Daily Norfloxacin Is More Effective thanWeekly Rufloxacin in Prevention of

Spontaneous Bacterial PeritonitisRecurrence

TILMAN M. BAUER, MD,* ANTONIO FOLLO, MD,* MIGUEL NAVASA, MD,* JORDI VILA, MD,†RAMON PLANAS, MD,‡ GERARDO CLEMENTE, MD,§ VICTOR VARGAS, MD,¶

FELIPE BORY, MD,** PERE VAQUER, MD,†† and JUAN RODES, MD, FRCP*

Our aim was to compare weekly rufloxacin with daily norfloxacin in the secondary prophylaxisof spontaneous bacterial peritonitis and to examine changes in antibiotic susceptibility in fecalEschericia coli. The method used was an open randomized clinical trial including 79 patientswho received either norfloxacin 400 mg/day or rufloxacin 400 mg/week and followed up forone year. E. coli counts, quinolone susceptibility, and drug concentrations in feces wereinvestigated in 12 patients. Cumulative one-year probability of peritonitis recurrence was26% for patients on norfloxacin and 36% for those on rufloxacin (P � 0.16). Norfloxacin wasmore effective in the prevention of peritonitis recurrence due to Enterobacteriaceae (0% vs22%, P � .01). At the end of follow-up, all 12 patients had E. coli resistant to quinolones intheir feces. In conclusion, weekly rufloxacin is not an alternative to daily norfloxacin in theprevention of peritonitis recurrence. The development of quinolone-resistant E. coli in fecesmay be an important problem in patients on long-term quinolone prophylaxis.

KEY WORDS: spontaneous bacterial peritonitis; norfloxacin; rufloxacin; quinolone resistant E. coli; secondaryprophylaxis.

Patients with liver cirrhosis recovering from an epi-sode of spontaneous bacterial peritonitis (SBP) havean indication for selective intestinal decontamination(SID) (1, 2). Quinolones are the antibiotics of choice

for the secondary prophylaxis of SBP. They are activeagainst Enterobacteriaceae without affecting the an-aerobic flora (3), an effect referred to as SID. Nor-floxacin at 400 mg/day was the first antibiotic shownto reduce the one-year probability of SBP recurrenceby 40% (1) and is currently widely used (4). However,other antibiotics such as trimethoprim–sulfamethox-azol (5) or ciprofloxacin 750 mg once a week (6) havealso been shown to be effective.

Rufloxacin is a fluoroquinolone-derivative with anantimicrobial spectrum similar to norfloxacin but alonger half-life due to enterohepatic circulation (7).The latter may provide increased small intestinal drugconcentrations over extended periods of time, thusrepresenting an advantage over other antibiotics. In

Manuscript received May 6, 2001; revised manuscript receivedSeptember 28, 2001; accepted November 12, 2001.

From the *Liver Unit, IMD, and †Department of Microbiology,Hospital Clinic, IDIBAPS, University of Barcelona; ‡HospitalTrias i Pujol, Badalona; §Hospital Gregorio Maranon, Madrid;¶Hospital de la Vall d’Hebron, Barcelona; **Hospital del Mar,Barcelona; and ††Hospital Son Dureta, Palma de Mallorca; Spain.

This study was supported in part with a Grant from the Fondo deInvestigaciones Sanitarias de la Seguridad Social (FISS 2088/97).T.M. Bauer is a grantholder of a Marie Curie Training MobilityResearch Grant of the European Commission.

Address for reprint requests: Dr. M. Navasa, Liver Unit, Hospi-tal Clinic, Villarroel 170, 08036. Barcelona, Spain.

Digestive Diseases and Sciences, Vol. 47, No. 6 (June 2002), pp. 1356–1361 (© 2002)

1356 Digestive Diseases and Sciences, Vol. 47, No. 6 (June 2002)0163-2116/02/0600-1356/0 © 2002 Plenum Publishing Corporation

fact, a single oral dose of 400 mg rufloxacin reducesfecal colony counts of E. coli for up to one week (8).These data suggest that rufloxacin at 400 mg once aweek might be an effective and comfortable alterna-tive to norfloxacin 400 mg once a day in the preven-tion of SBP recurrence.

An important problem in SBP prophylaxis in cir-rhotic patients undergoing SID is the emergence ofinfections caused by quinolone-resistant bacteria (9,10). Such infections have also been reported in neu-tropenic patients (11) and a mixed patient population(12) and were shown to be related to previousadministration of quinolones. Although some studieson long-term SID in cirrhotic patients with nor-floxacin (1, 13), ciprofloxacin (6), or trimethoprim-sulfamethoxazole (5) did not report the appearanceof bacteria resistant to these antibiotics, prolongedantibiotic prophylaxis is likely to result in the emer-gence and selection of antibiotic-resistant bacteria.

These data prompted us to perform a prospective,randomized study to compare the clinical efficacy ofrufloxacin, 400 mg once a week, with norfloxacin, 400mg once a day, in the prevention of SBP recurrence incirrhotic patients with ascites and to monitor thedevelopment of quinolone-resistant bacteria in thefecal flora of these patients.

MATERIALS AND METHODS

Selection of Patients. The study was performed in sixuniversity hospitals. All hospitalized patients with liver cir-rhosis diagnosed by liver biopsy and/or clinical, laboratory,and imaging studies recovering from an episode of SBPwere screened over a period of two years. SBP was diag-nosed on the basis of �250/mm3 polymorphonuclear leu-kocytes in the ascitic fluid without any clinical, radiological,or laboratory data suggesting secondary peritonitis or anyother abdominal pathology such as hemorrhage, pancreati-tis, peritoneal tuberculosis, or carcinomatosis. Positive as-citic fluid cultures were not necessary for the diagnosis ofSBP (14). Exclusion criteria included advanced liver failure(serum bilirubin �10 mg/dl, prothrombin activity �25%, orchronic hepatic encephalopathy grade �III), moderate orsevere renal failure (serum creatinine �2 mg/dl), presenceof hepatocellular carcinoma, known or suspected hypersen-sitivity to quinolones, and SID with norfloxacin prior to theindex episode of SBP. The trial was approved by the inves-tigation and ethics committees of each hospital involved,and all the patients gave informed written consent.

Treatment Protocol. Patients were allocated into thestudy groups by blocked randomization in blocks of four.Study medication was started one week after withdrawal ofthe antibiotic treatment of the SBP index episode. Group Rreceived 400 mg rufloxacin per os on days 1 and 4 and 7 ofthe first week and at weekly intervals thereafter. The ratio-nale for this dosing of rufloxacin was to assure a SID duringthe first week, followed by the weekly dose to maintain SID

as previously shown in a previous study (8). Group N wasgiven 400 mg norfloxacin per os daily. Patients were fol-lowed up in the out-patient clinic monthly during the firstsix months and then bimonthly until the completion of theone-year follow-up period. Medical history, physical exam-ination, and routine laboratory studies including liver andrenal function tests, complete differential blood count andurine analysis with sediment were performed before theinitiation of treatment and at each follow-up visit. Asciticfluid polymorphonuclear cell count and culture were per-formed before initiation of treatment and when SBP wassuspected. Medication was provided during the visits, andcompliance was assessed by pill counts.

The end point was the occurrence of SBP as definedabove. Patients were censored for any of the followingevents: liver transplantation, death, serious adverse events,loss to follow-up. Patients were considered noncompliantwhen they failed to take 25% or more of the pills during thewhole follow-up period or when they failed to attend two ormore follow-up visits. Patients admitted during the studyperiod for acute upper gastrointestinal hemorrhage re-ceived norfloxacin (400 mg/12 hr) for seven days, irrespec-tive of the study group. Study medication was restarted onday 8 of the bleeding episode.

Analysis of Fecal Flora and Drug Concentrations. In asubset of six patients in each treatment group, stool sampleswere obtained before therapy, at weekly intervals during thefirst month, and after three, six, and 10 months. Sampleswere collected in sterile containers and processed for quan-titative enumeration of E. coli by a method described else-where (15). Two colonies of E. coli were subjected tosusceptibility testing and analysis by repetitive extragenicpalindromic-PCR (REP-PCR). Susceptibility testing wasperformed by a microdilution method using cation-adjustedMueller Hinton broth in accordance with the guidelines ofthe National Committee for Clinical Laboratory Standards.Strains were considered resistant when the minimum inhib-itory concentration (MIC) exceeded 32 �g/ml for nalidixicacid or 1 �g/ml for ciprofloxacin. REP-PCR was used toscreen fecal E. coli isolates and was performed as describedelsewhere (16). Strains with identical patterns were exam-ined by low-frequency restriction analysis and pulsed fieldgel electrophoresis as described previously (17). For ampli-fication of the quinolone resistance-determining region ofthe gyrA and parC genes, PCR was performed as describedelsewhere (16, 18). The sample was processed for DNAsequencing with a TaqDyeDeoxyTerminator Cycle Se-quencing Kit (Amersham) and analyzed in an automaticDNA sequencer (Applied Biosystems 373A). Fecal drugconcentrations were determined by a bioassay as describedpreviously (19). Klebsiella pneumoniae ATCC 10031 wasused as the indicator strain. Reproducibility was 2.9%, 6.5%and 4.0% for rufloxacin and 2.2%, 7.1% and 4.7% fornorfloxacin at concentrations of 1, 16, and 128 �g/ml, re-spectively. Lower detection limits for rufloxacin and nor-floxacin were 1 and 0.5 �g/ml, respectively.

Statistical Analysis. The sample size was calculated onthe basis of a previous study on the cumulative probabilityof SBP recurrence in patients receiving norfloxacin of 20%(1), with the assumption that the null hypothesis should beretained if success rates of two different antibiotic regimens(in this case the cumulative one year probability of SBP

RUFLOXACIN IN SBP PROPHYLAXIS

1357Digestive Diseases and Sciences, Vol. 47, No. 6 (June 2002)

recurrence) differ by less than 15% (20), an anticipateddropout rate of 15%, and on the basis of type I and type IIerrors �0.05 (21). The estimated sample sized resulted in40 patients per group of treatment. The probability of SBPrecurrence was calculated according to the Kaplan-Meiermethod considering all patients included in the study andexcluding noncompliant patients. Curves were comparedusing the log-rank test; P � 0.05 was considered significant.The primary end point was the probability of a one-yearrecurrence of SBP. Comparisons between the study groupswere performed using the Student’s t test and the �2 test,respectively. Quantitative variables are expressed asmean � SEM.

RESULTS

Patient Characteristics. Two hundred forty-threepatients with liver cirrhosis who had recovered froman episode of SBP were evaluated. Reasons for ex-clusion were hepatocellular carcinoma (54 cases), ter-minal liver failure (32 cases), renal failure (22 cases),previous SID (40 cases), and refusal to participate orother reasons (16 cases). Thus, 79 patients were en-rolled and assigned to receive either norfloxacin(group N, 40 patients) or rufloxacin (group R, 39patients). There were no significant differences be-tween the two groups in clinical and laboratory dataat the time of enrollment and in the microbiologicalcharacteristics of the index SBP (Table 1).

Follow-Up and Censored Patients. The mean fol-

low-up period was 201 � 22 days in group N and208 � 23 days in group R (P � 0.81). A total of 17patients from group N and 11 from group R werecensored. Three patients in each group withdrewfrom the study during the first two months. Threepatients from group N and five from group R diedduring follow-up (P � 0.34), the causes being termi-nal liver failure (two and four cases in groups N andR, respectively) and upper gastrointestinal hemor-rhage (one case in group R). Drug administration wasstopped because of side effects in three patients fromgroup N (severe diarrhea, granulocytopenia, urti-caria) and three patients from group R (nausea andvomiting, headache). All side effects disappeared ondiscontinuation of the drugs. One patient in group Nand two patients in group R were considered non-compliant. Eight patients in group N vs zero patientsin group R were submitted to liver transplantation(P � 0.005). This difference could be partially ex-plained for the higher proportion of patients olderthan 65 years in group R, which was at the time of thestudy the upper age limit for liver transplantation.

Recurrence of SBP. SBP recurred in six patientsfrom group N (15%) and 12 patients from group R(31%, P � 0.09). Figure 1a shows the cumulative prob-ability of SBP recurrence in the two groups according tointention-to-treat analysis. The one-year probability ofSBP recurrence did not differ significantly betweengroups N and R (26% vs 36%, respectively, P � 0.16).There was a trend towards earlier recurrence in group R(mean time between start of prophylaxis and SBP re-currence in group N and R, 180 � 34 vs 93 � 25 days,respectively; P � 0.06), and the three-month probabilityof SBP recurrence differed significantly (3% vs 19% ingroups N and R, respectively; P � 0.03). When non-compliant patients were excluded from the analysis, theincidence of SBP recurrence was 15% in the norfloxacingroup vs 27% in the rufloxacin group (P � 0.21) whilethe one-year probability of SBP recurrence was, respec-tively, 26% vs 32% (P � 0.31).

The two groups did not differ significantly withrespect to ascitic fluid polymorphonuclear cell countat the time of SBP recurrence (6490 � 4970/mm3,range 340–31,065 in group N, and 4760 � 1990/mm3,range 480–23,680, in group R, P � 0.71) or to theproportion of culture-positive episodes (1 of 6 vs 7 of12 episodes, respectively, P � 0.15). Enterobacteri-aceae were cultured in the seven SBP recurrences ingroup R (E. coli, five cases; Klebsiella pneumoniae andAlcaligenes spp., one case each) but were not found inany patient from group N (58% vs 0%, P � 0.03).Consequently, norfloxacin was significantly more ef-

TABLE 1. CLINICAL AND LABORATORY DATA AT TIME OFINCLUSION*

Group N(N � 40)

Group R(N � 39)

Mean age (yr) 59 � 2 62 � 2Patients aged � 65 yr 75% 56%Male 26/14 29/10Etiology of cirrhosis

Alcohol abuse 11 (28%) 13 (33%)Chronic viral hepatitis B or C 27 (67%) 24 (62%)

Previous history ofAscites 36 (90%) 32 (82%)Hepatic encephalopathy 15 (38%) 14 (36%)Gastrointestinal bleeding 11 (28%) 12 (31%)

Serum bilirubin (mg/dl) 2.9 � 0.3 3.0 � 0.4Serum albumin (g/liter) 27 � 0.9 28 � 0.9Prothrombin activity (%) 58 � 3 57 � 2Serum creatinine (mg/dl) 1.1 � 0.1 1.0 � 0.1Child-Pugh score 10.4 � 0.1 10.9 � 0.1Organism responsible for the

index SBP:Gram-negative bacilli 11 11Other bacteria 6 7Culture-negative 23 21

*Group N, norfloxacin 400 mg/day; group R, rufloxacin 400 mg/72hr for one week, followed by 400 mg/week; SBP, spontaneousbacterial peritonitis. Quantitative variables are expressed asmean � SEM. No significant difference was found in any of thevariables analyzed.

BAUER ET AL

1358 Digestive Diseases and Sciences, Vol. 47, No. 6 (June 2002)

fective in preventing SBP recurrence due to Enter-obacteriaceae (0% vs 22%, P � 0.01) (Figure 1b).The only culture-positive episode in group N was dueto Streptococcus agalactiae. Of the seven gram-negative isolates found in group R, five were sensitiveto nalidixic acid and ciprofloxacin, but two (K. pneu-moniae and E. coli) were resistant to both quinolones.

Course of Disease During Follow-Up. Infectionsother than SBP occurred in seven patients from groupN (17%) and six patients from group R (15%, P �0.96). There were eight urinary tract infections, fourlower respiratory tract infections, and one case eachof dental abscess, pyelonephritis, and bacteremia. Atotal of 37 and 41 readmissions occurred in 33 and 29patients from group N and R, respectively. Causeswere ascites (six admissions in either group), gastro-intestinal hemorrhage (seven and nine admissions ingroups N and R, respectively), hepatic encephalopa-thy (10 and 12 admissions), SBP recurrence (five andeight cases), and various causes (nine and six admis-sions). One SBP recurrence in group N and fourepisodes in group R occurred nosocomially duringhospitalizations not initially related to SBP. None ofthe SBP episodes developed in the course of an acuteupper gastrointestinal hemorrhage. The number of

days spent in hospital did not significantly differ(11.1 � 1.8 vs 10.4 � 1.3 days; P � 0.77).

Fecal Drug Concentrations and Analysis of FecalFlora. Fecal drug concentrations at various timesduring follow-up are shown in Table 2. Norfloxacinachieved and maintained high concentrationsthroughout the study, whereas fecal concentrations ofrufloxacin were lower and decreased during the firstthree weeks, remaining steady thereafter. Figure 2shows the colony counts for E. coli obtained fromquantitative stool cultures performed in six patientsfrom each group. A decrease of colony counts by atleast 2 log units with respect to the initial culture wasobserved in four patients in group N but only onepatient in group R (P � 0.24). There was no com-plete eradication of E. coli in either group. Twopatients had nalidixic acid- and ciprofloxacin-resistantE. coli in their feces at the start of the study, neitherhaving previously received quinolones. In all of theremaining 10 patients, strains of E. coli showing re-sistance to nalidixic acid were recovered during fol-low-up. Six of these strains were resistant to cipro-floxacin, and two, strains 4.4 and 12.3, were shown byREP-PCR and low frequency restriction analysis, tobe identical to strains that were fully susceptible atthe start (strains 4.0 and 12.0, respectively, Figure 3).They were recovered after three and four weeks of

Fig 1. Kaplan Meier-curve showing the cumulative probability forrecurrence of spontaneous bacterial peritonitis in patients on se-lective intestinal decontamination with norfloxacin (group N, dot-ted line) and rufloxacin (group R, shaded line): (a) all episodes ofspontaneous bacterial peritonitis; (b) episodes culture-positive forgram-negative rods. Comparison by log-rank test.

TABLE 2. FECAL DRUG CONCENTRATIONS IN PATIENTS ONPROPHYLAXIS WITH NORFLOXACIN (GROUP N) AND RUFLOXACIN

(GROUP R)

Time after start

Drug conc �g/g, mean � SEM

Group N Group R

1 week 804 � 310 209 � 842 weeks 446 � 104 122 � 473 weeks 815 � 303 33 � 121 month 733 � 234 39 � 193 months 482 � 187 28 � 166 months 630 � 139 35 � 610 months 1146 � 574 30 � 11

Fig 2. Fecal Escherichia coli colony counts in patients on selectiveintestinal decontamination with norfloxacin (N � 6) and rufloxa-cin (N � 6). CFU: colony forming units.

RUFLOXACIN IN SBP PROPHYLAXIS

1359Digestive Diseases and Sciences, Vol. 47, No. 6 (June 2002)

prophylaxis, respectively. Analysis of their quinolone-resistance determining regions showed that both hadacquired a point mutation at Ser-83 of the gyrA gene,generating a substitution of this amino acid by Leu.Strain 4.4 also showed a mutation at amino acidcodons Asp-87 of the gyrA gene and Glu-84 of theparC gene. None of these mutations were present instrains 4.0 or 12.0.

DISCUSSION

Secondary prophylaxis of SBP with daily norfloxa-cin in patients with cirrhosis is widely accepted, costeffective, and may improve the survival of patientsawaiting liver transplantation (4). Encouraged by pre-vious microbiological studies, we postulated that onceweekly rufloxacin, a fluoroquinolone derivative with amarkedly prolonged half-life, might be as effective asnorfloxacin in SBP secondary prophylaxis at a lowercost. The results of the current study show that al-though both regimens did not significantly differ inthe one-year probability of SBP recurrence, norfloxa-cin was more effective in preventing SBP due togram-negative bacteria and recurrence occurring dur-ing the first three months of treatment, when the riskis highest (2). In addition, although a potential cost

savings and increased patient comfort associated witha once-weekly prophylactic regimen was initially ex-pected, once rufloxacin was commercialized (only inItaly), the cost per week of treatment was almostidentical for both antibiotics (cost/week/patient: €6.88for rufloxacin and €7.23 for norfloxacin). Therefore,the data presented in this study do not justify achange in the current practice. In the absence offurther data, norfloxacin at 400 mg/day should be main-tained as the standard regimen for secondary preven-tion of SBP in patients with cirrhosis and ascites.

Fecal drug concentrations as measured in a subsetof patients were markedly different between treat-ment groups, reflecting the difference in total drugconsumption between groups N and R. This couldexplain in part the lower efficacy of rufloxacin in theprevention of SBP recurrence by Enterobacteriaceae.However, it is interesting to note that despite highdrug concentrations found in patients on norfloxacin,exceeding the MIC of most fecal organisms, thisantibiotic failed to eliminate Enterobacteriaceaefrom the feces. Fecal drug concentrations of quino-lones may not correlate with their efficacy in obtain-ing SID because quinolones are bound to fecal fibersand bacteria and may therefore be pharmacologicallyinactive (22). In addition, and in contrast to earlierseries (1, 13, 15, 23) we observed the emergence ofresistant fecal enterobacteriaceal strains during pro-phylaxis with quinolones to be a common event. Twopatients had E. coli strains resistant to nalidixic acid atbaseline, and all other patients were found to becolonized with resistant strains during the follow-up.Owing to our study design, there is a possibility thatresistant strains were present at baseline in more thanthe two patients in whom they were actually detected.However, we provided clear evidence that resistancedue to mutations in the gyrA and/or parC geneevolved in two previously sensitive strains and thatthese changes occurred within the first three and fourweeks of antibiotic prophylaxis, respectively. We thusconfirm earlier reports which described the emer-gence of resistant fecal organisms in patients receiv-ing SID with quinolones (9, 10).

Quinolone-resistant Enterobacteriaceae can causesystemic infections in neutropenic patients on SIDwith quinolones (11), but there is a notable paucity ofreports on SBP due to such organisms (24). However,two of seven episodes of SBP with Enterobacteri-aceae in our series were caused by quinolone-resistant strains. As these two episodes occurred inpatients in whom fecal studies were not performed,we are unable to correlate them with fecal coloniza-

Fig 3. Pulsed-field gel electrophoresis (PGFE) of genomic DNAsubmitted to low-frequency restriction analysis by Xba. Lanes c andd and lanes e and f represent pairs of E. coli strains grown fromfecal cultures of patients 4 and 12 at the start of the study (strains4.0 and 12.0) and after four and three weeks, respectively (strains4.4 and 12.3). Changes in minimum inhibitory concentrations(MIC) to nalidixic acid (NAL) and ciprofloxacin (CIP) are shown.Lanes a and b; g and h, and i and j represent pairs of strains fromother patients that were not shown to be identical by PGFE.

BAUER ET AL

1360 Digestive Diseases and Sciences, Vol. 47, No. 6 (June 2002)

tion by resistant strains. These findings highlight,however, the fact that the emergence of quinolone-resistant Enterobacteriaceae in cirrhotic patients onSID may occur frequently and result in clinicallyrelevant infections. Therefore, long-term antibioticprophylaxis should be restricted to cirrhotic patientswith the highest risk for SBP (25), ie, those recoveringfrom their first episode (26). The results of this studyindicate that efforts should be directed at furtheranalysis of risk factors for first episodes of SBP (26).Additionally, alternative preventive strategies to an-tibiotic prophylaxis such as prokinetic or probioticagents should be explored. Meanwhile, norfloxacin ata dosage of 400 mg/day, continues to be the standardregimen for the secondary prevention of SBP in pa-tients with cirrhosis and ascites.

ACKNOWLEDGMENTS

The authors would like to thank Angeles Torrellas, PhD,and Lorenzo Aguilar, MD, for their valuable cooperation.

REFERENCES

1. Gines P, Rimola A, Planas R, Vargas V, Marco F, Almela M,Forne M, Miranda ML, Llach J, Salmeron JM: Norfloxacinprevents spontaneous bacterial peritonitis recurrence in cirrho-sis: Results of a double-blind, placebo-controlled trial. Hepa-tology 12:716–724, 1990

2. Tito L, Rimola A, Gines P, Llach J, Arroyo V, Rodes J:Recurrence of spontaneous bacterial peritonitis in cirrhosis:Frequency and predictive factors. Hepatology 8:27–31, 1988

3. Reeves DS: The effect of quinolone antibacterials on thegastrointestinal flora compared with that of other antibacteri-als. J Antimicrob Chemother 18 (suppl D):89–102, 1986

4. Gines P, Navasa M: Antibiotic prophylaxis for spontaneous bac-terial peritonitis: How and whom? J Hepatol 29:490–494, 1998

5. Singh N, Gayowski T, Yu VL, Wagener MM: Trimethoprim-sulfamethoxazole for the prevention of spontaneous bacterialperitonitis in cirrhosis: A randomized trial. Ann Intern Med122:595–598, 1995

6. Rolachon A, Cordier L, Bacq Y, Nousbaum JB, Franza A,Paris JC, Fratte S, Bohn B, Kitmacher P, Stahl JP: Ciprofloxa-cin and long-term prevention of spontaneous bacterial perito-nitis: Results of a prospective controlled trial. Hepatology22:1171–1174, 1995

7. Imbimbo BP, Broccali G, Cesana M, Crema F, Attardo-ParrinelloG: Inter- and intrasubject variabilities in the pharmacokinetics ofrufloxacin after single oral administration to healthy volunteers.Antimicrob Agents Chemother 35:390–393, 1991

8. Marco F, Gimenez MJ, Jimenez De Anta MT, Marcos MA,Salva P, Aguilar L: Comparison of rufloxacin and norfloxacineffects on faecal flora. J Antimicrob Chemother 35:895–901, 1995

9. Dupeyron C, Mangeney N, Sedrati L, Campillo B, Fouet P,Leluan: Rapid emergence of quinolone resistance in cirrhoticpatients treated with norfloxacin to prevent spontaneous bacterialperitonitis. Antimicrob Agents Chemother 38:340–344, 1994

10. Aparicio JR, Such J, Girona E, Gutierrez A, De Vera F,Arroyo MA, Plazas J, Palaz’on JM, Carnicer F, Perez-Mateo

M: Development of quinolone-resistant strains of Escherichiacoli in stools of patients with cirrhosis undergoing norfloxacinprophylaxis: clinical consequences. J Hepatol 31:277–283, 1999

11. Carratala J, Fernandez-Sevilla A, Tubau F, Callis M, Gudiol F:Emergence of quinolone-resistant Escherichia coli bacteremiain neutropenic patients with cancer who have received prophy-lactic norfloxacin. Clin Infect Dis 20:557–560, 1995

12. Muder RR, Brennen C, Goetz AM, Wagener MM, Rihs JD:Association with prior fluoroquinolone therapy of widespreadciprofloxacin resistance among gram-negative isolates in a Vet-erans Affairs medical center. Antimicrob Agents Chemother35:256–258, 1991

13. Soriano G, Guarner C, Teixido M, Such J, Barrios J, EnriquezJ, Vilardell F: Selective intestinal decontamination preventsspontaneous bacterial peritonitis. Gastroenterology 100:477–481, 1991

14. Runyon BA, Hoefs JC: Culture-negative neutrocytic ascites: Avariant of spontaneous bacterial peritonitis. Hepatology4:1209–1211, 1984

15. Gines P, Rimola A, Marco F, Almela M, Marques JM, Salm-eron JM, Gines A, Llach J, Jimenez De Anta MT, Arroyo V,Rodes J: Effect of norfloxacin administration on fecal flora inpatients with cirrhosis. Gastroenterol Hepatol 13:325–328, 1990

16. Vila J, Ruiz J, Goni P, De Anta MT: Detection of mutations inparC in quinolone-resistant clinical isolates of Escherichia coli.Antimicrob Agents Chemother 40:491–493, 1996

17. Gallardo F, Ruiz J, Marco F, Towner KJ, Vila J: Increase inincidence of resistance to ampicillin, chloramphenicol andtrimethoprim in clinical isolates of Salmonella serotype typhi-murium with investigation of molecular epidemiology andmechanisms of resistance. Antimicrob Agents Chemother48:367–374, 1999

18. Oram M, Fisher LM: 4-Quinolone resistance mutations in theDNA gyrase of Escherichia coli clinical isolates identified byusing the polymerase chain reaction. Antimicrob Agents Che-mother 35:387–389, 1991.

19. Cofsky RD, duBouchet L, Landesman SH: Recovery of norfloxa-cin in feces after administration of a single oral dose to humanvolunteers. Antimicrob Agents Chemother 26:110–111, 1984

20. Norrby SR: The design of clinical trials with antibiotics. EurJ Clin Microbiol Infect Dis 9:523–529, 1990

21. Pocock SJ: Clinical Trials, A Practical Approach, 1st ed. NewYork, John Wiley & Sons; 1983.

22. Edlund C, Lindqvist L, Nord CE: Norfloxacin binds to humanfecal material. Antimicrob Agents Chemother 32:1869–1874,1988

23. Nord CE: Effect of new quinolones on the human gastrointes-tinal microflora. Rev Infect Dis 10 (suppl 1):S193–S196, 1988

24. Llovet JM, Rodriguez-Iglesias P, Moitinho E, Planas R,Bataller R, Navasa M, Menacho M, Pardo A, Castells A, CabreE, Arroyo V, Gassull MA, Rodes J: Spontaneous bacterialperitonitis in patients with cirrhosis undergoing selective intes-tinal decontamination. A retrospective study of 229 spontane-ous bacterial peritonitis episodes. J Hepatol 26:88–95, 1997

25. Guamer C, Sola R, Soriano G, Andreu M, Novella M, Vila MC,Sabat M, Coll S, Ortiz J, Balanzo J: Risk of the first community-acquired spontaneous bacterial peritonitis in cirrhotics with lowascitic fluid protein. Gastroenterology 117:414–419, 1999

26. Rimola A, Garcia-Tsao G, Navasa M, Piddock LJ, Planas R,Bernard B, Inadomi JM: Diagnosis, treatment and prophylaxisof spontaneous bacterial peritonitis. A consensus document.International Ascites Club. J Hepatol 32:142–153, 2000

RUFLOXACIN IN SBP PROPHYLAXIS

1361Digestive Diseases and Sciences, Vol. 47, No. 6 (June 2002)