Spontaneous bacterial peritonitis in patients with cirrhosis undergoing selective intestinal decontamination

Journalof Hepatology1997;26: 88-95 Printed in Denmark Ail rights reserved Munksgaard Copenhagen

Copyright 0 European Association for the Study ofthe Liver I997 Journalof Hepatology ISSN 0168-8278

Spontaneous bacterial peritonitis in patients with cirrhosis undergoing selective intestinal decontamination A retrospective study of 229 spontaneous

bacterial peritonitis episodes

Josep M. Llovet’, Pilar Rodriguez-Iglesias’, Eduardo Moitinho’, Ram& Planas’, Ram6n Batalle2, Miquel Navasa*, Marga Menacho’, Albert Pardo’, Antoni Castells*, Eduard Cab&, Vicente Arroyo*, Miquel Angel Gassull’ and Joan Rod&* ‘Department of Gastroenterology, 2Hospital Germans Trias i Pujol, Badalona, and ‘Liver Unit, Hospital Clinic, Barcelona. Catalunya, Spain

Background/Aims: Selective intestinal decontamination with norfloxacin is widely used to prevent spontaneous bacterial infections in cirrhosis. The study was performed to compare the spontaneous bacterial peritonitis occurring in patients with and without prophylactic norfloxacin. Methods: Two hundred and twenty-nine consecutive episodes of spontaneous bacterial peritonitis, (193 in patients without (Group A) and 36 in patients with norfloxacin prophylaxis (Group B)), were retrospectively analyzed. In 100 episodes (86 and 14, respectively), the responsible organism was isolated in ascitic fluid. Results: Clinical and laboratory data at diagnosis were comparable in both groups. There were marked differences (p
polymicrobial episodes. Bacteria resistant to cefotaxime and gram-negative bacilli resistant to quinolones were isolated in ascitic fluid in nine (seven in Group A and two in Group B) and three episodes (all in Group A), respectively. No differences in the course of infection and patient survival were observed between groups. Concbsions: Spontaneous bacterial peritonitis in patients with and without prophylaxis with norfloxacin are not different in clinical features, response to treatment and prognosis. Spontaneous bacterial peritonitis caused by gram-negative organisms resistant to quinolones is extremely uncommon in patients with cirrhosis receiving prophylactic norfloxacin.

S

this infection (4-6) (primary prophylaxis) or SBP recurrence (7, 8) (secondary prophylaxis). SID is also effective in preventing bacterial infections in patients with cirrhosis with bleeding (9-l 1). It is well known that when SBP develops in patients submitted to SID, it is usually caused by gram-positive cocci (7). However, there has been no study specifically assessing the clinical course and prognosis of these patients. Therefore, the aim of the present study was to investigate the clinical, analytical and microbiological features, and the course of SBP, as well as the hospital survival in a large series of patients with cirrhosis, with and without SID with norfloxacin.

bacterial peritonitis (SBP) is a frequent complication of cirrhosis, with a prevalence ranging from 10% to 26% (l-3). Since enteric aerobic gram-negative bacteria are the most common causative organisms of SBP, selective intestinal decontamination (SID) with norfloxacin has been successfully used to prevent either SBP occurrence in patients with cirrhosis with a high risk of acquiring PONTANEOUS

Received 22 January; revised 13 May: accepted 20 May 1996

Correspondence: Ram6n Planas, M.D., Department of Gastroenterology, Hospital Universitari Germans Trias i Pujol, Carretera de1 Canyet s/n, 08916 Badalona, Spain. Tel.: 34-3-465 1385, Fax: 34-3-465 1385. 88

Key words: Cirrhosis; Gram-positive bacteria; Selective intestinal decontamination; Spontaneous bacterial peritonitis.

SBP in patients with cirrhosis on SID

Materials and Methods Two hundred and twenty-nine consecutive SBP episodes in 193 patients with cirrhosis (13 1 men, 62 women) diagnosed from September 1988 to July 1994 in two gastroenterological departments were analyzed. The mean age of the patients was 58 years (range, 24 to 89 years). The etiology of cirrhosis was alcoholic in 74 patients (38.3%), HBsAg or hepatitis C antibody (hepatitis C virus antibody assay was not available in 53 patients) associated in 67 (34.7%), both alcoholic and viral in 29 (15%), and miscellaneous or cryptogenic in 23 patients (12%). The diagnosis of cirrhosis was established by liver biopsy in 95 cases (49%), and by clinical, laboratory and exploratory findings, including ultrasonography, in the remaining patients. Active alcohol intake before the onset of SBP was present in 57 cases (25%). All patients had advanced liver cirrhosis: 55 patients (29%) belonged to the Child-Pugh class B, and 138 patients (71%) to the Child-Pugh class C. Most patients had a positive history of complications: 174 cases (90%) had had previous episodes of ascites, 65 (34%) of hepatic encephalopathy and 61 (32%) of gastrointestinal bleeding. Diagnosis of SBP was based on the presence of all the following criteria: 1) fever, abdominal pain and/ or hepatic encephalopathy; 2) ascitic fluid polymorphonuclear (PMN) count greater than 250 cells/mm3; and 3) absence of clinical, laboratory or radiological data suggesting secondary peritonitis (12). Samples of ascitic fluid were taken in the patient bedside for culture in blood-culture bottles before the start of antibiotic administration. A positive ascitic fluid culture was not considered necessary for the diagnosis, since culture-negative neutrocytic ascites is a wellrecognized variant of SBP (13, 14). The admission work-up for ascites in both hospitals includes diagnostic paracentesis in every patient. SBP was considered as community acquired when diagnostic criteria were present at admission (154 cases), and as hospital acquired when SBP developed during hospitalization in a patient with a previous normal ascitic fluid (75 cases). The diagnostic criteria and clinical types of SBP-induced renal impairment (SBP-RI) were defined according to the study by Follo et al. (15). Patients were empirically treated as soon as the diagnosis of SBP was established. Third-generation cephalosporines, mainly cefotaxime (iv, 2 g every 6 h), were administered in 190 cases. The remaining 39 SBP episodes were treated with fluoroquinolones (oral ofloxacin in 33 cases and oral or iv ciprofloxacin in the remaining 6 cases). The dosage of these antibiotics was adjusted to renal function

throughout the treatment period. Ascites was tapped every 48-72 h during antibiotic therapy to measure the ascitic PMN cell count. Resolution of SBP was considered when all clinical signs and symptoms of infection subsided and the ascitic fluid cell differential count showed less than 250 PMN/mm3. Antibiotic therapy was withdrawn 24 h after these criteria were fulfilled. All bacteria isolated in ascitic fluid were tested for susceptibility to the following antibiotics: third-generation cephalosporins (cefotaxime ceftazidime, ceftriaxone), amoxycillin-clavulanic acid and quinolones (ciprofloxacin). A microorganism was considered resistant to cefotaxime when the minimal inhibitory concentration was higher than 8ng/ml. Susceptibility to other antibiotics such as aminoglycosides (gentamycin, tobramycin, amikacin) or imipenem, vancomycin, aztreonam, erythromycin or rifampicin was tested when required. One hundred and ninety-three out of 229 SBP episodes (Group A) occurred in patients without SID. The remaining 36 (Group B) developed in patients with cirrhosis receiving nortloxacin either for preventing SBP recurrence (23 cases) or as prophylaxis of bacterial infections after upper gastrointestinal bleeding (13 cases). Norfloxacin was given in a dose of 400 mg/day for at least 7 days in patients with previous episodes of SBP, and of 800 mg/day for at least 3 days in bleeding patients with cirrhosis. SID was not considered for primary prophylaxis during hospitalization in patients with low ascitic fluid protein levels, No patient presenting SBP after gastrointestinal bleeding had previously received vasopressin or analogues for primary prophylaxis of early rebleeding. A total of 58 variables was recorded to compare SBP episodes in patients with and without SID (Table 1). Statistical methods Results are expressed as mean&SD. Groups were compared by means of the unpaired Student’s t-test for continuous variables with parametric distribution, the Mann-Whitney U-test for those with non-parametric distribution, and the x2 test for qualitative variables, applying the Yates’ correction when required. Differences were considered significant at the 0.05 level. All these analyses were performed with the BMDP statistical package (BMDP, Statistical software manual. Berkeley: University of California Press, 1985).

Results Characteristics of the patients Epidemiological, clinical and biochemical features in the two groups at the time of diagnosis of SBP are 89

J. M. Llovet et al. TABLE 1 Variables included in the analysis Epidemiological: age, sex. Clinical: active alcoholism, etiology of cirrhosis and time since diagnosis, history of gastrointestinal hemorrhage, hepatic encephalopathy, hepatocarcinoma, ascites and functional renal impairment, diuretic therapy, previous episodes of SBP, norfloxacin-induced SID, other antibiotic previous SBP, site of SBP acquisition (community or hospital acquired), cause of admission, G-I bleeding before SBP, abdominal pain, fever, encephalopathy, diarrhea, abdominal tenderness, ileus, shock, Child-Pugh score. Biochemical: WBC count (total, segmented and band neutrophils), prothrombin rate, MCV, serum albumin, total bilirubin, AST, ALT, GGT, alkaline phosphatase, serum cholesterol, serum creatinine, serum sodium and potassium, serum urea, ascitic fluid WBC count, ascitic fluid PMN cells/mm3 and %, AF total protein. Bacteriological: positive ascitic fluid culture, gram-negative gram-positive AF culture, positive blood culture.

or

Outcome: episodes with SBP-RI (transient, steady or progressive), antibiotic therapy and length, length of hospital stay, SBP resolution and hospital survival.

summarized in Table 2. Fever and abdominal pain were the most common symptoms, occurring in 154 (67%) and 132 (58%) cases, respectively. Hepatic encephalopathy at the time of diagnosis of SBP was present in 89 cases (38.8%), and was the sole symptom of SBP in 15 cases (6.5%). In five cases SBP was

symptomless. Patients of group B had a significantly lower serum sodium (128 rnmol/l vs. 13 1 mmol/l, p=O.Ol) and ascitic fluid protein concentrations (8.2 g/l vs. 11.6 g/l, p=O.O14). As expected, hospitalacquired SBP was most frequent in the group of patients submitted to SID, because it included patients developing SBP after gastrointestinal hemorrhage. The duration of norfloxacin prophylaxis was 7.16*4 days in patients with hemorrhage, and 114.4k118.78 in those patients in whom it was used as prevention of SBP occurrence. Microbiological data A positive ascitic fluid culture was obtained in 100 out of 207 episodes (48%). In 22 cases the amount of ascitic fluid obtained was insufficient to perform bacteriological analysis. In 20 out of the 107 patients with culture-negative neutrocytic ascites, a positive blood culture was obtained. Therefore, a bacteriological diagnosis was obtained in 58% (120/207) of the cases. Organisms isolated from ascites and blood cultures are shown in Table 3. In all groups of patients, aerobic gram-negative bacilli were responsible for 60% of culture-positive SBP episodes (those with positive ascitic fluid culture). The frequency of SBP-caused by gram-negative bacilli was higher in patients without than in

TABLE 2 Clinical and laboratory characteristics of the two groups obtained at the time of diagnosis of SBP Group A (n= 193) Mean age (years) Sex (M/F) Etiology (alcoholic/other) Time since diagnosis of cirrhosis (months) Site of SBP acquisition (community/hospital) Child-Pugh score Clinical findings Fever Abdominal pain Encephalopathy Blood analysis Leukocyte count (cell/mm3) PMN leukocytes (cell/mm3) Prothrombin rate (%) Serum albumin (g/l) Total bilirubin (mg/dl) Serum creatinine (mg/dl) Serum urea (mg/dl) Serum sodium (mmol/l) Ascitic fluid analysis Leukocyte count (celVmm3) PMN leukocytes (cell/mm3) Total protein (g/l) PMN leukocytes 72 h after starting the therapy

90

58.6fll 13U61 781115 51.2*49 137156 10.7f1.8 134 113 74 10631tiO85 8665ti426 48+16 27.4k4.1 6ti.5 1.3dsl.7 68f53 131+7 5095+6819 454of6491 11.6k7.1 1626zt2 898

Group B (n = 36) 58d9.1 26110 lU24 52.7+ 40 17/19 10.5k1.8 20 19 15

P NS NS NS NS 0.005 NS NS NS NS

10962k3766 9355f3685 52.5+17 26f4.3 6fi.8 1.4M.9 69zb43 128+7.1

NS NS NS NS NS NS NS 0.01

3 043?3 098 2642+3030 8.2f8.8 619f1008

NS NS 0.01 NS

SBP in patients with cirrhosis on SID

those with SID (67.4% vs. 14.3%, respectively, cocci were isolated in p=O.O002). G ram-positive 78.6% of culture-positive SBP episodes occurring in patients with SID, and in only 30.2% of the SBP episodes developing in patients without SID (p=0.0008). Almost all organisms isolated in the ascitic fluid were susceptible in vitro to cefotaxime. In fact, bacteria resistant to cefotaxime were only isolated in nine episodes of SBP (seven in group A and two in group B, Table 4). Table 4 also shows those gram-negative bacteria isolated in ascitic fluid resistant to quinolones. Among them, only three episodes of SBP, occurring in patients without SID, showed bacteria resistant

in vitro to quinolones (ciprofloxacin was the quinolone tested). Therefore, none of the 3 gram-negative bacteria isolated in patients submitted to SID showed in vitro resistance to quinolones. No superinfection during cefotaxime therapy was detected. Empirical antibiotic treatment had to be changed in 2 1 cases (Group A: 18, Group B: 3) due to treatment failure. In only five of these cases did the isolated bacteria show in vitro resistance to cefotaxime. Incidence of SBP-RI and prognosis The overall incidence of SBP-RI was 46.4% (106/ 229), with no differences between groups (Group A: 47% vs. Group B: 41%, n.s.). Regarding the type of

TABLE 3 Isolated bacteria from the ascitic fluid in culture-positive SBP, and blood cultures in culture-negative neutrocytic ascites Bacteria

Group A

Group B

P

Culture-negative neutrocytic ascites Culture-positive SBP Gram-negative bacilli E. coli Klebsiella pneumoniae Salmonella spp. Pseudomona aeruginosa Klebsiella oxytoca Citrobacterfreundii Enterobacter aerogenes Proteus mirabilis Morganella morganii Aeromona sobria Pasteurella pneumotropica

86 (50%) 86 (50%) 58 (67.4%) 41 4 3 2 2 1 1 1 1 1 1

21 (60%) 14 (40%) 2 (14.3%) 0 1 0 0 0 1 0 0 0 0 0

NS NS 0.0002

Gram-positive bacteria Streptococcus pneumoniae Enterococcus spp. Other streptococci Staphylococcus aureus Staphylococcus epidennidis Listeria monocytogenes

26 (30.2%) 11 1 10 3 0 1

11 (78.6%) 4 2 4 0 1 0

0.0008

ASCITIC FLUID

Polymicrobial

2* (2.3%)

’

1** (7.1%)

NS

BLOOD CULTURES Culture-positive Gram-negative bacteria E. coli Klebsiella pneumoniae Salmonella spp. Campylobacter jejuni Haemophilus infkenzae Gram-positive bacteria Streptococcus pneumoniae Enterococcus spp. Other streptococci Staphylococcus epidennidis

16 11 (68.7%) 7 1 1 1 1

4 3 (75%) 2 0 0 1 0

5 (31.3%) 2 0 2 1

1 (25%) 0 1 0 n

NS

NS

NS

* Klebsiella pneumoniae + Morganella morganii, E. coli + Enterococcus. ** E. coli + Enterococcus.

91

J. M. Llovet et al.

TABLE 4 Microorganisms resistant to cefotaxime and gram-negative bacteria resistant to quinolones isolated in ascitic fluid Bacteria

Group A

Group B

Microorganisms resistant to cefotaxime

(7186)

(204)

1 1 1

0 0 0

2

0

2

2

(3/58)

(O/2)

2

0

1

0

E. coli’ Enterobacter aerogenes= Morganella morganii3 Pseudomona aeruginosa4 Enterococcus ~pp.~

Gram-negative bacteria resistant to quinolones E. coli Klebsiella pneumoniae

Antibiotic susceptibility: ’ Ciprofloxacin, amikacin. ’ Ciprofloxacin, amikacin. 3 Ciprofloxacin, imipenem, aztreonam, amikacin. 4 Ceftazidime, piperacillin, ciprofloxacin, imipenem, aztreonam, amikacin. ’ Imipenem, vancomycin, erythromycin, rifampicin. TABLE 5 SBP resolution and hospital survival

Duration of antibiotic therapy (days) SBP resolution Length of hospital stay (days) Hospital survival

Group A (n= 193)

Group B (n = 36)

P

7.8 f 3.7 176 (91%) 21 f 13.4 143 (74%)

7.1 f 2.4 36 (100%) 22.2 f 14.8 29 (80%)

NS NS NS NS

SBP-RI, in 26% of the episodes it was transient, in 35% it was steady, and in 38% it was progressive. The incidence and types of SBP-RI were comparable in patients who did and did not receive norfloxacin prior to the infection. There were also no significant differences between groups regarding length of treatment (Group A: 7.8 f3.7 days, Group B: 7.kk2.4 days), duration of hospital stay (Group A: 2M3.4, Group B: 22.2f14.8 days), rate of SBP resolution and hospital survival (Table 5). No statistical differences regarding hospital survival were observed when SID patients were considered separately according to decontamination indications (GI bleeding or SBP recurrence). The overall hospital survival rate was 75% (74% in Group A and 80% in Group B). Fifty-seven patients died during hospitalization. In 17 cases (all from Group A) death occurred before SBP resolution, due to treatment failure (16 cases) or upper gastrointestinal hemorrhage (one case). Forty patients (17%) died after SBP resolution (33 in group A, seven in group B, p=N.S.). The main causes of death were: hepatorenal syndrome in 13 cases, severe infection acquired after SBP resolution in 10, liver failure in seven, hepatocarcinoma in five , gastroin92

testinal bleeding in four, and respiratory distress in one case. There were no significant differences between group A and B patients in relation to the causes of death.

Discussion The prognosis of patients with SBP has changed dramatically within the last two decades. The rate of infection resolution has increased from 50% to 92%, and the hospital mortality has decreased from almost 100% to less than 30% (l-3, 16, 17). These changes are attributed to various factors, including better recognition and early diagnosis of SBP (1,12), the use of third-generation cephalosporins (mainly cefotaxime), which are extremely effective against gramnegative bacteria and have no nephrotoxic effect, as empiric treatment of the infection (17), and an improvement in the management of the complications related to the underlying cirrhosis. Intestinal decontamination for prophylaxis of bacterial infection is another measure which has improved the management of patients with cirrhosis. It was introduced in 1985 by Rimola et al. (10) in patients with gastrointestinal hemorrhage. These

SBP in patients with cirrhosis on SID

authors demonstrated in a randomized trial that the oral administration of nonabsorbable antibiotics (gentamycin-vancomycin-nystatin or neomycin-colistin-nystatin) during and immediately after the bleeding episode significantly reduced the incidence of infection by gram-negative bacilli in these patients. The incidence of infections caused by gram-positive cocci, however, was not affected. Subsequently, Gin&s et al. (7) showed in a double-blind placebo controlled trial that long-term norlloxacin, a quinolone that induces SID (eradication of gram-negative bacilli preserving the anaerobic flora), was associated with a dramatic reduction in the probability of SBP recurrence (from 68% to 20% within the first year after resolution of the index SBP). This effect was due to the disappearance of SBP recurrences caused by gram-negative organisms. Again, in this study the incidence of SBP recurrences caused by gram-positive cocci was not modified. Subsequent studies have confirmed these results, and have shown that SID with norlloxacin is also effective in preventing bacterial infection and SBP due to gram-negative bacilli in patients with cirrhosis with gastrointestinal hemorrhage and in non-bleeding patients with cirrhosis with a high risk of developing SBP (cases with poor liver function and/or low ascitic fluid protein concentration) (4,9,11). Finally, a recent controlled trial has suggested that long-term administration of oral ciprofloxacin (750 mg per week) is also effective in the prevention of SBP (18). The clinical course and outcome of SBP arising in patients submitted to SID is a subject which has never been assessed. The extremely low number of patients developing SBP while on norfloxacin treatment probably accounts for this lack of information. In fact, we have been able to collect only 36 episodes of SBP in patients submitted to SID within a 6-year period. However, this is an important subject. First, the widespread use of norfloxacin for primary and secondary prophylaxis of SBP will increase the number of these patients. Second, the empiric antibiotic treatment that should be given to patients developing SBP while on SID may differ from that in patients not receiving norfIoxacin. Finally, the prevalence of quinolone-resistant gram-negative bacteria in the fecal flora among the general population and also in patients with cirrhosis under long-term SID with norfloxacin is increasing (19). Therefore, the development of SBP caused by gram-negative bacilli resistant to quinolones could be an emerging problem. The results of the current study indicate that SBP in patients on SID is clinically indistinguishable from

the SBP occurring in patients without SID. The clinical signs of infection, the leucocyte and PMN count in blood and in ascitic fluid, and the proportion of culture-positive and culture-negative SBP were comparable in the two groups. The clinical course of the patients, as estimated either by the incidence of SBP-RI or the episode and hospital mortality, was also similar in the two groups. Therefore, SID does not modify the natural course of SBP. Most SBP episodes (79%) in our patients with SID were caused by gram-positive cocci. In fact, in only three out of the 14 patients with culture-positive SBP was a gram-negative bacillus isolated. However, the percentage of patients responding to cefotaxime was very high. The empiric initial antibiotic treatment (mainly cefotaxime) had to be modified due to treatment failure in only 8.3% of patients on SID and in 9.3% of patients without SID. Moreover, only 9 isolated bacteria (7 out of the 86 isolated in the group without SID and 2 out the 14 in the SID group) were resistant in vitro to cefotaxime. Therefore, SID does not modify the initial empiric treatment of SBP Following the study of Gin& et al. (7), published in 1990 and performed in our services, we routinely use SID with norfloxacin in all patients admitted with gastrointestinal hemorrhage (for 7 days, beginning immediately after emergency endoscopy) and in those recovering from an episode of SBP (long-term SID until liver transplantation or death). Since the current study is a retrospective analysis, we did not attempt to estimate the number of patients submitted to SID during recent years. However, it is surprising that no case of SBP caused by a gram-negative organism resistant to quinolones was observed in patients submitted to SID. Two possibilities may arise to explain this outstanding feature. The first is that the development of gram-negative organisms resistant to quinolones in the fecal flora in patients submitted to SID may not be as common as recently reported (19). In fact, in the study of Gin& et al. (7) in six patients on SID followed up with monthly quantitative analysis of the fecal flora for 1 year, organisms resistant to quinolones were isolated at low concentrations in only 3 out of the 82 specimens obtained. A second possibility is that norfIoxacin interferes the bacterial translocation (BT) process from the intestinal lumen to the submucosal lymphatics, which is thought to be the initial step of the spontaneous infections (SBP and spontaneous bacteremia) in cirrhosis (20-23). Bacterial translocation is a complex process involving the interaction of bacterial adhesins to specific receptors in the enterocytes, followed by 93

J.

M. Llovet et al.

endocytosis of the organisms (24, 25). Once in the submucosa, bacteria reach the intestinal lymphatics, the thoracic duct and the systemic circulation. Recent studies have implicated BT in the pathogenic hypothesis of SBP (20, 21), and have observed that ascitic cirrhotic rats on norfloxacin-induced SID showed a decreased incidence of gram-negative bacterial translocation (22,23), a phenomenon that could account for the significantly lower incidence of SBP due to these organisms observed in patients on SID. Several mechanisms that have proved to favor BT in experimental models, such as endotoxemia (26), histological changes in the intestinal mucosa secondary to portal hypertension (27), malnutrition (28) and intestinal bacterial overgrowth (29), could explain why patients with cirrhosis and ascites are predisposed to bacterial translocation. In this respect, the recent experimental study of Guarner et al. (30), showing that bacterial translocation of a specific bacterium is almost always associated with intestinal bacterial overgrowth of that organism,is of special relevance. It is possible that, although patients submitted to SID may develop gram-negative organisms resistant to quinolones in the fecal flora, their concentration is insufficient to translocate. In summary, the results of the present study indicate that SBP developing in patients with and without SID are not different in clinical features, response to treatment and prognosis. SBP during norlloxacininduced SID is usually caused by grampositive cocci. SBP caused by gram-negative organisms resistant to quinolones is extremely uncommon in patients with cirrhosis submitted to SID.

Acknowledgements This work was presented in part at the Annual Congress of the American Gastroenterological Association in San Diego, California on May 14-17, 1995.

References 1. Garcia-Tsao G. Spontaneous bacterial peritonitis. Gastroenterol Clin North Am 1992; 21: 257-75. 2. Almdal TP, Skinhoj P. Spontaneous bacterial peritonitis in cirrhosis. Incidence, diagnosis, and prognosis. Stand J Gastroenterol 1987; 22: 295-300. 3. Llovet JM, Planas R, Morillas R, Quer JC, Cab& E, Boix J, Humbert P, Guilera M, Domenech E, Bert& X, Gassull MA. Short-term prognosis of cirrhotics with spontaneous bacterial peritonitis: multivariate study. Am J Gastroenterol 1993; 88: 388-92. 4. Soriano G, Guarner C, Teixid6 M, Such J, Barrios J, Enrfquez J, Vilardell J. Selective intestinal decontamination prevents spontaneous bacterial peritonitis. Gastroenterology 1991; 100: 447-81.

94

5. Andreu M, Sola R, Sitges-Serra A, Alia C, Gallen M, Vila C, Co11 S, Oliver M. Risk factors for spontaneous bacterial peritonitis in cirrhotic patients with ascites. Gastroenterology 1993; 104: 1133-8. 6. Llach J, Rimola A, Navasa M, Gines P, Salmer6n JM, Gines A, Arroyo V, Rod& J. Incidence and predictive factors of first episode of spontaneous bacterial peritonitis in cirrhosis with ascites: relevance of ascitic fluid concentration. Hepatology 1992; 16: 724-727. 7. Gin&s P, Rimola A, Planas R, Vargas V, Marco F, Almela M, For& M, Miranda ML, Llach J, Salmer6n JM, Esteve M, Marques JM, Jimenez de Anta MT, Arroyo V, Rod& J. Norfloxacin prevents spontaneous bacterial peritonitis recurrence in cirrhosis: results of a double-blind, placebo-controlled trial. Hepatology 1990; 12: 716-24. 8. Tit6 LI, Rimola A, Gin& P, Llach J, Arroyo V, Rod& J. Recurrence of spontaneous bacterial peritonitis in cirrhosis: frequency and predictive factors. Hepatology 1988; 8: 27-3 1. 9. Soriano G, Guamer C, Tomas A, Villanueva C, Torras X, Gonzalez D, Sainz S, Anguera A, Cuss6 X, Balanzo , Vilardell F. Notlloxacin prevents bacterial infection in cirrhotits with gastrointestinal hemorrhage. Gastroenterology 1992; 103: 1267-72. 10. Rimola A, Bory F, Teres J, Perez-Ayuso RM, Arroyo V, Rodts J. Oral nonabsorbable antibiotics prevent infection in cirrhotics with gastrointestinal hemorrhage. Hepatology 1985; 5: 463-67. 11. Blaise M, Pateron D, Trinchet JC, Levacher S, Beaugrand M, Pourriat J. Systemic antibiotic therapy prevents bacterial infection in cirrhotic patients with gastrointestinal hemorrhage. Hepatology 1994; 20: 34-38. 12. Hoefs JC, Runyon BA. Spontaneous bacterial peritonitis. DM-Dis Mon 1985; 31: l-48. 13. Runyon BA, Hoefs JC. Culture-negative neutrocytic ascites: A variant of spontaneous bacterial peritonitis. Hepatology 1984; 4: 1209-11. 14. Pelletier G, Salmon D, Ink 0, Hannoun S, Attali P, Buffet C, Etienne JP. Culture-negative neutrocytic ascites: a less severe variant of spontaneous bacterial peritonitis. J Hepatol 1990; 10: 327-31. 15. Follo A, Llovet JM, Navasa M, Planas R, Foms X, Francitorra A, Rimola A, Gassull MA, Arroyo V, Rod& J. Renal impairment after spontaneous bacterial peritonitis in cirrhosis: incidence, clinical course, predictive factors and prognosis. Hepatology 1994; 20: 1495-501. 16. Toledo C, Salmer6n JM, Rimola A, Navasa M, Arroyo V, Llach J, Gines A, Gines P, Rod& J. Spontaneous bacterial peritonitis in cirrhosis: predictive factors of infection resolution and survival in patients treated with cefotaxime. Hepatology 1993; 17: 251-7. 17. Felisart J, Rimola A, Arroyo V, Perez-Ayuso RM, Quintero E, Gin& P, Rod& J. Cefotaxime is more effective than ampicillin-tobramycin in cirrhotics with severe infections. Hepatology 1985; 5: 457-62. 18. Rolachon A, Cordier L, Bacq Y, Nousbaum JP, Franza A, Paris JC, Fratte S, Bohn B, Kitmacher P, Stahl JP, Zarski JP Ciprofloxacin and long-term prevention of spontaneous bacterial peritonitis: results of a prospective controlled trial. Hepatology 1995; 22: 1171-4. 19. Dupeyron C, Mangeney N, Sedrati L, Campillo B, Foulet P, Leluan G. Rapid emergence of quinolone resistance in cirrhotic patients treated with nortloxacin to prevent spontane-

SBP in patients with cirrhosis on SID

20.

21.

22.

23.

24.

ous bacterial peritonitis. Antimicrob Agents Chemother 1994; 38: 340-4. Llovet JM, Bartoli R, Planas R, Cabre E, Jimenez M, Urban A, Ojanguren I, Arnal J, Gassull MA. Bacterial translocation in cirrhotic rats. Its role on the development of spontaneous bacterial peritonitis. Gut 1994; 35: 1648-52. Runyon B, Squier S, Borzio M. Translocation of gut bacteria in rats with cirrhosis to mesenteric lymph nodes partially explains the pathogenesis of spontaneous bacterial peritonitis. J Hepatol 1994; 21: 792-6. Llovet JM, Bartoli R, Planas R, Vifiado B, Amal J, Ojanguren I, Ausina V, Gassull MA. Selective intestinal decontamination with norlloxacin reduces bacterial translocation in ascitic cirrhotic rats exposed to haemorrhagic shock. Hepatology 1996; (in press). Runyon B, Borzio M, Young S, Squier SU, Guarner C, Runyon M. Effect of selective bowel decontamination with norlloxacin on spontaneous bacterial peritonitis, translocation, and survival in an animal model of cirrhosis. Hepatology 1995; 21: 1719-24. Berg RD, Garlington AW. Translocation of certain indigenous

25.

26.

27.

28. 29.

30.

bacteria from the gastrointestinal tract to the mesenteric lymph nodes and other organs in gnotobiotic mouse model. Infect Immun 1979; 23: 403-l 1. Wells C, Maddaus M, Simmons R. Proposed mechanisms for the translocation of intestinal bacteria. Rev Infect Dis 1988; 5: 958-79. Deitch E, Berg R, Specian R. Endotoxin promotes the translocation of bacteria from the gut. Arch Surg 1987; 122: 1859. Garcia-Tsao G, Albillos A, Barden GE, West B. Bacterial translocation in acute and chronic portal hypertension. Hepatology 1993;17: 1081-5. Deitch EA. Bacterial translocation: the influence of dietary variables. Gut 1994; Suppl 1: S23-7. Steffen E, Berg R. Relationship between caecal population levels of indigenous bacteria and translocation to mesenteric lymph nodes. Infect Immun 1983; 39: 1252-9. Guamer C, Runyon BA, Young S, Heck M, Sheikh MY. Intestinal bacterial overgrowth and bacterial translocation in an experimental model of cirrhosis in rats. Hepatology 1995; 22: 238 (A).

95