recent advances in clinical practice spontaneous bacterial peritonitis: recent ... · 2014. 11....

Spontaneous bacterial peritonitis:recent guidelines and beyondR Wiest,1 A Krag,2 A Gerbes3

INTRODUCTIONSpontaneous bacterial peritonitis (SBP) is the mostfrequent and life-threatening infection in patientswith liver cirrhosis requiring prompt recognitionand treatment. It is defined by the presence of >250polymorphonuclear cells (PMN)/mm3 in ascites inthe absence of an intra-abdominal source of infec-tion or malignancy. In this review we discuss thecurrent opinions reflected by recent guidelines(American Association for the Study of LiverDiseases, European Association for the Study of theLiver, Deutsche Gesellschaft für Verdauungs- undStoffwechselkrankheiten),1e4 with particular focuson controversial issues as well as open questionsthat need to be addressed in the future. First,diagnostic criteria and tools available for rapid andaccurate diagnosis are reviewed. Second, sinceprophylaxis is of crucial relevance when trying toimprove survival, we discuss who should betreated, when, how and for how long to preventepisodes of SBP. Identification of risk factors andindividualisation of timing and selection ofprophylactic measures are the key to successwithout major development of resistant bacteria.Finally, effective therapy is essential since treat-ment failure is associated with poor outcome. Sincethe emergence and spread of drug-resistant bacteriahas accelerated, criteria for the choice of antibioticregimen in the individual patient are pivotal foroptimising therapy.

EPIDEMIOLOGY AND PROGNOSIS OF SBPSBP is the most frequent bacterial infection incirrhosis, accounting for 10e30% of all reportedbacterial infections in hospitalised patients.5e7 Inoutpatients without symptoms the prevalence islow (3.5%8 or lower9 10), but the prevalenceincreases in the nosocomial setting, ranging from8% to 36%.11 12 Bacterascites, defined as positiveculture results but no increase in the PMN count inthe ascitic fluid, occurs with a prevalence of 2e3%in outpatients8e10 and in up to 11% in hospitalisedpatients.11 13 In-hospital mortality for the firstepisode of SBP ranges from 10% to 50%, dependingon various risk factors.7 14e18 One-year mortalityafter a first episode of SBP has been reported to be31% and 93%.8 17 19e21 In fact, the occurrence ofSBP or other severe bacterial infections markedlyworsens the prognosis in patients with cirrhosisand it has been proposed that a new prognosticstage of cirrhosis not reflected in current stagingsystems should be defined, the so-called ‘criticallyill cirrhotic’.22 Patients at this late stage have to beevaluated for the possibility of liver trans-plantation. Predictive factors reported for a poor

prognosis in various cohorts of patients with SBPare summarised in figure 1 and include age,16 20

Child score,18 20 23 intensive care,16 18 nosocomialorigin,18 24 hepatic encephalopathy,25 elevatedserum creatinine and bilirubin,26 lack of infectionresolution/need to escalate treatment and culturepositivity27e29 as well as the presence of bacter-aemia30 and CARD15/NOD2 variants as a geneticrisk factor.31 It is important to stress in this contextthat the only factors that are modifiable in thisscenario are timely diagnosis and effective first-linetreatment.

Bacterial translocation (BT) and pathophysiologyBacterial translocation (BT) is the most commoncause of SBP.32 33 However, particularly in nosoco-mial SBP, other sources such as transient bacter-aemia due to invasive procedures can lead to SBP.Limited BT to mesenteric lymph nodes (MLN) isa physiological phenomenon, whereas any increasein the rate and severity of BT may be deleteriousfor the patient and thus should be termed ‘patho-logical BT’. Only a few intestinal bacteria areable to translocate into MLN, including Escherichiacoli, Klebsiella pneumoniae and other Enter-obacteriaceae.34 Interestingly, these species mostfrequently cause SBP, and DNA sequencing studiesreveal genotypic identity of bacteria in MLN andascites in the vast majority of cases.35 36 Thissuggests that pathological BT is the underlyingcause and source of SBP in cirrhosis and supportsthe view that the route of pathological BT leadingto SBP is largely lymphatic. Three factors have beenimplicated in the development of pathological BTin liver cirrhosis32: (1) alterations in gut microbiota;(2) increased intestinal permeability; and (3)impaired immunity.

MicrobiotaLiver cirrhosis is associated with distinct changes infaecal microbial composition37 38 including anincreased prevalence of potentially pathogenicbacteria such as Enterobacteriaceae. Moreover,small intestinal bacterial overgrowth (SIBO),defined as >105 colony forming units/ml jejunalaspirate and/or colonic-type species, is frequentlypresent in advanced stages of liver cirrhosis andhas been linked with pathological BT, SBP andendotoxinaemia.39e41 In cirrhosis, factorspromoting these changes may include deficienciesin paneth cell defensins,41a reduced intestinalmotility, decreased pancreatobiliary secretions andportal-hypertensive enteropathy. In experimentalcirrhosis, in the absence of SIBO, BT occurs rarely(0e11%) and at rates comparable to healthy

1Department for visceral surgeryand medicine, UniversityHospital Bern, Switzerland2Department ofGastroenterology, HvidovreUniversity Hospital,Copenhagen, Denmark3Klinikum of the University ofMunich, Munich, Germany

Correspondence toProfessor Dr Reiner Wiest,Department for visceral surgeryand medicine, UniversityHospital Bern, 3010 Bern,Switzerland;[email protected]

Published Online First6 December 2011

Gut 2012;61:297e310. doi:10.1136/gutjnl-2011-300779 297

Recent advances in clinical practice

group.bmj.com on November 26, 2014 - Published by http://gut.bmj.com/Downloaded from group.bmj.com on November 26, 2014 - Published by http://gut.bmj.com/Downloaded from group.bmj.com on November 26, 2014 - Published by http://gut.bmj.com/Downloaded from

http://gut.bmj.com/http://group.bmj.comhttp://gut.bmj.com/http://group.bmj.comhttp://gut.bmj.com/http://group.bmj.com

conditions. However, BT does not occur in up tohalf of the animals with SIBO and, thus, SIBO isnecessary but not sufficient for BT to occur.

Intestinal permeabilityCirrhosis is associated with structural and func-tional alterations in the intestinal mucosa thatincrease permeability to bacteria and bacterialproducts. In particular, changes in enterocytemitochondrial function and increased oxidativestress of the intestinal mucosa have beenidentified.42 43

Host defenceFor translocation to become clinically significantdthat is, for it to lead to SBP or bacteraemiadafailure of local and systemic immune defencesappears to be the most important prerequisite (seebelow).

Local ascitic-peritoneal host defence in peritonitisThe peritoneal cavity probably has the most severelack of host defence compared with othercompartments in decompensated cirrhosis. In fact,ascites per se may be considered a risk factor for thedevelopment of peritonitis. In healthy conditions,peritoneal host defence mechanisms are very effi-cient and intraperitoneal injection of variousnumbers of single organisms does not cause peri-tonitis unless adjuvant substances or ascites arepresent.44 In cirrhosis, deficiencies in local defencemechanisms against bacteria, including dysfunc-tion of cellular and humoral immunity, limitperitoneal bacterial clearance.Since the absolute number of PMN per mm3

ascitic fluid defines SBP, the mechanisms ofchemotaxis mediating PMN influx into the perito-neal cavity are important. The degree of PMN

migration and accumulation in the peritonealcavity combating invading bacteria depends ona number of factors. Resident macrophages are thefirst to phagocytose bacteria, they further help toattract PMN by release of chemotactic factors andactivate complement. For instance, monocytechemotactic protein 1 is one of the most potentchemokines, and a functional polymorphism hasbeen proposed as a risk factor for SBP in alcoholiccirrhosis.45 A chemotactic gradient is necessary toachieve appropriate neutrophil recruitment into theperitoneal cavity. In fact, PMN chemoattractantssuch as zymosan are very effective in preventingthe death of animals with E coli-induced peritonitiswhen administered locally but not systemically.46

Unfortunately, little is known about the influx,efflux and kinetics of neutrophils in ascitic fluid incirrhosis and its dependency on type, extent andduration of bacterial stimulus as well as hostfactors.Besides influx of PMN, bacterial clearance is

determined by the overall killing capacity which isdependent on opsonisation, burst activity andinflammatory response. A marked reduction inopsonic and bactericidial activity is well-known incirrhosis. In particular, low C3 levels in cirrhoticascites correlate strongly with opsonic activity47

and have been shown to predispose to SBP.48

However, the total protein content aslo mirrorsopsonic activity and has been shown to be predic-tive of the development of SBP.49 At a protein levelof >1.5 g/dl ascitic fluid, the incidence rates of SBPhave been consistently reported to be lower than1%. In contrast, at protein levels

chemotactic, opsonic, phagocytic and killingcapacity.58 59 Furthermore, genetic variants influ-encing host defence mechanisms such as CARD15/NOD231 60 and TLR261 have been reported to beassociated with an enhanced probability ofacquiring SBP. TLR2 polymorphisms and NOD2variants seem to represent supplementary riskfactors since the simultaneous presence of bothunfavourable polymorphisms markedly increasesthe risk of SBP.61 This underlines the knowninteraction of NOD2 and TLRs, in particular themodulation of TLR2-dependent cytokine responsesby NOD2.62

Medication can also affect the chances of devel-oping SBP. The use of proton pump inhibitors (PPI)has been proposed to facilitate SIBO and thus tocontribute to pathological BT. In fact, retrospectivecaseecontrol studies reveal a potential associationbetween the use of PPI and development ofSBP.63 64 Considering the frequently inadequateoveruse of PPI in patients with cirrhosis, wetherefore recommend restricting their use to indi-cations of proven benefit. In contrast, non-selectiveb-blockers (NSBB) may prevent SBP.65 66 It istempting to speculate that this benefit relates to animprovement in chemotaxis, proinflammatorycytokine release and killing capacity reported forb-adrenergic antagonists in various experimentalsettings.67 68 Since the sympathetic nervous systemaffects PMN chemotaxis, the question arises as tohow treatment with NSBB affects the validity ofdiagnosing SBP based on PMN count in the asciticfluid.

DIAGNOSIS OF SBPSymptoms and signs are frequently absent inpatients with SBP,69 so a diagnostic paracentesisshould be performed in all patients with ascitesadmitted to hospital regardless of whether or notthere is clinical suspicion. Diagnosis should beprompt and treatment must not be delayed untilthe microbiology results are available. Thus, in allthe available guidelines, diagnosis is based on a fixeddefined cut-off PMN count in the ascitic fluid.1e4 Inpatients with haemorrhagic ascites (ie, red bloodcell count >10 000/mm3), subtraction of one PMNper 250 red blood cells should be made to adjust forthe presence of blood in ascites. Owing to the shortlifespan of PMN, their ascitic count is independentof diuretics and/or other modulations of ascitesvolume. In contrast, lymphocytes which havea long lifespan increase in concentration duringdiuresis.70 Moreover, differential diagnoses ofpredominant lymphocytosis in ascitic fluid includetuberculous peritonitis, neoplasms, congestiveheart failure, pancreatitis and myxedema, but notusually SBP. PMN are therefore used to define SBP,and the greatest sensitivity is reached at a cut-offvalue of 250 PMN/mm3, although the best speci-ficity has been reported with a cut-off of 500 PMN/mm3.71e74 However, since it is important not tomiss a case of SBP, the most sensitive cut-off valueis used. Nonetheless, this upper limit has been setquite arbitrarily since it was tested in the setting of

culture-positive peritonitis. Thus, the range ofPMN in truly non-infected ascitesdthat is, theascitic PMN count that is clinically relevant forthe patientdis not known. Moreover, SBP causedby Gram-positive cocci has been reportedfrequently to have a PMN count below thethreshold of 250/mm3.75 Interestingly, bactDNAfrom Gram-negative bacteria in ascitic fluid isassociated with a higher ascitic PMN count thanbactDNA from Gram-positive bacteria,76 under-scoring the differences in stimulatory capacity forPMN migration depending on the type of bacteria.

Microscopy versus automated cell counterAscitic PMN cell counts can be determined eitherby a traditional haematological method usinga light microscope and a manual counting chamberor by automated cell counters.77e79 Currentguidelines either do not state specifically themethod to be used2 4 or recommend microscopy asthe preferred method.1 However, microscopic eval-uation is labour-intensive, time-consuming and hashigh intraoperator and interoperator variability. Incontrast, automated cell counters, if available, areeasily accessible in emergencies and provide resultswithin minutes at low cost. Their use hasrecently been validated in patients with cirrhoticascites,77 79 revealing sufficient sensitivity fordetection of SBP, and thus should be recommended.However, it is important to stress that not allautomated cell counters fulfil the quality criteria.These include sufficient functional sensitivity, testprecision and accuracy, particularly for automatedleucocyte counts in ascites even with low cellconcentrations (eg, XE-5000 (Sysmex, Mundelein,IL, USA), Advia 120 (Erlangen, Germany), IrisiQ200 (Chatsworth, CA, USA), CellDyn-4000(Wiesbaden, Germany)).None of the recent guidelines recommends the

use of reagent test strips to assess leucocyte esteraseactivity of activated PMNs for the diagnosis of SBPowing to unacceptable rates of false negativeresults.80 However, most of the strips used to datehave been developed for urinary tract infectionswith a threshold of >50 PMN/mm3.81 Recently,a reagent strip test has been calibrated for asciticfluid with a cut-off of 250 PMN/mm3.82 Validityscores achievable were reported to be 100% sensi-tivity and 100% negative predictive value.However, this needs to be confirmed in largemulticentre trials and, furthermore, the test wasnot interpretable in bloody, chylous or biliousascitic fluid.

Bacterial DNA detection and culture techniquesDetection of bacterial DNA (bactDNA) usingvarious approaches has recently been proposed inthe ascitic fluid of patients with cirrhosis.83e85 Theadvantage of such a system would be the imme-diate identification of the causative bacteria, thusenabling more accurately targeted antibiotic treat-ment. BactDNA is found in the ascitic fluid ofabout 40% of patients with cirrhosis, being derivedmainly from Gram-negative bacteria.84 85 However,

Gut 2012;61:297e310. doi:10.1136/gutjnl-2011-300779 299


group.bmj.com on November 26, 2014 - Published by http://gut.bmj.com/Downloaded from

http://gut.bmj.com/http://group.bmj.com

detection of bactDNA in ascites or serum was notassociated with an enhanced incidence of SBP anddoes not appear to predict the development ofbacterial infections.86

Culture techniquesGram staining of peritoneal fluid is rarely helpful87

and is not recommended. In contrast, culture is therecommended procedure. Although only a fewspecies and genera are found to cause SBP, morethan 70 different microbial species have beenisolated from the ascitic fluid of patients withbacteriologically-confirmed SBP.88 Classical culturetechniques fail to grow bacteria in up to 65% ofneutrocytic ascites. Bedside inoculation of ascitesinto blood culture bottles has been shown toincrease the sensitivity to nearly 80%.89e91 In thisregard, non-radiometric (eg, colorimetric BacTec)systems in particular have improved the time todiagnosis since they are faster than conventionalblood culture bottles.89 Handling processes influ-ence culture results and delay in transport increasesfalse negative results.92 Separate and simultaneousblood cultures should be collected since 30e58% ofSBP cases are associated with bacteraemia.30 93

Other markers of inflammation and secondaryperitonitisOther markers found to be indicative of SBP includeascitic pH, lactate dehydrogenase, lactate (andcorresponding arterialeascitic gradients), but noneof these is sufficiently predictive or discriminativeand may be increased in malignancy-relatedascites.72 74 94 95 Proteins such as granulocyteelastase96 and lactoferrin97 released by PMN uponactivation have likewise been shown to beincreased in SBP. Lactoferrin was reported to giverates of sensitivity and specificity of 95.5% and97%, respectively, using a cut-off value of 242 ng/ml and to decrease to below the cut-off value inpatients responding to treatment.97 However,because of the small number of SBP cases in thisinvestigation, confirmation is required in multi-centre trials including assessment of its accuracy inhaemorrhagic and coexisting malignant ascites.Differentiation of SBP from secondary peritonitis

due to perforation or inflammation of an intra-abdominal organ is clinically very relevant as theassociated mortality is exceedingly high.98 In fact,all patients with perforated secondary peritonitisnot undergoing timely surgery have been reportedto die during hospitalisation and, thus, delayeddiagnostic investigation is fatal. However, theproposed criteria to suspect secondary peritonitis(eg, inadequate response to therapy, multipleorganisms)1 3 are identified too late and thereforerapid and accurate ‘chemical’ parameters availableat the time of paracentesis are needed. Parametersproposed by Runyon et al are neutrocytic asciteswith at least two of the following three criteria:ascitic fluid total protein >1 g/dl (in contrast toSBP), glucose 225 mU/ml.99 The sensitivity of these criteria can be less

than 68%98 99 and thus can be optimised. In addi-tion, Wu et al reported that ascitic fluid with eitheralkaline phosphatase >240 U/l or carcinoem-bryonic antigen >5 ng/ml in 80% of cases reflectsperitonitis of secondary origin.100 Although no dataare available on the diagnostic accuracy of thecombined criteria (ie, those of either Wu et al or

Runyon et al), they are likely to improve sensitivityand should be tested prospectively. In themeantime, we strongly recommend performing anabdominal CT scan as soon as any of these featuresare present.101

TREATMENT OF SBPTreatment has to be started immediately afterdiagnosis of SBP and therefore is empirical sinceculture results are not available at this time point.The strain of bacteria causing SBP mainly dependson the site of acquisition. However, none of theinternational guidelines to date differentiates

Box 1

Key messages established unequivocally< Clinical judgement does not rule out SBP and

thus a diagnostic paracentesis should beperformed in all patients with cirrhosis andascites at hospital admission and/or in case ofgastrointestinal bleeding, shock signs of inflam-mation, worsening of liver/renal function orhepatic encephalopathy.

< SBP is defined by >250 PMN/mm3 andbacterascites by positive culture results ofascitic fluid in the absence of PMN >250/mm3.

< Ascitic fluid culture is important to guideantibiotic therapy and should be performed inall patients before starting antibiotic treatmentby inoculation of ascites into blood culturebottles at the patient’s bedside.

Controversial but proposed< PMN count in ascitic fluid can be determined

either by microscope OR appropriate automatedcell counters. Reagent strips currently cannot berecommended for rapid diagnosis of SBP butascites-calibrated sticks may become available.

< Bacterial DNA is not useful in detecting orpredicting the occurrence of SBP.

Questions to be addressed in the future< Are there potential differences in the detection

of SBP dependent on the use of b-blockers andthe type of causative bacteria (Gram-positive vsGram-negative)?

< Is the fixed cut-off PMN count used for definingSBP the best choice, or is the chemotacticcapacity of each individual patient relevant?

< Which parameters are sufficiently sensitive toguide rapid imaging for detection of secondaryperitonitis?

300 Gut 2012;61:297e310. doi:10.1136/gutjnl-2011-300779




between nosocomial and community-acquired SBPwith regard to the type of antibiotic regimen to use.This may be deleterious since nosocomial infectionsare associated with high rates of bacterial multi-resistance and mortality (J G Acevedo, personalcommunication, 2009).24 102 Patients with cirrhosisare also at increased risk of healthcare-associatedinfections,103 but studies are needed to determinethe associated risk for multiresistant bacteriacausing SBP.

Community-acquired SBP: complicated anduncomplicated casesHistorically, Gram-negative bacteriadalmostexclusively Enterobacteriaceaedhave been isolatedin the overwhelming majority of SBP cases. Morerecently, several studies have found an increasingrate of infections with Gram-positive bacteria andresistant microorganisms (J G Acevedo, personalcommunication, 2009).24 29 102 However, inpatients with no previous hospitalisation and noprior antibiotic treatment, the causative bacteriastill usually belong to the easily treatable Enter-obacteriaceae family of bacteria. Several antibioticshave been recommended for the initial treatment ofSBP in these cases including cefotaxime or otherthird-generation cephalosporins, amoxicillin-clav-ulanic acid or quinolones. Although earlier trialshave shown comparable efficacy of intravenousamoxicillin/clavulanic acid (1/0.2 g every 8 h) andintravenous cefotaxime in the treatment of SBP,recent increases in resistance to aminopenicillin/b-lactamase inhibitors104 may limit their useful-ness. In patients presenting without complicatingfactors that may worsen therapeutic efficacy, oraltreatment with quinolones appears sufficient incountries with a relatively low rate of quinolone-resistant strains of E coli. Possible complicatingfactors include shock, ileus, gastrointestinalbleeding, severe hepatic encephalopathy or renaldysfunction (serum creatinine >3 mg/dl).105

Nosocomial SBP: treatment failure, risk factors andrecommendationsIn nosocomial SBP, use of the antibiotics recom-mended above (third-generation cephalosporins,amoxicillin/clavulanic acid or quinolones) hasrecently led to disappointing and unacceptably lowrates of resolution (J G Acevedo, personal commu-nication, 2009).29 106 Resistance to third-generationcephalosporins and quinolones has been reported toincrease continuously and to reach levels of23e44% and 38e50%, respectively, in some insti-tutions and countries (J G Acevedo, personalcommunication, 2009).24 29 106 107 In addition, theincidence of extended-spectrum b-lactamase(ESBL)-producing bacteria as well as multiresistantGram-positive bacteria such as Enterococcus faeciumor methicillin-resistant Staphylococcus aureus(MRSA) causing nosocomial SBP is alarming (table1). MRSA has been found in 24e27% of cases ofSBP, with detection of S aureus in ascites severalyears ago.75 112 Fortunately, the numbers aredecreasing in most European countries.113 In

contrast, the Study for Monitoring AntimicrobialResistance Trends reported that hospital-acquiredESBL-positive E coli in any intra-abdominal infec-tion have increased in Europe from 4.3% in 2002 to11.8% in 2008.114 115 ESBLs cause resistance tovarious types of newer b-lactam antibioticsincluding third-generation cephalosporins andmonobactams and, in addition, frequently alsocarry genes encoding resistance even to other anti-biotics including quinolones, tetracyclines andantifolates.116 ESBL resistance genes/plasmidsrapidly spread around the world, with foreigntravel being associated with intestinal colonisationrates as high as 32% in Asia (and 88% specifically inIndia).117 118 Moreover, colonisation of these resis-tant organisms persists in a large proportion ofpatients for many months117 and any antibiotictreatment causes selective pressure, accelerating theclinical relevance of these bacteria.119 For SBP,ESBL-positive strains are not yet as frequent as inAsia but have been reported to cause up to 22% ofnosocomial infections in Spain (J Fernandez,personal communication, 2010). However, amongEuropean countries and even among institutions inthe same country, there are wide differences inresistance rates. For instance, for E coli isolates,susceptibility rates of ciprofloxacin or ampicillin/sulbactam are 90% and 65%, respectively, inEstonia but are 52% and 32% in Turkey.114

The clinical relevance of these numbers isreflected in the associated morbidity, healthcare-associated costs and mortality. In a number ofindependent investigations, in-hospital mortalityand/or 30-day mortality have been shown to beincreased in nosocomial SBP caused by multi-resistant bacteria compared with commonbacteria (J G Acevedo, personal communication,2009).75 102 106 111 In some series, most patientswith SBP due to multiresistant bacteria died withinthe first 5 days after the diagnosis of SBP was madeand, indeed, none of the patients with persistentinfection survived.29 A meta-analysis of recentlypublished data found a four times increased risk ofmortality associated with bacterial resistance inSBP (figure 2). Nosocomial SBP due to ESBL strainsor to multiresistant bacteria is often associatedwith a failure of first-line empirical antibiotictreatment.29 102 109 Indeed, the need for escalationof treatment associated with poor survival ispredictive of in-hospital mortality24 29 and there-fore must be avoided. The use of carbapenems andglycopeptides would be safest and easiest since noresistance has so far been reported in cases of SBP,but this is not practical and the choice of antibi-otics needs to be stratified for parameters definingthe risk of resistant bacteria. This includes hostfactors as well as validated knowledge of theresistance profile of bacteria acting in the setting inwhich the patient is diagnosed and treated.Reported independent risk factors for bacterialmultiresistance are previous hospitalisation(particularly within 3 months and intensive caretreatment) and prior prophylactic or therapeuticantibiotic treatment (figure 3).24 29 120 121 It is

Gut 2012;61:297e310. doi:10.1136/gutjnl-2011-300779 301




therefore suggested that, in patients with cirrhosiswho develop nosocomial SBP and present with suchrisk factors, a more effective first-line empiricalantibiotic therapy with a broader spectrum shouldbe used, namely carbapenems. However, thisregimen should be de-escalated as soon as possible ifmicrobiological results reveal non-resistant easilytreatable causative microorganisms. This minimisesresistance selection pressure on the carbapenemsand underlines the paramount importance ofobtaining appropriate microbiological cultures.Global susceptibility statistics from intra-abdom-inal infections show that the susceptibilities ofGram-negative isolates to the carbapenems haveremained stable over the past years, with E coli andK pneumoniae isolates, including ESBL-positiveisolates, being 98e100% susceptible.122 Imple-menting carbapenems as first-line treatment inpatients with nosocomial SBP with risk factors formultiresistant bacteria can therefore save lives. Thishas also been recommended in recent guidelines onthe treatment of sepsis,123 aiming at rapid initia-tion of an antibiotic regimen likely to cover allexpected causative microorganisms. The sameshould be even more true for patients withdecompensated cirrhosis who have an enhancedproinflammatory response to bacterial stimuli124

and exhibit an increased susceptibility for anyvasodilatory stimulus due to the already highlyhyperdynamic splanchnic circulation.125

Treatment of bacterascitesIt is controversial whether culture-positive resultsin the absence of an increased PMN count in theascitic fluid require immediate initiation of antibi-otic therapy. Some guidelines recommend antibi-otic treatment only in patients with signs ofinfection or inflammation.4 Otherwise, a follow-upparacentesis should establish whether SBP ispresent (PMN count >250/mm3) and thus whethertreatment is indicated. However, this is based ona single-centre observational cohort study126 andhas not been addressed prospectively. Until then wethink that considering the lack of symptoms ina large number of cirrhotic patients even in pres-ence of severe bacterial infection antibiotic treat-ment should be used in case of bacterascites.

Use of albumin as adjuvant treatmentIn patients with cirrhosis with SBP, a prospectiverandomised comparative study reported that adju-vant administration of high-dose albumin (1.5 g/kgon day 1 and 1 g/kg on day 3) with antibiotictreatment prevented worsening of renal functionwith a concomitant improvement in in-hospitaland 3-month survival.108 However, this regimen ismainly effective in high-risk patients characterisedby serum bilirubin >4 mg/dl. In addition, in unse-lected patients with SBP, even low-dose albumin(10 g/day on days 1e3) has been shown to reducetumour necrosis factor and interleukin 6 levels inserum and ascites and to prevent increases in serumNOx induced by SBP.110 Therefore, future trialsneed to determine whether other patients withTa

ble1

Dataon

bacterialresistance

inSBPandassociated

mortalityinlast

decade

Reference

Country

Num

berof

patients/num

ber

SBPepisodes

(yearwhere

stated)

Resistanceratesforantibiotics

(inculturable

bacteria)

ESBL(%

ofcultured

bacteria)

MRSA(%

ofcultured

bacteria)

Candida

spp

Mortalitywithvs

without

multiresistant

bacteria

(orresistance*/failure

first-

linetherapyy

)RR(95%

CI)

Singh

etal,2002

108

USA

42/61

25(1991e

1995)

17(1996e

2001)

Multiresistant:19%(overall)

8.3%

(1991e

1995)

38.5%(1996e

2001)

8.1%

5.4%

12%

30-day

mortality:

4/7vs

5/30,RR3.43

(1.23to

9.56)

Park

etal,2003

109

Korea

75/87(1995)

195/222(1998)

207/271(1999)

Cefotaximez

:7%

(1995)

and28%(1999)

Ciprofloxacinz

:10%(1995)

and32%

(1999)

Ampicillinz

:83%(1995)

and76%(1999)

7.9%

(1995)

9.7%

(1998)

19.8%(1999)

Not

stated

Not

stated

In-hospitalmortality:

ESBL94%vs

others

notstated

Angeloniet

al,2008

106

Italy

32/38

Cefotaxime:

44%

33%

00

Not

stated

Heo

etal,2009

110

Korea

145/157

Cefotaxime:

15.6%z

Ciprofloxacin:

20.4%z

Ampicillin:

61.1%z

11.1%

01.9%


20/23vs

13/132,

RR8.83

(5.15to

15.15)y

4/6forESBLvs

29/151

forothers,

RR3.47

(1.81to

6.67)

Cheonget

al,2009

24

Korea

236/236

Third-generationcephalosporin:16.3%z

4.7%

0Not

stated

30-day

mortality:

46/61vs

69/175,

RR1.91

(1.52to

2.41)y

Umgelteret

al,2009

29

Germany

101/101

Cefotaxime:

33%

Ciprofloxacin:

45.2%Amoxicillin/

clavulanic

acid:38.6%

Multiresistant:26.8%

00

4.9%

8/17

vs1/12,RR5.65

(0.81to

39.42)*

(88.9%

with

multiresistant

bacteria;

personalcommunication)

Fernandezet

al,2011

111

Spain

100/126

Multiresistantbacteria:22%

(nosocom

ial);

3%(com

munity-acquired)

16%/2%

00


17/116

vs5/10,

RR3.41

(1.60to

7.29)

PooledxRR3.87

(1.76to

8.52)

zFor

Gram-negativebacteria.

xRandom

effect

meta-analyses

performed

inReview

Manager

Version

5.1.

Copenhagen:

TheNordicCochraneCentre,

TheCochraneCollaboration,

2011.StatisticalmethodMantel-H

aenszel.

ESBL,

extended-spectrum

b-lactam

ase;

MRSA,methicillin-resistantStaphylococcusaureus;SBP,

spontanous

bacterialperitonitis.

302 Gut 2012;61:297e310. doi:10.1136/gutjnl-2011-300779




cirrhosis could also benefit and to establish the doseand timing of albumin needed to give most benefitto the individual patient.

Duration of treatment and control of treatmentsuccessAntibiotic treatment can safely be discontinuedafter the ascites PMN count has decreased to

Table2

Publishedrandom

ised

controlledtrialsincludingprimaryprophylaxisof

SBPtreatm

ent

Reference

Intervention,

controls,comparison

Patients

Follow-up

Previous

SBP

GIbleed

Ascites

protein(g/dl)

Child

class

(A/B/C/score)

Exclusion

criteria

Incidence

ofSBP

Survival

As

Ac

BO

IS

C/D

Soriano

etal,1991

136

Norfloxacin400mg/day

vsno

treatm

ent

32/31

Nodata

6%4/61

(6.6%)

All<1.5

TP:0.7160.3

Con:0.6760.3

TP:2/3/17

Con:1/14/16

Recentinfection

GIbleed

TP:0/32

(0%)

Con:7/31

(22.5%

)p<

0.05

TP:30/32

(93.7%

)Con:26/31

(83.9%

)NS

??

e+

+e

Rolachon

etal,1995

137

Ciprofloxacin750mg/week

vsplacebo

28/32

6months

11%

No

All<1.5

TP:0.9460.3

Con:1.0360.3

TP:0/17/11

Con:1/18/13

HCC

GIbleed

HE

TP:1/28

(3.6%)

Con:7/32(22%

)p<

0.05

TP:24/28

(85.7%

)Con:26/32

(81.2%

)NS

??

++

++

Non-com

pliant

(n¼3

),withdraw

alor

lost

tofollow-up

(n¼5

),overall13%

Singh

etal,1995

134

Trimethoprim-

sulfamethoxazole

double-strength13

/day

(5days/week)

vsno

treatm

ent

30/30

90days

(7e682)

22%

13%

Nodata

Nodata

Nodata

TP:1/30

(3%)

Con:7/30

(23.3%

)(8/30(27%

)for

endpoint*

p<0.05

TP:28/30

(93%

)Con:24/30

(80%

)NS

??

e?

+e

Novella

etal,1997

51


continuous

vsnorfloxacin

400mg/dayin-hospital

56/53

43+3week

No

23/109

(>21%)

TP:1.06

0.2

Con:0.96

0.1

TP:0/29/27

Con:0/24/29

HCC

Bilirubin

>15

mg/dl

TP:1/56

(1.8%)

Con:9/53

(16.9%

)p<

0.01

TP:75%

Con:62%

NS

?+

e?

?e

Drop-out:>10%

Grange

etal,1998

138


vsplacebo

53/54

6months

No

No

All<1.5

TP:0.9360.29

Con:1.0460.0.3

Nodata

GIbleed

HCC

SBPor

bacteraemiay

TP:1/53

(1.9%)

Con:9/54

(16.7%

)p<

0.05

TP:45/53

(84.9%

)Con:44/54

(81.5%

)NS

??

?+

++

Lost

tofollow-up

(4/53and4/54)

Drop-out:7.5%

non-compliant

(3/53and2/54),

withdraw

al(2/53),

overall14%

Alvarez

etal,2005

139


vstrimethoprim-

sulfamethoxazole

160/800mg5days/week

32/25

3e547days

39%

No

Alsopts

with

>1.5

Norfloxacin:

0.9660.55

SMT:

1.3760.84

p<0.05

Norfloxacin:

1/10/21

SMT:

0/8/17

Antibiotic

within

2weeks

GIbleed

within

1week

HCC/m

alignancy

Norfloxacin:

3/32

(9.4%)

SMT:

4/25

(16%

)NS

Norfloxacin:

25/32(78.1%

)SMT:

20/25

(75%

)NS

++

e?

?e

Nodata

Fernandez

etal,2007

140


vsplacebo

35/33

12months

No

3/68

(4.4%)

All<1.5

TP:0.9360.29

Con:1.0460.3

TP:9.96

1.5

Con:10.461.5

HCC,HIV

organic

renaldisease

TP:2/35

(5.7%)

Con:10/33(30.3%

)p<

0.05

TP:25/35

(71.4%

)Con:20/33

(60.6%

)NS

++

++

+?

Lost

tofollow-up

(3/35and2/33),

oneprotocol

violation,

non-

compliant

(3/35

and3/33)

Terg

etal,2008

141

Ciprofloxacin500mg/day

vsplacebo

50/50

12months

No

Nodata

All<1.5

TP:0.8460.01

Con:0.85+0.4

TP:8.36

1.3

Con:8.56

1.5

HE,

HCC/

malignancy

creatinine

>3mg/dl

platelets

<98

000

TP:2/50

(4%)

Con:7/50

(14%

)NS

TP:44/50

(80%

)Con:36/50

(72%

)p<

0.05

*SBPor

spontaneousbacteraemia.

yPrim

aryendpointisGram-negativeinfections

butno

inform

ationon

SBPalone.

As,

allocationsequence;Ac,

allocationconcealment;B,blinding;O,outcom

edata

setcomplete?;I,intentionto

treat;S,samplesize

calculation;

C/D,compliance/drop-out

rate.

Con,control;GI,gastrointestinal;HCC,hepatocellularcarcinom

a;HE,

hepatic

encephalopathy;NS,notsignificant;SBP,

spontanous

bacterialperitonitis;

SMT,

trimethoprim-sulfamethoxazole;TP,therapy.

304 Gut 2012;61:297e310. doi:10.1136/gutjnl-2011-300779




a higher rate of quinolone-resistant organisms131

and, in our view, should therefore be avoided. Datasupporting the use of trimethoprim/sulfamethox-azole are weak,132 while its side effects are poten-tially dangerous and probably under-reported.133

Moreover, resistance to this class of antibiotics hasincreased to a degree that it is no longer recom-mended as the first-line choice for the empiricaltreatment of urinary tract infections in somecountries.134 In patients with cirrhosis withgastrointestinal haemorrhage, quinolones are mostfrequently used and have been found to decreasethe incidence of severe infections (SBP and/orsepticaemia) and mortality. However, in patientswith bleeding necessitating invasive procedures,infections are increasingly caused by Gram-positivebacteria and intravenous delivery may be moreappropriate than the oral route. In fact, the third-generation cephalosporin ceftriaxone administeredintravenously has been shown to be superior to oralnorfloxacin in patients with advanced cirrhosis(ie, with at least two of the following: ascites,severe malnutrition, encephalopathy or bilirubin>3 mg/dl).135

With regard to the use of antibiotics for primaryprophylaxis in the setting of low protein ascites(

Nonetheless, guidelines state very cautiously thatthe long-term use of norfloxacin can be justified4 orshould be considered in these selected patients.1 3

However, since this trial fulfils the highest qualitycriteria (Jadad score 5) and represents a well-definedgroup of patients, we consider the use of norflox-acin for primary prophylaxis as a standard of careprocedure.

Limitations in antibiotic prophylaxis andalternativesThe longer the duration of antibiotic treatment, thegreater is the risk for selection of resistant strainsand the lower is the chance of reducing the inci-dence of SBP. In fact, survival advantage usingnorfloxacin as primary prophylaxis in highlyselected patients is most marked during the first3 months of treatment (94% vs 62%, p¼0.003) anddecreases over time.138 145 We therefore proposethat the use of norfloxacin for primary prophylaxisshould also be considered in unselected patientswith low protein ascites if liver transplantation isa realistic option within a few months. Althoughthere are no long-term data, the same time courseof antibiotic efficacy is likely to be present as insecondary prophylaxis. Its use is recommended tobe continued until liver transplantation or until thedisappearance of ascites (eg, in alcoholics stoppingalcohol ingestion).2 3 In any other case, antibiotictreatment guidelines support long-term use but, inour view, improvement in liver disease should leadto interruption of treatment.Overall, the continuous use of a single antibiotic

appears not to be the optimal solution and effortsshould be made to seek alternatives which couldinclude antibiotic cycling. The basic principle ofcycling antibiotics is that bacteria acquiring resis-tance to the first course of treatment would remainsusceptible to the second regimen, and so on. Inthis context, future trials should test the use ofrifaximin since (a) it belongs to a different antibi-otic class from the antibiotics tested prospectivelyso far; (b) it exerts a broad range of antimicrobialactivity including Gram-positive bacteria141; (c) itappears to cause considerably less bacterial resis-tance147 148; and (d) it acts predominantly in thesmall intestine,147e149 the site of bacterial over-growth in cirrhosis. Finally, as has been pointed outby others,150 151 effective non-antibiotic approachesin reducing the incidence of SBP represent the HolyGrail. Interestingly, a significant decrease in theincidence of postoperative infections has beenreported in a cohort study of patients with cirrhosistreated with propranolol and ciprofloxacincompared with ciprofloxacin alone after laparo-scopic surgery.152 Moreover, NSBB have beenreported to ameliorate pathological BT in experi-mental cirrhosis.153 Finally, recent meta-analyses ofavailable data indicate that NSBB lower the risk ofSBP in patients with cirrhosis which may occurindependently of the haemodynamic responseachieved.65 66 However, the use of NSBB in patientswith refractory ascites has been suggested toworsen prognosis154 155 and to be associated with

haemodynamic adverse effects after large-volumeparacentesis.156 Future prospective trials thereforeneed to address these questions in detail in order toestablish the use of NSBB in the right patient at theright time.Cisapride, a serotonin 5-HT4 receptor agonist

and intestinal prokinetic drug, has been shownto decrease SIBO and BT in experimentalcirrhosis41 157 but was abandoned due to cardiacside effects. Nonetheless, these encouraging resultsshould stimulate human prospective trials investi-gating other prokinetics such as the new highlyselective 5-HT4 receptor agonist prucalopride whichshowed no interaction at other receptor sites.158

Other promising approaches reported to amelio-rate BT in experimental cirrhosis include orallyadministered conjugated bile acids, chol-ylsarcosine,159 insulin-like growth factor I)160 andanti-tumour necrosis factor.161 Probiotics have beenreported to correct bacterial overgrowth, stabilisemucosal barrier function, improve neutrophilfunction and decrease BT in experimental liverfailure.162 163 In patients with cirrhosis, symbiotictreatment significantly reduced endotoxin levelsand improved the Child-Pugh functional class innearly 50% of cases.164 Similarly, the addition offibre to lactobacilli decreased postoperative bacterialinfections after liver transplantation.165 Probioticsmay even be helpful in limiting the development ofbacterial resistance, and trials are ongoing toinvestigate their efficacy in eradicating carba-penem-resistant bacteria as well as the decolonisa-tion of MRSA in carrier patients (NCT00722410/NCT00941356).

Box 3

Key messages: established unequivocally< In patients with gastrointestinal haemorrhage,

antibiotic prophylaxis is mandatory using third-generation cephalosporins (eg, ceftriaxone) insevere liver disease or quinolones in less severeand uncomplicated cases.

< Secondary prophylaxis is recommended afterresolution of SBP with the strongest evidencesupporting use of norfloxacin.

Controversial but proposed< Primary prophylaxis can be justified in patients

with low protein ascites (

Acknowledgements The authors thank Professor Thomas Glueck(Hospital Trostberg, Germany) for helpful discussions, scientificadvice and for reviewing the manuscript.

Competing interests None.

Contributors RW and AG wrote the manuscript. AK performed thestatistical analysis and meta-analysis.

Provenance and peer review Commissioned; externally peerreviewed.

REFERENCES1. Fernandez J, Navasa M, Planas R, et al. Primary prophylaxis of

spontaneous bacterial peritonitis delays hepatorenal syndromeand improves survival in cirrhosis. Gastroenterology2007;133:818e24.

2. Grange JD, Roulot D, Pelletier G, et al. Norfloxacin primaryprophylaxis of bacterial infections in cirrhotic patients withascites: a double-blind randomized trial [see comments]. JHepatol 1998;29:430e6.

3. Terg R, Fassio E, Guevara M, et al. Ciprofloxacin in primaryprophylaxis of spontaneous bacterial peritonitis: a randomized,placebo-controlled study. J Hepatol 2008;48:774e9.

4. Runyon BA. Management of adult patients with ascites due tocirrhosis: an update. Hepatology 2009;49:2087e107.

5. Fernandez J, Navasa M, Gomez J, et al. Bacterial infections incirrhosis: epidemiological changes with invasive procedures andnorfloxacin prophylaxis. Hepatology 2002;35:140e8.

6. Caly WR, Strauss E. A prospective study of bacterial infectionsin patients with cirrhosis. J Hepatol 1993;18:353e8.

7. Pinzello G, Simonetti RG, Craxi A, et al. Spontaneous bacterialperitonitis: a prospective investigation in predominantlynonalcoholic cirrhotic patients. Hepatology 1983;3:545e9.

8. Evans LT, Kim WR, Poterucha JJ, et al. Spontaneous bacterialperitonitis in asymptomatic outpatients with cirrhotic ascites.Hepatology 2003;37:897e901.

9. Castellote J, Girbau A, Maisterra S, et al. Spontaneous bacterialperitonitis and bacterascites prevalence in asymptomatic cirrhoticoutpatients undergoing large-volume paracentesis. JGastroenterol Hepatol 2008;23:256e9.

10. Jeffries MA, Stern MA, Gunaratnam NT, et al. Unsuspectedinfection is infrequent in asymptomatic outpatients withrefractory ascites undergoing therapeutic paracentesis. Am JGastroenterol 1999;94:2972e6.

11. Conn HO, Fessel JM. Spontaneous bacterial peritonitis incirrhosis: variations on a theme. Medicine (Baltimore)1971;50:161e97.

12. Rimola A, Garcia-Tsao G, Navasa M, et al. Diagnosis, treatmentand prophylaxis of spontaneous bacterial peritonitis: a consensusdocument. International Ascites Club. J Hepatol2000;32:142e53.

13. Chu CM, Chang KY, Liaw YF. Prevalence and prognosticsignificance of bacterascites in cirrhosis with ascites. Dig Dis Sci1995;40:561e5.

14. Follo A, Llovet JM, Navasa M, et al. Renal impairment afterspontaneous bacterial peritonitis in cirrhosis: incidence, clinicalcourse, predictive factors and prognosis. Hepatology1994;20:1495e501.

15. Nobre SR, Cabral JE, Gomes JJ, et al. In-hospital mortality inspontaneous bacterial peritonitis: a new predictive model. Eur JGastroenterol Hepatol 2008;20:1176e81.

16. Thuluvath PJ, Morss S, Thompson R. Spontaneous bacterialperitonitisein-hospital mortality, predictors of survival, and healthcare costs from 1988 to 1998. Am J Gastroenterol2001;96:1232e6.

17. Tito L, Rimola A, Gines P, et al. Recurrence of spontaneousbacterial peritonitis in cirrhosis: frequency and predictive factors.Hepatology 1988;8:27e31.

18. Toledo C, Salmeron JM, Rimola A, et al. Spontaneous bacterialperitonitis in cirrhosis: predictive factors of infection resolutionand survival in patients treated with cefotaxime. Hepatology1993;17:251e7.

19. Andreu M, Sola R, Sitges-Serra A, et al. Risk factors forspontaneous bacterial peritonitis in cirrhotic patients with ascites.Gastroenterol 1993;104:1133e8.

20. Silvain C, Besson I, Ingrand P, et al. Prognosis and long-termrecurrence of spontaneous bacterial peritonitis in cirrhosis. JHepatol 1993;19:188e9.

21. Terg R, Levi D, Lopez P, et al. Analysis of clinical course andprognosis of culture-positive spontaneous bacterial peritonitis and

neutrocytic ascites. Evidence of the same disease. Dig Dis Sci1992;37:1499e504.

22. Arvaniti V, D’Amico G, Fede G, et al. Infections in patients withcirrhosis increase mortality four-fold and should be used indetermining prognosis. Gastroenterology 2010;139:1246e56,1256.e1e5.

23. Altman C, Grange JD, Amiot X, et al. Survival after a firstepisode of spontaneous bacterial peritonitis. Prognosis ofpotential candidates for orthotopic liver transplantation. JGastroenterol Hepatol 1995;10:47e50.

24. Cheong HS, Kang CI, Lee JA, et al. Clinical significance andoutcome of nosocomial acquisition of spontaneous bacterialperitonitis in patients with liver cirrhosis. Clin Infect Dis2009;48:1230e6.

25. Almdal TP, Skinhoj P. Spontaneous bacterial peritonitis incirrhosis. Incidence, diagnosis, and prognosis. Scand JGastroenterol 1987;22:295e300.

26. Terg R, Gadano A, Cartier M, et al. Serum creatinine and bilirubinpredict renal failure and mortality in patients with spontaneousbacterial peritonitis: a retrospective study. Liver Int2009;29:415e19.

27. Franca AV, De Souza JB, Silva CM, et al. Long-term prognosisof cirrhosis after spontaneous bacterial peritonitis treated withceftriaxone. J Clin Gastroenterol 2001;33:295e8.

28. Kamani L, Mumtaz K, Ahmed US, et al. Outcomes in culturepositive and culture negative ascitic fluid infection in patientswith viral cirrhosis: cohort study. BMC Gastroenterol 2008;8:59.

29. Umgelter A, Reindl W, Miedaner M, et al. Failure of currentantibiotic first-line regimens and mortality in hospitalized patientswith spontaneous bacterial peritonitis. Infection 2009;37:2e8.

30. Cho JH, Park KH, Kim SH, et al. Bacteremia is a prognosticfactor for poor outcome in spontaneous bacterial peritonitis.Scand J Infect Dis 2007;39:697e702.

31. Appenrodt B, Grunhage F, Gentemann MG, et al. Nucleotide-binding oligomerization domain containing 2 (NOD2) variants aregenetic risk factors for death and spontaneous bacterialperitonitis in liver cirrhosis. Hepatology 2010;51:1327e33.

32. Wiest R, Garcia-Tsao G. Bacterial translocation (BT) in cirrhosis.Hepatology 2005;41:422e33.

33. Wiest R. Role of bacterial infections for hepatorenal syndrome.In: Gerbes A, ed. Frontiers in Gastrointestinal Research. Aszites,Hyponatremia and heptorenal syndrome. Basel, Switzerland:Krager, 2010.

34. Wells CL. Relationship between intestinal microecology and thetranslocation of intestinal bacteria. Antonie Van Leeuwenhoek1990;58:87e93.

35. Guarner C, Gonzalez-Navajas JM, Sanchez E, et al. Thedetection of bacterial DNA in blood of rats with CCl4-inducedcirrhosis with ascites represents episodes of bacterialtranslocation. Hepatology 2006;44:633e9.

36. Llovet JM, Bartoli R, March F, et al. Translocated intestinalbacteria cause spontaneous bacterial peritonitis in cirrhoticrats: molecular epidemiologic evidence. J Hepatol1998;28:307e13.

37. Chen Y, Yang F, Lu H, et al. Characterization of fecal microbialcommunities in patients with liver cirrhosis. Hepatology2011;54:562e72.

38. Yan AW, Fouts DE, Brandl J, et al. Enteric dysbiosis associatedwith a mouse model of alcoholic liver disease. Hepatology2011;53:96e105.

39. Bauer TM, Steinbruckner B, Brinkmann FE, et al. Small intestinalbacterial overgrowth in patients with cirrhosis: prevalence andrelation with spontaneous bacterial peritonitis. Am JGastroenterol 2001;96:2962e7.

40. Guarner C, Runyon BA, Young S, et al. Intestinal bacterialovergrowth and bacterial translocation in cirrhotic rats withascites. J Hepatol 1997;26:1372e8.

41. Pardo A, Bartoli R, Lorenzo-Zuniga V, et al. Effect of cisapride onintestinal bacterial overgrowth and bacterial translocation incirrhosis. Hepatology 2000;31:858e63.

41a. Teltschik Z, Wiest R, Beisner J, et al. Intestinal bacterialtranslocation in cirrhotic rats is related to compromised Panethcell antimicrobial host defence. Hepatology (in press).

42. Chiva M, Guarner C, Peralta C, et al. Intestinal mucosal oxidativedamage and bacterial translocation in cirrhotic rats. Eur JGastroenterol Hepatol 2003;15:145e50.

43. Ramachandran A, Prabhu R, Thomas S, et al. Intestinal mucosalalterations in experimental cirrhosis in the rat: role of oxygen freeradicals. Hepatology 2002;35:622e9.

44. Dunn DL, Barke RA, Ahrenholz DH, et al. The adjuvant effect ofperitoneal fluid in experimental peritonitis. Mechanism and clinicalimplications. Ann Surg 1984;199:37e43.

Gut 2012;61:297e310. doi:10.1136/gutjnl-2011-300779 307




45. Gabele E, Muhlbauer M, Paulo H, et al. Analysis of monocytechemotactic protein-1 gene polymorphism in patients withspontaneous bacterial peritonitis. World J Gastroenterol2009;15:5558e62.

46. Joyce LD, Hau T, Hoffman R, et al. Evaluation of the mechanismof zymosan-induced resistance to experimental peritonitis.Surgery 1978;83:717e25.

47. Runyon BA, Morrissey RL, Hoefs JC, et al. Opsonic activity ofhuman ascitic fluid: a potentially important protective mechanismagainst spontaneous bacterial peritonitis. Hepatology1985;5:634e7.

48. Such J, Guarner C, Enriquez J, et al. Low C3 in cirrhotic ascitespredisposes to spontaneous bacterial peritonitis. J Hepatol1988;6:80e4.

49. Runyon BA. Low-protein-concentration ascitic fluid ispredisposed to spontaneous bacterial peritonitis. Gastroenterol1986;91:1343e6.

50. Llach J, Rimola A, Navasa M, et al. Incidence and predictivefactors of first episode of spontaneous bacterial peritonitis incirrhosis with ascites: relevance of ascitic fluid proteinconcentration. Hepatology 1992;16:724e7.

51. Novella M, Sola R, Soriano G, et al. Continuous versus inpatientprophylaxis of the first episode of spontaneous bacterialperitonitis with norfloxacin. Hepatology 1997;25:532e6.

52. Wiest R, Leidl F, Kopp A, et al. Peritoneal fluid adipokines: readyfor prime time? Eur J Clin Invest 2009;39:219e29.

53. Guarner C, Sola R, Soriano G, et al. Risk of a first community-acquired spontaneous bacterial peritonitis in cirrhotics withlow ascitic fluid protein levels. Gastroenterology1999;117:414e19.

54. Obstein KL, Campbell MS, Reddy KR, et al. Association betweenmodel for end-stage liver disease and spontaneous bacterialperitonitis. Am J Gastroenterol 2007;102:2732e6.

55. Riordan SM, Skinner N, Nagree A, et al. Peripheral bloodmononuclear cell expression of toll-like receptors and relation tocytokine levels in cirrhosis. Hepatology 2003;37:1154e64.

56. Wasmuth HE, Kunz D, Yagmur E, et al. Patients with acute onchronic liver failure display “sepsis-like” immune paralysis. JHepatol 2005;42:195e201.

57. Berres ML, Schnyder B, Yagmur E, et al. Longitudinal monocytehuman leukocyte antigen-DR expression is a prognostic marker incritically ill patients with decompensated liver cirrhosis. Liver Int2009;29:536e43.

58. Hassner A, Kletter Y, Jedvab M, et al. Impaired monocytefunction in liver cirrhosis. Lancet 1979;1:329e30.

59. Rajkovic IA, Williams R. Abnormalities of neutrophilphagocytosis, intracellular killing and metabolic activity inalcoholic cirrhosis and hepatitis. Hepatology 1986;6:252e62.

60. Bruns T, Peter J, Reuken PA, et al. NOD2 gene variants area risk factor for culture-positive spontaneous bacterial peritonitisand monomicrobial bacterascites in cirrhosis. Liver Int2011:10e3231.

61. Nischalke HD, Berger C, Aldenhoff K, et al. Toll-like receptor(TLR) 2 promotor and intron 2 polymorphisms are associated withincreased risk for spontaneous bacterial peritonitis in livercirrhosis. J Hepatol 2011;55:1010e16.

62. Netea MG, Ferwerda G, de Jong DJ, et al. Nucleotide-bindingoligomerization domain-2 modulates specific TLR pathways forthe induction of cytokine release. J Immunol 2005;174:6518e23.

63. Bajaj JS, Zadvornova Y, Heuman DM, et al. Association ofproton pump inhibitor therapy with spontaneous bacterialperitonitis in cirrhotic patients with ascites. Am J Gastroenterol2009;104:1130e4.

64. Trikudanathan G, Israel J, Cappa J, et al. Association betweenproton pump inhibitors and spontaneous bacterial peritonitis incirrhotic patientsda systematic review and meta-analysis. Int JClin Pract 2011;65:674e8.

65. Senzolo M, Cholongitas E, Burra P, et al. beta-Blockers protectagainst spontaneous bacterial peritonitis in cirrhotic patients:a meta-analysis. Liver Int 2009;29:1189e93.

66. Krag A, Wiest R, Albillos A, et al. Reduced mortality with non-selective betablockers (NSBB) compared to banding is notrelated to prevention of bleeding or bleeding related mortality:a hypothesis for non-hemodynamic effects of NSBB. J Hepatol2011;54:S72.

67. Sanders VM, Straub RH. Norepinephrine, the beta-adrenergicreceptor, and immunity. Brain Behav Immun 2002;16:290e332.

68. Worlicek M, Knebel K, Linde HJ, et al. Splanchnicsympathectomy prevents translocation and spreading of E colibut not S aureus in liver cirrhosis. Gut 2010;59:1127e34.

69. Chinnock B, Afarian H, Minnigan H, et al. Physician clinicalimpression does not rule out spontaneous bacterial peritonitis in

patients undergoing emergency department paracentesis. AnnEmerg Med 2008;52:268e73.

70. Hoefs JC. Increase in ascites white blood cell and proteinconcentrations during diuresis in patients with chronic liverdisease. Hepatology 1981;1:249e54.

71. Albillos A, Cuervas-Mons V, Millan I, et al. Ascitic fluidpolymorphonuclear cell count and serum to ascites albumingradient in the diagnosis of bacterial peritonitis. Gastroenterology1990;98:134e40.

72. Garcia-Tsao G, Conn HO, Lerner E. The diagnosis of bacterialperitonitis: comparison of pH, lactate concentration and leukocytecount. Hepatology 1985;5:91e6.

73. Stassen WN, McCullough AJ, Bacon BR, et al. Immediatediagnostic criteria for bacterial infection of ascitic fluid. Evaluationof ascitic fluid polymorphonuclear leukocyte count, pH, andlactate concentration, alone and in combination. Gastroenterology1986;90:1247e54.

74. Yang CY, Liaw YF, Chu CM, et al. White count, pH and lactate inascites in the diagnosis of spontaneous bacterial peritonitis.Hepatology 1985;5:85e90.

75. Campillo B, Richardet JP, Kheo T, et al. Nosocomialspontaneous bacterial peritonitis and bacteremia in cirrhoticpatients: impact of isolate type on prognosis and characteristicsof infection. Clin Infect Dis 2002;35:1e10.

76. Gonzalez-Navajas JM, Bellot P, Frances R, et al. Presence ofbacterial-DNA in cirrhosis identifies a subgroup of patients withmarked inflammatory response not related to endotoxin. JHepatol 2008;48:61e7.

77. Angeloni S, Nicolini G, Merli M, et al. Validation of automatedblood cell counter for the determination of polymorphonuclear cellcount in the ascitic fluid of cirrhotic patients with or withoutspontaneous bacterial peritonitis. Am J Gastroenterol2003;98:1844e8.

78. Riggio O, Angeloni S, Parente A, et al. Accuracy of theautomated cell counters for management of spontaneousbacterial peritonitis. World J Gastroenterol 2008;14:5689e94.

79. Cereto F, Genesca J, Segura R. Validation of automated bloodcell counters for the diagnosis of spontaneous bacterialperitonitis. Am J Gastroenterol 2004;99:1400.

80. Nousbaum JB, Cadranel JF, Nahon P, et al. Diagnostic accuracyof the Multistix 8 SG reagent strip in diagnosis of spontaneousbacterial peritonitis. Hepatology 2007;45:1275e81.

81. Hiscoke C, Yoxall H, Greig D, et al. Validation of a method for therapid diagnosis of urinary tract infection suitable for use in generalpractice. Br J Gen Pract 1990;40:403e5.

82. Mendler MH, Agarwal A, Trimzi M, et al. A new highly sensitivepoint of care screen for spontaneous bacterial peritonitis usingthe leukocyte esterase method. J Hepatol 2010;53:477e83.

83. Bruns T, Sachse S, Straube E, et al. Identification of bacterialDNA in neutrocytic and non-neutrocytic cirrhotic ascites bymeans of a multiplex polymerase chain reaction. Liver Int2009;29:1206e14.

84. Frances R, Benlloch S, Zapater P, et al. A sequential study ofserum bacterial DNA in patients with advanced cirrhosis andascites. Hepatology 2004;39:484e91.

85. Frances R, Zapater P, Gonzalez-Navajas JM, et al. BacterialDNA in patients with cirrhosis and noninfected ascites mimicsthe soluble immune response established in patients withspontaneous bacterial peritonitis. Hepatology 2008;47:978e85.

86. Zapater P, Frances R, Gonzalez-Navajas JM, et al. Serum andascitic fluid bacterial DNA: a new independent prognostic factorin noninfected patients with cirrhosis. Hepatology2008;48:1924e31.

87. Chinnock B, Fox C, Hendey GW. Gram’s stain of peritoneal fluidis rarely helpful in the evaluation of the ascites patient. AnnEmerg Med 2009;54:78e82.

88. Conn HO. Unusual presentations of SBP. In: Conn HO, Rodes J,Navasa M, eds. London: Informa Healthcare, 2000:47e74.

89. Ortiz J, Soriano G, Coll P, et al. Early microbiologic diagnosis ofspontaneous bacterial peritonitis with BacT/ALERT. J Hepatol1997;26:839e44.

90. Siersema PD, de Marie S, van Zeijl JH, et al. Blood culturebottles are superior to lysis-centrifugation tubes forbacteriological diagnosis of spontaneous bacterial peritonitis. JClin Microbiol 1992;30:667e9.

91. Castellote J, Xiol X, Verdaguer R, et al. Comparison of twoascitic fluid culture methods in cirrhotic patients withspontaneous bacterial peritonitis. Am J Gastroenterol1990;85:1605e8.

92. Klaerner HG, Eschenbach U, Kamereck K, et al. Failure of anautomated blood culture system to detect nonfermentative gram-negative bacteria. J Clin Microbiol 2000;38:1036e41.

308 Gut 2012;61:297e310. doi:10.1136/gutjnl-2011-300779




93. Pelletier G, Salmon D, Ink O, et al. Culture-negative neutrocyticascites: a less severe variant of spontaneous bacterial peritonitis.J Hepatol 1990;10:327e31.

94. Gerbes AL, Jungst D, Xie YN, et al. Ascitic fluid analysis forthe differentiation of malignancy-related and nonmalignantascites. Proposal of a diagnostic sequence. Cancer1991;68:1808e14.

95. Wang SS, Lu CW, Chao Y, et al. Malignancy-related ascites:a diagnostic pitfall of spontaneous bacterial peritonitis by asciticfluid polymorphonuclear cell count. J Hepatol 1994;20:79e84.

96. Casafont F, Rivero M, Fernandez MD, et al. Granulocyte elastasein cirrhotic patients with spontaneous bacterial peritonitis. Dig DisSci 1999;44:1985e9.

97. Parsi MA, Saadeh SN, Zein NN, et al. Ascitic fluid lactoferrin fordiagnosis of spontaneous bacterial peritonitis. Gastroenterology2008;135:803e7.

98. Soriano G, Castellote J, Alvarez CE, et al. Secondary bacterialperitonitis in cirrhosis: a retrospective study of clinical andanalytical characteristics, diagnosis and management. J Hepatol2009;52:39e44.

99. Runyon BA, Hoefs JC. Ascitic fluid analysis in the differentiationof spontaneous bacterial peritonitis from gastrointestinal tractperforation into ascitic fluid. Hepatology 1984;4:447e50.

100. Wu SS, Lin OS, Chen YY, et al. Ascitic fluid carcinoembryonicantigen and alkaline phosphatase levels for the differentiation ofprimary from secondary bacterial peritonitis with intestinalperforation. J Hepatol 2001;34:215e21.

101. Wiest R, Schoelmerich J. Secondary peritonitis in cirrhosis: “Oilin fire”. J Hepatol 2009;52:7e9.

102. Song KH, Jeon JH, Park WB, et al. Clinical outcomes ofspontaneous bacterial peritonitis due to extended-spectrum beta-lactamase-producing Escherichia coli and Klebsiella species:a retrospective matched case-control study. BMC Infect Dis2009;9:41.

103. Merli M, Lucidi C, Giannelli V, et al. Cirrhotic patients are at riskfor health care-associated bacterial infections. Clin GastroenterolHepatol 2010;8:979e85.

104. Kresken M, Hafner D, Schmitz FJ, et al. Resistenzsituation beiklinisch wichtigen Infektionserregern gegenüber Antibiotika inDeutschland und im mitteleuropäischen Raum. 2010. http://www.p-e-g.org

105. Navasa M, Follo A, Llovet JM, et al. Randomized, comparativestudy of oral ofloxacin versus intravenous cefotaxime inspontaneous bacterial peritonitis [see comments].Gastroenterology 1996;111:1011e17.

106. Angeloni S, Leboffe C, Parente A, et al. Efficacy of currentguidelines for the treatment of spontaneous bacterialperitonitis in the clinical practice. World J Gastroenterol2008;14:2757e62.

107. Castellote J, Ariza X, Girbau A, et al. Antibiotic-resistantbacteria in spontanous bacterial peritonitis. Is it time to change?J Hepatol 2010;52:S69.

108. Singh N,Wagener MM, Gayowski T. Changing epidemiology andpredictors of mortality in patients with spontaneous bacterialperitonitis at a liver transplant unit. Clin Microbiol Infect2003;9:531e7.

109. Park YH, Lee HC, Song HG, et al. Recent increase in antibiotic-resistant microorganisms in patients with spontaneous bacterialperitonitis adversely affects the clinical outcome in Korea. JGastroenterol Hepatol 2003;18:927e33.

110. Heo J, Seo YS, Yim HJ, et al. Clinical features and prognosis ofspontaneous bacterial peritonitis in korean patients with livercirrhosis: a multicenter retrospective study. Gut Liver2009;3:197e204.

111. Campillo B, Dupeyron C, Richardet JP. Epidemiology of hospital-acquired infections in cirrhotic patients: effect of carriage ofmethicillin-resistant Staphylococcus aureus and influence ofprevious antibiotic therapy and norfloxacin prophylaxis. EpidemiolInfect 2001;127:443e50.

112. Piroth L, Pechinot A, Minello A, et al. Bacterial epidemiology andantimicrobial resistance in ascitic fluid: a 2-year retrospectivestudy. Scand J Infect Dis 2009;41:847e51.

113. Kock R, Becker K, Cookson B, et al. Methicillin-resistantStaphylococcus aureus (MRSA): burden of disease and controlchallenges in Europe. Euro Surveill 2010;15:19688.

114. Hawser SP, Bouchillon SK, Hoban DJ, et al. Incidence andantimicrobial susceptibility of Escherichia coli and Klebsiellapneumoniae with extended-spectrum beta-lactamases incommunity- and hospital-associated intra-abdominal infections inEurope: results of the 2008 Study for Monitoring AntimicrobialResistance Trends (SMART). Antimicrob Agents Chemother2010;54:3043e6.

115. Coque TM, Baquero F, Canton R. Increasing prevalence of ESBL-producing Enterobacteriaceae in Europe. Euro Surveill2008;13:19044.

116. Pitout JD, Laupland KB. Extended-spectrum beta-lactamase-producing Enterobacteriaceae: an emerging public-healthconcern. Lancet Infect Dis 2008;8:159e66.

117. Tangden T, Cars O, Melhus A, et al. Foreign travel is a major riskfactor for colonization with Escherichia coli producing CTX-M-typeextended-spectrum beta-lactamases: a prospective study withSwedish volunteers. Antimicrob Agents Chemother2010;54:3564e8.

118. Walsh TR, Weeks J, Livermore DM, et al. Dissemination ofNDM-1 positive bacteria in the New Delhi environment and itsimplications for human health: an environmental point prevalencestudy. Lancet Infect Dis 2011;11:355e62.

119. World Health Organisation. Combat Drug Resistance: No ActionToday Means No Cure Tomorrow. World Health Day 2011, 2011.http://www.who.int/mediacentre/multimedia/podcasts/en/

120. Rodriguez-Bano J, Picon E, Gijon P, et al. Risk factors andprognosis of nosocomial bloodstream infections caused byextended-spectrum-beta-lactamase-producing Escherichia coli.J Clin Microbiol 2010;48:1726e31.

121. Rodriguez-Bano J, Picon E, Gijon P, et al. Community-onsetbacteremia due to extended-spectrum beta-lactamase-producingEscherichia coli: risk factors and prognosis. Clin Infect Dis2010;50:40e8.

122. Hoban DJ, Bouchillon SK, Hawser SP, et al. Susceptibility ofgram-negative pathogens isolated from patients withcomplicated intra-abdominal infections in the United States,2007e2008: results of the Study for Monitoring AntimicrobialResistance Trends (SMART). Antimicrob Agents Chemother2010;54:3031e4.

123. Capp R, Chang Y, Brown DF. Effective Antibiotic TreatmentPrescribed by Emergency Physicians in Patients Admitted to theIntensive Care Unit with Severe Sepsis or Septic Shock: Where isthe Gap? J Emerg Med. Published Online First: 1 March 2011.PMID: 21371846.

124. Deviere J, Content J, Denys C, et al. Excessive in vitro bacteriallipopolysaccharide-induced production on monokines in cirrhosis.Hepatology 1990;11:628e34.

125. Wiest R, Das S, Cadelina G, et al. Bacterial translocation incirrhotic rats stimulates eNOS-derived NO production and impairsmesenteric vascular contractility. J Clin Invest 1999;104:1223e33.

126. Runyon BA. Monomicrobial nonneutrocytic bacterascites:a variant of spontaneous bacterial peritonitis. Hepatology1990;12:710e15.

127. Runyon BA, McHutchison JG, Antillon MR, et al. Short-courseversus long-course antibiotic treatment of spontaneous bacterialperitonitis. A randomized controlled study of 100 patients.Gastroenterology 1991;100:1737e42.

128. Runyon BA, Hoefs JC. Spontaneous vs secondary bacterialperitonitis. Differentiation by response of ascitic fluid neutrophilcount to antimicrobial therapy. Arch Intern Med1986;146:1563e5.

129. Fong TL, Akriviadis EA, Runyon BA, et al. Polymorphonuclear cellcount response and duration of antibiotic therapy in spontaneousbacterial peritonitis. Hepatology 1989;9:423e6.

130. Gines P, Rimola A, Planas R, et al. Norfloxacin preventsspontaneous bacterial peritonitis recurrence in cirrhosis: results ofa double-blind, placebo- controlled trial. Hepatology1990;12:716e24.

131. Terg R, Llano K, Cobas SM, et al. Effects of oral ciprofloxacin onaerobic gram-negative fecal flora in patients with cirrhosis:results of short- and long-term administration, with daily andweekly dosages. J Hepatol 1998;29:437e42.

132. Singh N, Gayowski T, Yu VL, et al. Trimethoprim-sulfamethoxazole for the prevention of spontaneous bacterialperitonitis in cirrhosis: a randomized trial. Ann Intern Med1995;122:595e8.

133. Nunnari G, Celesia BM, Bellissimo F, et al. Trimethoprim-sulfamethoxazole-associated severe hypoglycaemia: a sulfonylurea-like effect. Eur Rev Med Pharmacol Sci 2010;14:1015e18.

134. Wagenlehner FM, Schmiemann G, Hoyme U, et al. [National S3guideline on uncomplicated urinary tract infection:recommendations for treatment and management ofuncomplicated community-acquired bacterial urinary tractinfections in adult patients] (In German). Urologe A2011;50:153e69.

135. Fernandez J, Ruiz d A, Gomez C, et al. Norfloxacin vsceftriaxone in the prophylaxis of infections in patients withadvanced cirrhosis and hemorrhage. Gastroenterology2006;131:1049e56.

Gut 2012;61:297e310. doi:10.1136/gutjnl-2011-300779 309




136. Saab S, Hernandez JC, Chi AC, et al. Oral antibiotic prophylaxisreduces spontaneous bacterial peritonitis occurrence andimproves short-term survival in cirrhosis: a meta-analysis. Am JGastroenterol 2009;104:993e1001.

137. Loomba R, Wesley R, Bain A, et al. Role of fluoroquinolones inthe primary prophylaxis of spontaneous bacterial peritonitis:meta-analysis. Clin Gastroenterol Hepatol 2009;7:487e93.

138. Runyon BA. A pill a day can improve survival in patients withadvanced cirrhosis. Gastroenterology 2007;133:1029e31.

139. Soriano G, Guarner C, Teixido M, et al. Selective intestinaldecontamination prevents spontaneous bacterial peritonitis.Gastroenterology 1991;100:477e81.

140. Terg R, Fassio E, Guevara M, et al. Ciprofloxacin in primaryprophylaxis of spontaneous bacterial peritonitis: a randomized,placebo-controlled study. J Hepatol 2008;48:774e9.

141. Koo HL, Dupont HL. Rifaximin: a unique gastrointestinal-selectiveantibiotic for enteric diseases. Curr Opin Gastroenterol2010;26:17e25.

142. Alvarez RF, Mattos AA, Correa EB, et al. Trimethoprim-sulfamethoxazole versus norfloxacin in the prophylaxis ofspontaneous bacterial peritonitis in cirrhosis. Arq Gastroenterol2005;42:256e62.

143. Rolachon A, Cordier L, Bacq Y, et al. Ciprofloxacin andlong-term prevention of spontaneous bacterial peritonitis:results of a prospective controlled trial. Hepatology1995;22:1171e4.

144. Krag A, Wiest R, Gluud LL. Fluoroquinolones in the primaryprophylaxis of spontaneous bacterial peritonitis. Am JGastroenterol 2010;105:1444e5.

145. Fernandez J, Navasa M, Planas R, et al. Primary prophylaxis ofspontaneous bacterial peritonitis delays hepatorenal syndromeand improves survival in cirrhosis. Gastroenterology2007;133:818e24.

146. Grange JD, Roulot D, Pelletier G, et al. Norfloxacin primaryprophylaxis of bacterial infections in cirrhotic patients withascites: a double-blind randomized trial [see comments]. JHepatol 1998;29:430e6.

147. Dupont HL, Jiang ZD, Okhuysen PC, et al. A randomized, double-blind, placebo-controlled trial of rifaximin to prevent travelers’diarrhea. Ann Intern Med 2005;142:805e12.

148. Dupont HL, Jiang ZD. Influence of rifaximin treatment on thesusceptibility of intestinal Gram-negative flora and enterococci.Clin Microbiol Infect 2004;10:1009e11.

149. Darkoh C, Lichtenberger LM, Ajami N, et al. Bile acids improvethe antimicrobial effect of rifaximin. Antimicrob Agents Chemother2010;54:3618e24.

150. Garcia-Tsao G. Bacterial infections in cirrhosis: treatment andprophylaxis. J Hepatol 2005;42(Suppl 1):S85e92.

151. Riordan SM, Williams R. The intestinal flora and bacterialinfection in cirrhosis. J Hepatol 2006;45:744e57.

152. Chelarescu O, Chelarescu D, Ticovenau E, et al. Propranololadministration on post surgical infections in cirrhotic patients.J.Hepatol 2003;38:A173, (Abstract).

153. Perez-Paramo M, Munoz J, Albillos A, et al. Effect ofpropranolol on the factors promoting bacterial translocation incirrhotic rats with ascites. Hepatology 2000;31:43e8.

154. Serste T, Melot C, Francoz C, et al. Deleterious effects of beta-blockers on survival in patients with cirrhosis and refractoryascites. Hepatology 2010;52:1017e22.

155. Angeli P. beta-Blockers and refractory ascites in cirrhosis: themessage of a team of true scientists. J Hepatol 2011;55:743e4.

156. Serste T, Francoz C, Durand F, et al. Beta-blockers causeparacentesis-induced circulatory dysfunction in patients withcirrhosis and refractory ascites: a cross-over study. J Hepatol2011;55:794e9.

157. Zhang SC, Wang W, Ren WY, et al. Effect of cisapride onintestinal bacterial and endotoxin translocation in cirrhosis. WorldJ Gastroenterol 2003;9:534e8.

158. Tack J, van Outryve M, Beyens G, et al. Prucalopride (Resolor) inthe treatment of severe chronic constipation in patientsdissatisfied with laxatives. Gut 2009;58:357e65.

159. Lorenzo-Zuniga V, Bartoli R, Planas R, et al. Oral bile acidsreduce bacterial overgrowth, bacterial translocation, andendotoxemia in cirrhotic rats. Hepatology 2003;37:551e7.

160. Lorenzo-Zuniga V, Rodriguez-Ortigosa CM, Bartoli R, et al.Insulin-like growth factor-I improves intestinal barrier function incirrhotic rats. Gut 2006;55:1306e52.

161. Frances R, Chiva M, Sanchez E, et al. Bacterial translocation isdownregulated by anti-TNF-alpha monoclonal antibodyadministration in rats with cirrhosis and ascites. J Hepatol2007;46:797e803.

162. Adawi D, Kasrawi FB, Molin G, et al. Effect of lactobacillussupplementation with and without arginine on liver damage andbacterial translocation in an acute liver injury model in the rat.Hepatology 1997;25:642e7.

163. Wang XD, Soltesz V, Molin G, et al. The role of oral administrationof oatmeal fermented by Lactobacillus reuteri R2LC on bacterialtranslocation after acute liver failure induced by subtotal liverresection in the rat. Scand J Gastroenterol 1995;30:180e5.

164. Liu Q, Duan ZP, Ha DK, et al. Synbiotic modulation of gut flora:effect on minimal hepatic encephalopathy in patients withcirrhosis. Hepatology 2004;39:1441e9.

165. Rayes N, Seehofer D, Theruvath T, et al. Supply of pre- andprobiotics reduces bacterial infection rates after livertransplantationea randomized, double-blind trial. Am J Transplant2005;5:125e30.

PAGE fraction trail=13.75

310 Gut 2012;61:297e310. doi:10.1136/gutjnl-2011-300779




guidelines and beyondSpontaneous bacterial peritonitis: recent

R Wiest, A Krag and A Gerbes

doi: 10.1136/gutjnl-2011-3007792012 61: 297-310 originally published online December 6, 2011Gut

http://gut.bmj.com/content/61/2/297Updated information and services can be found at:

These include:

References #BIBLhttp://gut.bmj.com/content/61/2/297

This article cites 159 articles, 14 of which you can access for free at:

serviceEmail alerting

box at the top right corner of the online article. Receive free email alerts when new articles cite this article. Sign up in the

Errata

http://gut.bmj.com/content/61/4/636.full.pdf or: page

nextAn erratum has been published regarding this article. Please see

CollectionsTopic Articles on similar topics can be found in the following collections

(312)Cirrhosis (436)Hepatic cancer

(71)GUT Recent advances in clinical practice (56)Gut Education

Notes

http://group.bmj.com/group/rights-licensing/permissionsTo request permissions go to:

http://journals.bmj.com/cgi/reprintformTo order reprints go to:

http://group.bmj.com/subscribe/To subscribe to BMJ go to:


http://gut.bmj.com/content/61/2/297http://gut.bmj.com/content/61/2/297#BIBLhttp://gut.bmj.com/content/61/4/636.full.pdfhttp://gut.bmj.com//cgi/collection/gut_educationhttp://gut.bmj.com//cgi/collection/gut_recent_advances_in_clinical_practicehttp://gut.bmj.com//cgi/collection/hepatic_cancerhttp://gut.bmj.com//cgi/collection/cirrhosishttp://group.bmj.com/group/rights-licensing/permissionshttp://journals.bmj.com/cgi/reprintformhttp://group.bmj.com/subscribe/http://gut.bmj.com/http://group.bmj.com

in situ hybridisation, the authors foundbacteria deeply infiltrating the appendix.Fusobacteria (mainly Fusobacterium nucle-atum/necrophorum) were specific componentsof epithelial and submucosal infiltrates in62% of patients and were not found invarious controls. The presence of Fusobac-teria correlated positively with the severityof appendicitis. Conversely, main faecalmicrobiota including Faecalibacterium praus-nitzii groups were significantly decreasedwith an inverse relationship with theseverity of the disease.1

Altogether, these observations point to thepresence of a local appendiceal dysbiosiswith more bacteria with inflammatoryproperties and fewer bacteria with anti-inflammatory properties associated withacute appendicitis. The genus Fusobacteriumis characterised by high proteolytic activityand comprises different distinct species. Themost frequently encountered is F nucleatum,which is frequent in the oral sphere andimplicated in periodontitis. F necrophorum hasa high pathogenic potential and is implicatedin life-threatening infections such asLemierre’s syndrome. In cattle, it is found infootrot disease and is also frequent in liverabscesses. The third important species is Fvarium. All species are part of the normalintestinal microflora. By contrast, F praus-nitzii, which showed decreased numbers inappendicitis, is a bacterium with anti-inflammatory properties. Its numbers arealso reduced in patients with inflammatorybowel disease and it is associated withpostoperative recurrence of Crohn’s disease.2

Over 30 studies have now analysed theassociation between appendectomy andulcerative colitis (UC) and the majority ofthe studies support a highly significantinverse relationship.3 It is also well estab-lished that the protective effect of appen-dectomy depends on the inflammatoryconditions (appendicitis or lymphadenitis)that were the indication for appendectomyrather than on appendectomy itself.4 Theavailable data regarding whether or notappendectomy performed after the onset ofUC can modulate its clinical course arey stilllimited and conflicting and properlycontrolled trials are needed.5 Despite accu-mulating clinical evidence, the mechanismlinking appendicitis, appendectomy and UCremains elusive.

Interestingly, a link between Fusobacteriaand UC has been reported in several studies.In 2002, F varium was reported to be presentin the colonic mucosa of a high proportion(84%) of UC patients.6 Using immunoblot-ting with a F varium antigen Minami et alfound positive signals with sera from 45(40.2%) of 112 UC patients versus 20(15.6%) of 128 healthy controls (p

recent advances in clinical practice spontaneous bacterial peritonitis: recent ... · 2014. 11....

Documents