SBP and low cardiacoutput have been shown to be at high risk for HRS,18 andNSBB treatment has previously been found to further dete-riorate cardiac output in patients with cirrhosis.21 These
results provide a pathophysiologic explanation for thedetrimental effect of NSBB treatment after SBP development,compromising systemic hemodynamics and predisposing forHRS and AKI development, as observed in our study popu-lation. Our study supports previous hypotheses with obser-vational data.40,41 Becausewe only considered group C AKI inour analyses, there was a substantial overlap of AKI and HRS.However, the proportion of patients who fulfilled the criteriafor group C AKI was numerically smaller than the proportionof patients with HRS, which might be explained by therestrictive time criterion included in the diagnostic criteria ofAKI and the higher creatinine increase threshold for thediagnosis of group C AKI.
The main limitations of our study arise from the retro-spective design. As patients have not been prospectivelyfollowed, drug adherence and intermittent interruptions inNSBB treatment cannot be ruled out entirely. However, thebeneficial effect of transplant-free survival in the no-SBPgroup can serve as an internal validation for our approach.
This is the first study to provide observational evidencefor the detrimental effect of NSBB treatment after SBPdevelopment. It supports the pathophysiologically compel-ling window hypothesis by Krag et al,13 defining SBP as aclinical event that closes the therapeutic window for NSBBtreatment. However, whether the therapeutic window forNSBB treatment reopens after resolving the SBP episoderemains unclear. Prospective studies are highly encouragedto investigate the appropriate, stage-dependent use of thiscornerstone in the treatment of portal hypertension.
Supplementary MaterialNote: To access the supplementary material accompanyingthis article, visit the online version of Gastroenterology atwww.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. Theburden of liver disease in Europe: a review of availableepidemiological data. J Hepatol 2013;58:593–608.
2. Asrani SK, Larson JJ, Yawn B, et al. Underestimation ofliver-related mortality in the United States. Gastroenter-ology 2013;145:375–382, e1–2.
3. D’Amico G, Garcia-Tsao G, Pagliaro L. Natural historyand prognostic indicators of survival in cirrhosis: a sys-tematic review of 118 studies. J Hepatol 2006;44:217–231.
4. Ge PS, Runyon BA. The changing role of beta-blockertherapy in patients with cirrhosis. J Hepatol 2014;60:643–653.
5. Peck-Radosavljevic M, Angermayr B, Datz C, et al.Austrian consensus on the definition and treatment ofportal hypertension and its complications (Billroth II).Wien Klin Wochenschr 2013;125:200–219.
6. de Franchis R. Revising consensus in portal hyperten-sion: report of the Baveno V consensus workshop onmethodology of diagnosis and therapy in portal hyper-tension. J Hepatol 2010;53:762–768.
June 2014 b Blockers in Spontaneous Bacterial Peritonitis 1689
CLINICAL
LIVE
R
7. Garcia-Tsao G, Sanyal AJ, Grace ND, et al. Preventionand management of gastroesophageal varices and vari-ceal hemorrhage in cirrhosis. Hepatology 2007;46:922–938.
8. Lebrec D, Poynard T, Hillon P, et al. Propranolol forprevention of recurrent gastrointestinal bleeding in pa-tients with cirrhosis: a controlled study. N Engl J Med1981;305:1371–1374.
9. Hernandez-Gea V, Aracil C, Colomo A, et al. Develop-ment of ascites in compensated cirrhosis with severeportal hypertension treated with beta-blockers. Am JGastroenterol 2012;107:418–427.
10. Reiberger T, Ferlitsch A, Payer BA, et al. Non-selectivebetablocker therapy decreases intestinal permeabilityand serum levels of LBP and IL-6 in patients withcirrhosis. J Hepatol 2013;58:911–921.
11. Senzolo M, Cholongitas E, Burra P, et al. beta-Blockersprotect against spontaneous bacterial peritonitis incirrhotic patients: a meta-analysis. Liver Int 2009;29:1189–1193.
12. Serste T, Melot C, Francoz C, et al. Deleterious effects ofbeta-blockers on survival in patients with cirrhosis andrefractory ascites. Hepatology 2010;52:1017–1022.
13. Krag A, Wiest R, Albillos A, et al. The window hypothesis:haemodynamic and non-haemodynamic effects of beta-blockers improve survival of patients with cirrhosis dur-ing a window in the disease. Gut 2012;61:967–969.
14. D’Amico G, Pagliaro L, Bosch J. Pharmacological treat-ment of portal hypertension: an evidence-basedapproach. Semin Liver Dis 1999;19:475–505.
15. Arvaniti V, D’Amico G, Fede G, et al. Infections in patientswith cirrhosis increase mortality four-fold and should beused in determining prognosis. Gastroenterology 2010;139:1246–1256, 1256 e1–5.
16. Ferlitsch A, Pleiner J, Mittermayer F, et al. Vasocon-strictor hyporeactivity can be reversed by antioxidants inpatients with advanced alcoholic cirrhosis of the liver andascites. Crit Care Med 2005;33:2028–2033.
17. Bellot P, Garcia-Pagan JC, Frances R, et al. Bacterial DNAtranslocation is associated with systemic circulatory ab-normalities and intrahepatic endothelial dysfunction inpatients with cirrhosis. Hepatology 2010;52:2044–2052.
18. Ruiz-del-Arbol L, Urman J, Fernandez J, et al. Systemic,renal, and hepatic hemodynamic derangement incirrhotic patients with spontaneous bacterial peritonitis.Hepatology 2003;38:1210–1218.
19. Krag A, Bendtsen F, Henriksen JH, et al. Low cardiacoutput predicts development of hepatorenal syndromeand survival in patients with cirrhosis and ascites. Gut2010;59:105–110.
20. Krag A, Moller S, Burroughs AK, et al. Betablockersinduce cardiac chronotropic incompetence. J Hepatol2012;56:298–299.
21. Bendtsen F, Henriksen JH, Sorensen TI. Propranolol andhaemodynamic response in cirrhosis. J Hepatol 1991;13:144–148.
22. Ferlitsch M, Reiberger T, Hoke M, et al. von Willebrandfactor as new noninvasive predictor of portal hyperten-sion, decompensation and mortality in patients with livercirrhosis. Hepatology 2012;56:1439–1447.
23. Kamath PS, Kim WR. The model for end-stage liverdisease (MELD). Hepatology 2007;45:797–805.
24. Child C, Turcotte J. Surgery and portal hypertension. In:Child C, ed. The Liver and Portal Hypertension. Phila-delphia: Saunders, 1964:50–64.
25. Peck-Radosavljevic M, Trauner M, Schreiber F. Austrianconsensus on the definition and treatment of portal hy-pertension and its complications. Endoscopy 2005;37:667–673.
26. European Association for the Study of the Liver. EASLclinical practice guidelines on the management of asci-tes, spontaneous bacterial peritonitis, and hepatorenalsyndrome in cirrhosis. J Hepatol 2010;53:397–417.
27. Runyon BA. Introduction to the revised American Asso-ciation for the Study of Liver Diseases Practice Guidelinemanagement of adult patients with ascites due tocirrhosis 2012. Hepatology 2013;57:1651–1653.
28. Fagundes C, Barreto R, Guevara M, et al. A modifiedacute kidney injury classification for diagnosis and riskstratification of impairment of kidney function in cirrhosis.J Hepatol 2013;59:474–481.
29. Mehta RL, Kellum JA, Shah SV, et al. Acute Kidney InjuryNetwork: report of an initiative to improve outcomes inacute kidney injury. Crit Care 2007;11:R31.
30. Wong F, Salerno F. Beta-blockers in cirrhosis: friend andfoe? Hepatology 2010;52:811–813.
31. Kurt M. Deleterious effects of beta-blockers on survival inpatients with cirrhosis and refractory ascites. Hepatology2011;53:1411–1412.
32. Lo GH. Effects of beta-blockers on survival for patientswith cirrhosis and refractory ascites. Hepatology 2011;53:1068–1069.
33. AgustinA, Javier Z.Beta-blockers inpatientswith cirrhosisand refractory ascites. Hepatology 2011;53:1783.
34. Serste T, Francoz C, Durand F, et al. Beta-blockerscause paracentesis-induced circulatory dysfunction inpatients with cirrhosis and refractory ascites: a cross-over study. J Hepatol 2011;55:794–799.
35. Robins A, Bowden A, Watson W, et al. Propranolol atmodest doses does not impair survival in patients withcirrhosis and refractory ascites. Hepatology in press.
36. Banares R, Moitinho E, Matilla A, et al. Randomizedcomparison of long-term carvedilol and propranololadministration in the treatment of portal hypertension incirrhosis. Hepatology 2002;36:1367–1373.
37. Reiberger T, Ulbrich G, Ferlitsch A, et al. Carvedilol forprimary prophylaxis of variceal bleeding in cirrhotic pa-tients with haemodynamic non-response to propranolol.Gut 2013;62:1634–1641.
38. Galbois A, Das V, Thabut D, et al. Beta-blockers have noeffect on outcomes in patients with cirrhosis and severeinfections. Hepatology 2011;53:1412–1413.
39. Llach J, Gines P, Arroyo V, et al. Prognostic value ofarterial pressure, endogenous vasoactive systems, andrenal function in cirrhotic patients admitted to the hos-pital for the treatment of ascites. Gastroenterology 1988;94:482–487.
40. Efe C, Purnak T, Ozaslan E. The deleterious effects ofpropranolol on patients with cirrhosis. Hepatology 2011;53:371–372.
1690 Mandorfer et al Gastroenterology Vol. 146, No. 7
CLINICALLIVER
41. Krag A, Bendtsen F, Moller S. Does cardiac dysfunc-tion explain deleterious effects of beta-blockers incirrhosis and refractory ascites? Hepatology 2011;53:370–371.
Author names in bold designate shared co-first authors.
Received December 23, 2013. Accepted March 7, 2014.
Reprint requestsAddress requests for reprints to: Thomas Reiberger, MD, Division ofGastroenterology and Hepatology, Department of Internal Medicine III,Medical University of Vienna, Waehringer Guertel 18-20, 1090, Vienna,Austria. e-mail: [email protected]; fax: (þ43) 1 40400 4735.
Conflicts of interestThe authors disclose no conflicts.
Supplementary Figure 1. Influence of NSBB treatment onSBP incidence.
June 2014 b Blockers in Spontaneous Bacterial Peritonitis 1690.e1
