Antibiotic Prophylaxis in Neutropenic Patients

1743

Antibiotic Prophylaxis in Neutropenic Patients New Evidence, Practical Decisions

Leonard Leibovici, MD1,2 Mical Paul, MD1,2 Michael Cullen, MD3 Giampaolo Bucaneve, MD4 Anat Gafter-Gvili, MD1,2 Abigail Fraser, MPH1 Winfried V. Kern, MD5

New evidence shows that antibiotic prophylaxis in neutropenic patients reduces mortality, febrile episodes, and bacterial infections. For patients with acute leukemia or those who undergo bone marrow transplantation, prophylaxis with fluoroquinolones diminished the risk of death from any cause by 33% (95% confidence interval [95% CI], 2–54%). Thus, 55 patients who have acute leukemia or who undergo bone marrow transplantation must receive prophylaxis to prevent 1 death. In 4 studies that included patients with solid tumors or lymphoma, prophylaxis reduced the rate of death during the first month (relative risk, 0.51; 95% CI,

1

Department of Medicine E, Beilinson Campus, Rabin Medical Center, Petah-Tiqva, Israel.

0.27–0.97), and 82 patients had to receive prophylaxis to prevent 1 death. The main

2

received prophylaxis did not experience more infections caused by resistant strains than patients in the control group. The recent GIMEMA study was con-

Department of Medicine, Sackler Faculty of Medicine, Tel-Aviv University, Ramat-Aviv, TelAviv, Israel. 3

University Hospital Birmingham Cancer Centre, Birmingham, United Kingdom. 4

Istituto di Medicina Interna e Scienze Oncologiche, Policlinico Monteluce, Perugia, Italy. 5

Center for Infectious Diseases and Travel Medicine, Department of Medicine, University Hospital, Freiburg, Germany.

argument brought against prophylaxis is the induction of resistance. Patients who

ducted in a population with a nearly 50% resistance to fluoroquinolones in all pathogens and 20% resistance in gram-negative isolates, thus indicating that prophylaxis should be offered in settings with similar or less resistance. Prophylaxis with fluoroquinolones was efficacious in reducing infections caused by gram-positive bacteria. Patients who are treated for acute leukemia should be offered prophylaxis with ciprofloxacin or levofloxacin. Prophylaxis to cover the expected period of neutropenia may be considered for the first cycle of treatment in patients with solid tumors or lymphoma who regularly receive regimens that cause severe neutropenia. Excessive local levels of resistance to fluoroquinolones or high local incidence of infections caused by Clostridium difficile and related to fluoroquinolones should prompt a reconsideration of this policy. Cancer 2006;107: 1743–51. 2006 American Cancer Society.

KEYWORDS: neutropenia, antibiotic prophylaxis, ciprofloxacin, levofloxacin.

N

Dr. Cullen has received honoraria from Sanofi Aventis. We thank Dr. Vivianne Tjan-Heijnen and Ms. Muriel Debois for supplying us with full data regarding mortality from their study. Address for reprints: Leonard Leibovici, MD, Department of Medicine E, Beilinson Campus, Rabin Medical Center, 49100 Petah-Tiqva, Israel; Fax: (011) 972 39376512; E-mail: leibovic@post.tau.ac.il Received April 26, 2006; revision received June 23, 2006; accepted July 10, 2006.

ª 2006 American Cancer Society

ew evidence shows that antibiotic prophylaxis in neutropenic patients reduces mortality, febrile episodes, and bacterial infections. A systematic review and meta-analysis of randomized controlled trials1 demonstrated that death from all causes was reduced by 34% (95% confidence interval [95% CI], 21–45%) in neutropenic patients who received any antibiotic prophylaxis and by 40% (95% CI, 17–56%) in patients who received quinolones for prophylaxis. in addition, the occurrence of febrile episodes and bacterial infections decreased significantly (Table 1). Most of the studies that were included in that meta-analysis addressed patients with hematologic malignancies. In the most recent and largest randomized controlled trial that included patients in whom neutropenia was expected to occur for more than 7 days (mainly with acute leukemia and autologous peripheral blood stem cell transplantation),2 patients who received levofloxacin as prophylaxis had a relative risk (RR) of 0.54 (95% CI, 0.25–1.16) for mortality compared with the placebo group: a

DOI 10.1002/cncr.22205 Published online 14 September 2006 in Wiley InterScience (www.interscience.wiley.com).

1744

CANCER

October 15, 2006 / Volume 107 / Number 8

TABLE 1 Prophylaxis with Fluoroquinolones in Neutropenic Patients* Patients (Study)/Event

RR (95% CI)

95% CI

Absolute risk in the control group (%)

No. of treated patients needed to prevent 1 event

All patients (Gafter-Gvili et al., 20051)y Death from any cause 0.52 0.35–0.77 8.7 24 Febrile episode 0.67 0.56–0.81 72 4 Bacterial infection 0.50 0.35–0.70 45 5 Patients with expected prolonged neutropenia: Mainly acute leukemia and stem cell transplantation (Bucaneve et al., 20052) Death from any cause 0.54 0.25–1.16 5 43 Febrile episode 0.76 0.69–0.83 85 5 Bacterial infection 0.56 0.44–0.71 39 6 Patients with solid tumors and lymphomas: No stem cell transplantation (Cullen et al., 20053) 0.67 0.32–1.38 2.3 132 Death from any cause{ Febrile episode§ 0.71 0.55–0.92 15 23 Bacterial infection§ 0.82 0.73–0.94 42 13 RR indicates relative risk; 95% CI, 95% confidence interval. * Values shown are the RR and the number of treated patients needed to prevent 1 death, a febrile episode, and a bacterial infection (according to a meta-analysis [Gafter-Gvili et al., 20051] and the 2 largest, recent, randomized controlled trials [Bucaneve et al., 20052; Cullen et al., 20053]). y The 2 recent trials (Bucaneve et al., 20052; Cullen et al., 20053) were not included. { During the first 30 days. § In all cycles of treatment.

decrease of 46%. The occurrence of febrile episodes and bacterial infections was reduced significantly (Table 1). The close similarity of the point estimates between the meta-analysis and this large recent study supports the robustness of the results. In the SIGNIFICANT trial,3 1565 patients with solid tumors and lymphomas were randomized to receive either levofloxacin or placebo after chemotherapy during up to 6 cycles of chemotherapy. In patients who received levofloxacin, the 30-day mortality rate was 1.5% (12 of 781 patients); and, in the placebo group, the 30-day mortality rate was 2.3% (18 of 784 patients; RR, 0.67; 95% CI, 0.32–1.38). At the end of follow-up the mortality rates were 4% (31 of 781 patients) in the levofloxacin group and 4.6% (36 of 784 patients) in the placebo group (RR, 0.86; 95% CI, 0.54–1.38; unpublished data). Significant reductions in febrile episodes and bacterial infections were reported during the first cycle of chemotherapy and for all cycles (Table 1). Hospitalizations were reduced by 27% (95% CI, 10–41%) for all cycles. Current guidelines4 do not recommend antibiotic prophylaxis for neutropenia. The main concerns expressed are that a reduction in mortality has not been demonstrated, whereas the widespread use of prophylactic antibiotics may increase antibacterial resistance. The first argument no longer is valid. Antibiotic prophylaxis reduces mortality, and low numbers of treated patients are required to prevent 1 death. Practitioners should consider antibiotic prophylaxis for neutropenic patients: They should consider which

patients should be offered prophylaxis and which drugs to use, and they should appraise whether major determinants of prophylaxis efficacy are sufficiently different in their practice from the randomized controlled trials to justify withholding antibiotic prophylaxis. Finally, they should balance the gains from prophylaxis with its costs and drawbacks. We also should ask what further research is warranted. We address several points in this report to help with these considerations.

Which Patients Should Be Offered Antibiotic Prophylaxis? Patients treated for acute leukemia or undergoing ablative bone marrow treatment before transplantation We have updated our meta-analysis1 with data from recent trials2,3 and divided the trials5–22 according to the type of patient (Fig. 1). For studies that included only patients with acute leukemia and those who received high-dose chemotherapy with stem cell transplantation, the risk of death from any cause was diminished by prophylaxis with fluoroquinolones by 33% (95% CI, 2–54%). Because mortality during neutropenia has declined in recent years, a rate of 5.5% deaths in the control group was used (derived from studies that were performed after 2000) to estimate that 55 patients with acute leukemia or who received high-dose chemotherapy with stem cell transplantation needed to receive prophylaxis with a fluoroquinolone to prevent 1 death. The drawbacks associated with antibiotic treatment are the cost of the drug, its toxicity, and the in-

Antibiotic Prophylaxis in Neutropenia/Leibovici et al.

1745

FIGURE 1. This chart illustrates the relative risks (RR) with 95% confidence intervals (95% CI) for all-cause mortality in patients who received fluoroquinolones versus patients who received either placebo or no treatment (the 2 recent trials [Bucaneve et al., 20052; Cullen et al., 20053] were included). Studies are listed according to lead the author and the year (see the list of References). BMT indicates bone marrow transplantation; Chi2, chi-square test; df, degrees of freedom; I2, test for heterogeneity.

duction of antimicrobial resistance. Because the numbers needed to treat to prevent 1 death are low, cost is a secondary consideration. Patients who received fluoroquinolones for prophylaxis suffered more side effects than patients who received placebo or no treatment (RR, 1.41; 95% CI, 1.09–1.83), resulting in 25 as the number needed to harm.1 However, all of those side effects were minor. The use of fluoroquinolone prophylaxis did not increase the risk of fungal infections.1 Fluoroquinolones are drugs in common use that have an acceptable safety profile, especially in a situation in which mortality is reduced. Thus, the main consideration against the use of prophylaxis is the induction of resistance. This may occur on 3 levels: colonization and then infection with resistant microorganisms in the neutropenic individual, thus affecting health and chances of survival; change in the flora of the ward or unit in which this practice is adopted, thus potentially harming other patients; and a change in resistance to fluoroquinolones in the population at large. In trials that reported resistance data, patients who received fluoroquinolones developed no more infections with pathogens resistant to the drug than patients who received placebo (RR, 1.04; 95% CI, 0.73–1.5) (Fig. 2). Of all infections that developed among patients who received fluoroquinolone prophylaxis, 30% were quinolone-resistant; thus, fluoroquinolones cannot be used for the treatment of febrile neutropenia in patients who receive prophylaxis. The total antibiotic consumption is not necessarily increased in patients who receive prophylaxis, because prophylaxis reduces the need for

therapeutic interventions.23 Because the overall mortality was reduced by prophylaxis, the danger of infection caused by resistant pathogens to a particular patient evidently was much smaller than the gain. Many reports have documented the emergence of bacteria resistant to fluoroquinolones in units in which prophylaxis is practiced (e.g., see Cometta et al., Kern et al., Baum et al., and Razonable et al.24–27). There are no quantitative data indicating the extent to which patients came to harm as a result. The ethics of withholding a clearly effective antibiotic drug from current patients to reserve it for future patients are far from clear.28 Withholding antibiotic treatment that reduces mortality is not the rationale that guides our daily antibiotic practice: We use wide-spectrum antibiotics and even antibiotics of last resort in patients who likely will derive a survival benefit, even though we are sure that consumption drives resistance. Thus, neutropenic patients with acute leukemia or who receive high-dose chemotherapy with stem cell transplantation should be offered antibiotic prophylaxis if no other considerations prevail. In the population at large, the quantity of fluoroquinolones to be given as prophylaxis to patients with acute leukemia is negligible compared with the consumption of fluoroquinolones for other infections (e.g., of the urinary tract). For example, the yearly incidence of acute leukemia in the U.S. is 0.012%, and the yearly incidence for cancer of all sites is 0.47%.29 The incidence of urinary tract infections is approximately 2% in men and 13% in women, and >33% of those infections are treated with fluoroquinolones.30,31

1746

CANCER


FIGURE 2. This chart illustrates the relative risks (RR) with 95% confidence intervals (95% CI) for an infection caused by pathogen(s) resistant to the fluoroquinolone used for prophylaxis (the 2 recent trials [Bucaneve et al., 20052; Cullen et al., 20053] were included). Studies are listed according to the lead author and the year (see the list of References). Chi2 indicates chi-square test; df, degrees of freedom; I2, test for heterogeneity.

Patients treated for solid tumors or lymphoma Four studies that compared fluoroquinolones with placebo included only patients who had solid tumors or lymphoma and reported all-cause mortality (Fig. 1). In those studies, prophylaxis reduced the rate of death during the first month (RR, 0.51; 95% CI, 0.27–0.97). Using the rate of death in the placebo groups of studies that were performed after 2000 (2.5%), the number of treated patients needed to prevent 1 death during the first month of treatment for solid tumors and lymphomas was 82. The heterogeneity between the results of the studies was small (I2, 0%). However, the small studies showed larger effects than the 1 large, recent study.3 However, even using only the point estimates from the SIGNIFICANT trial,3 the number of treated patients needed to prevent 1 death during the first month of treatment is 132. The SIGNIFICANT trial3 addressed the question of antibiotic prophylaxis for all cycles of treatment in these patients. The RR for mortality for the duration of treatment was closer to 1 (0.86; 95% CI, 0.54–1.38). In the same trial, 47% of deaths occurred during the first month, the median follow-up was 3.7 months, and the longest duration of follow-up was 7.6 months. Fluoroquinolone prophylaxis reduced the occurrence of febrile episodes (RR, 0.74; 95% CI, 0.69–0.78) (Fig. 3). The lowest incidence of febrile episodes in controls (16%) was observed in the most recent, largest study.3 Using this estimate, treating 25 patients would prevent 1 febrile episode. A febrile episode generally involves hospitalization (although many times a short one) and antibiotic treatment (although many times an oral antibiotic32). Existing evidence points to an advantage to prophylaxis in patients who receive chemotherapy for solid tumors or lymphoma at least for the first cycle of chemotherapy; in these patients, prophylaxis reduces the chances of death significantly during the first month. Approximately 50% of all deaths during treatment occur during this period. Febrile episodes,

infections, and hospitalizations also are reduced. For the whole duration of chemotherapy, the evidence is less compelling. An ongoing post-hoc analysis of the SIGNIFICANT trial3 will help define patient characteristics, chemotherapy regimens, and cycles in which the benefit of prophylaxis is greatest.

Choice of Drug and Duration of Treatment Choice of drug Fluoroquinolones were compared with trimethoprimsulfamethoxazole in 10 studies, all of which were conducted >10 years ago. All-cause mortality and episodes of fever were similar.1 Any side effects, including those that required discontinuation of treatment, were significantly more common in patients who received trimethoprim-sulfamethoxazole. Using an indirect comparison, studies that compared fluoroquinolones with placebo or no treatment yielded an RR for all-cause mortality of 0.62 (95% CI, 0.45–0.86) (meta-analysis1 updated with the 2 new studies2,3); whereas studies that used trimethoprim-sulfamethoxazole yielded an RR of 0.71 (95% CI, 0.49–1.02).1 All studies of trimethoprimsulfamethoxazole versus placebo or no treatment were performed before 1993. Because the susceptibility to trimethoprim-sulfamethoxazole declined in the last 2 decades worldwide (albeit at different rates), and because of the rate of side effects and possibly greater efficacy for fluoroquinolones, the latter probably will be the practical choice for most practitioners. Prophylaxis with trimethoprim-sulfamethoxazole may prevent infections caused by Pneumocystis jirovecii, Stenotrophomonas maltophilia, Listeria monocytogenes, and Nocardia spp. However, none of the studies confirmed its preventive capability, and it did not translate into a benefit for trimethoprim-sulfamethoxazole. Only 5 studies compared 1 fluoroquinolone with another, but those studies were too small to reach any conclusion about the equivalence or advantage of 1 drug.1 Norfloxacin was compared with either placebo or no treatment in 4 small studies, and no effect was


1747

FIGURE 3. This chart illustrates the relative risks (RR) with 95% confidence intervals (95% CI) for a febrile episode in patients who received fluoroquinolones versus patients who received either placebo or no treatment (the 2 recent trials [Bucaneve et al., 20052; Cullen et al., 20053] were included). Studies are listed according to the lead author and the year (see the list of References). Chi2 indicates chi-square test; df, degrees of freedom; I2, test for heterogeneity.

FIGURE 4. This chart illustrates the relative risks (RR) with 95% confidence intervals (95% CI) for death in neutropenic patients who received various fluoroquinolones (the 2 recent trials [Bucaneve et al., 20052; Cullen et al., 20053] were included). Studies are listed according to the lead author and the year (see the list of References). Chi2 indicates chi-square test; df, degrees of freedom; I2, test for heterogeneity.

reported (RR for mortality, 1.03; 95% CI, 0.58–1.81) (Fig. 4). Ciprofloxacin was compared with either placebo or no treatment in 6 studies and reduced allcause mortality (RR, 0.32; 95% CI, 0.13–0.82) (Fig. 4). The 2 new, larger trials used levofloxacin. Ciprofloxa-

cin was used in doses ranging between 500 mg per day and 750 mg twice daily. The 2 new, large trials used levofloxacin at a single daily dose of 500 mg. A practical decision would be to use either ciprofloxacin or levofloxacin for prophylaxis. Practitioners

1748

CANCER


may choose between them based on the theoretical advantage of ciprofloxacin against Pseudomonas aeruginosa or the theoretical advantage of levofloxacin against gram-positive bacteria, according to the local distribution of pathogens in neutropenic patients. Levofloxacin has the additional advantage of requiring once-daily administration, enhancing compliance, which is important in prophylactic settings.

Commencement and end of prophylaxis In the majority of studies that included patients with acute leukemia or patients who underwent bone marrow transplantation, prophylaxis was started at the initiation of chemotherapy and was stopped at the resolution of neutropenia or the appearance of fever. In 3 of 4 studies3,19,22 of patients with solid tumors and lymphoma, prophylaxis was started later to cover the anticipated nadir period of neutrophils. No differences in the reduction of risk for mortality or febrile episodes were evident between studies in which prophylaxis was started with the initiation of chemotherapy and studies in which prophylaxis was started with the appearance of neutropenia. In an effort to reduce total consumption of antibiotics, prophylaxis may be started with appearance of neutropenia; however, this approach necessitates frequent determinations of the neutrophil count. Local Factors that May Affect Decisions Resistance to fluoroquinolones The high prevalence of resistance to fluoquinolones probably will annul the efficacy of prophylaxis. What is the threshold of resistance above which prophylaxis no longer will be effective? For an approximation of background resistance in settings in which prophylaxis did work, we looked at the percentage of infections caused by pathogens that were resistant to fluoroquinolones in the control group (who received either placebo or no treatment) in randomized controlled trails. The GIMEMA trial2 is the most recent and has the highest percentage of resistance. In the control group from that trial, 47% of 68 isolates were resistant to levofloxacin, and the resistance rate among gram-negative aerobic bacteria was 17%. Six other trials reported on resistant pathogens in the control group, all of them with a lower percentage of resistance and a few isolates. Several observational studies have examined what happens to outcomes of neutropenic patients in settings with high resistance to fluoroquinolones when the practice of antibiotic prophylaxis is stopped.33–38 At a university hospital in Germany 2 attempts to discontinue prophylaxis were stopped because there were more bacteremia and deaths during the periods

when prophylaxis was not given.33,34 Discontinuation of prophylaxis in a Spanish and Swiss hospital35,36 led to an increase in febrile episodes and bacteremia, but the differences did not reach statistical significance. In the Swiss study, there were no deaths among 41 patients who received ciprofloxacin prophylaxis compared with 3 deaths among 71 patients who received no prophylaxis. In the Spanish study, mortality was not described. Gomez et al. reported that the incidence of febrile episodes and bacteremia did not increase when prophylaxis was stopped in another hospital in the same region in Spain.37 In a third observational study from Spain, there was a tendency for reduced mortality in patients who received levofloxacin compared with patients who received either trimethoprimsulfamethoxazole or no prophylaxis.38 Long-term surveillance for the development of ciprofloxacin resistance was conducted after a randomized controlled trial of antibiotic prophylaxis in hematologic patients.39 After 10 years of continued prophylaxis with ciprofloxacin plus colistin, only 5 ciprofloxacin-resistant, gram-negative bacteria were isolated from 66 episodes of microbiologically documented infection among 47 patients during approximately 700 neutropenic episodes. None of those infections were fatal. We have no solid evidence to point out a measure and percentage of resistance that may serve as a threshold for deciding not to use prophylaxis. We can state with a high degree of confidence that prophylaxis with fluoroquinolones was as effective in the large, recent GIMEMA study2 as it was in previous studies.1 The GIMEMA study was conducted in a population with nearly 50% resistance to fluoroquinolones in all pathogens and 20% resistance in gram-negative isolates in the control group and in a country with a baseline resistance of approximately 20% in gram-negative isolates from the community40 and medical departments.41 Prophylaxis should be considered in locations that have similar or less resistance.

Percentage of gram-positive infections in neutropenic patients Fluoroquinolones are active mainly against gram-negative pathogens. In many centers, gram-positive bacteria are common causes of infection in neutropenic patients. Will a baseline high percentage of infections caused by gram-positive bacteria annul the efficacy of prophylaxis with fluoroquinolones? Prophylaxis with fluoroquinolones was efficacious in reducing infections caused by gram-positive bacteria (RR, 0.37; 95% CI, 0.30–0.46) (Fig. 5). The efficacy of fluoroquinolones was mainly because of a reduction of streptococcal and methicillin-sensitive Staphylococcus aureus infections.


1749

FIGURE 5. This chart illustrates the relative risks (RR) with 95% confidence intervals (95% CI) of a gram-positive infection in neutropenic patients who received fluoroquinolone prophylaxis versus patients who received either placebo or no treatment (the 2 recent trials [Bucaneve et al., 20052; Cullen et al., 20053] were included). Studies are listed according to the lead author and the year (see the list of References). BMT indicates bone marrow transplantation; Chi2, chisquare test; df, degrees of freedom; I2, test for heterogeneity.

The ratio of gram-positive infections to gram-negative infections in the control group of the GIMEMA study2 was 61:47 (1.3); whereas, in the other studies that have been published since 2000, the ratio was close to 1. In a meta-regression analysis, we observed no correlation between the percentage of gram-positive pathogens in the control group and the protective effect of prophylaxis against death and occurrence of fever. Even in settings with relatively high prevalence of gram-positive pathogens, prophylaxis with fluoroquinolones is effective.

Infections caused by Clostridium difficile A virulent new strain of Clostridium difficile has caused severe disease and death in several locations in the U.S., Canada, the Netherlands, and the U.K.42–45 The new strain is resistant to fluoroquinolones, and fluoroquinolone consumption is a risk factor for diarrhea.43–45 It is an additional concern that the use of fluoroquinolones for prophylaxis may lead to outbreaks of the new strain of Clostridium difficile. Data on Clostridium difficile infections were available in 5 studies in which a fluoroquinolone for prophylaxis was compared with either placebo or no treatment.2,3,12,18,19 Seven documented infections were observed in 1250 patients who were randomized to receive prophylaxis in those studies, and there were 5 infections in 1279 patients who were randomized to receive either placebo or no treatment (RR, 1.44; 95% CI, 0.45–4.63). In the 2 new and larger studies2,3 no Clostridium difficile infections were detected. In locations in which the new strain did not appear, Clostridium difficile infections were not a deterrent to the use of prophylaxis. Clinics in which the strain has appeared will have to take it into account in their decisions whether to offer prophylaxis. We may assume that, in addition to the use of fluoroquinolones, the spread of

the new strain of Clostridium difficile is associated with breaches in infection control.46 Infection-control measures should be emphasized and practiced meticulously in neutropenic patients.

Further Research Further placebo-controlled, randomized, controlled trials in patients with acute leukemia or patients who undergo autologous stem cell transplantation are justified only to examine new drugs versus existing fluoroquinolones or to compare antibiotics with placebo in locations with high resistance to fluoroquinolones. This resistance should be higher than that experienced in the GIMEMA study.2 The exception is patients who undergo allogeneic (hematopoietic or peripheral) stem cell transplantation. For patients with solid tumors and lymphomas, post-hoc analyses of existing trials and observational studies should attempt to identify which patients are more likely to benefit from prophylaxis. Further trials may be warranted among such patients. A worthwhile study may be a comparison between prophylaxis and early, patient-initiated treatment of febrile episodes. We also could learn from prospective, observational, longitudinal studies in centers with different prophylaxis policies and strategies. Data on outcomes of neutropenic patients, including pathogens and resistance, should be collected prospectively using uniform protocols and definitions. In conclusion, prophylaxis with ciprofloxacin or levofloxacin should be considered in patients who are treated for acute leukemia or who received high-dose chemotherapy with autologous stem cell transplantation. Pending further research, prophylaxis to cover the expected period of neutropenia can be considered for the first cycle of treatment in patients with solid tumors or lymphoma. Excessive local levels of resistance to

1750

CANCER


fluoroquinolones or high local incidence of infections caused by Clostridium difficile and related to fluoroquinolones should prompt a reconsideration of this policy.

REFERENCES 1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

Gafter-Gvili A, Fraser A, Paul M, Leibovici L. Meta-analysis: antibiotic prophylaxis reduces mortality in neutropenic patients. Ann Intern Med. 2005;142:979–995. Bucaneve G, Micozzi A, Menichetti F, et al. Levofloxacin to prevent bacterial infection in patients with cancer and neutropenia. N Engl J Med. 2005;353:977–987. Cullen M, Steven N, Billingham L, et al. Antibacterial prophylaxis after chemotherapy for solid tumors and lymphomas. N Engl J Med. 2005;353:988–998. Hughes WT, Armstrong D, Bodey GP, et al. 2002 Guidelines for the use of antimicrobial agents in neutropenic patients with cancer. Clin Infect Dis. 2002;34:730–751. Brodsky AL, Minissale CJ, Melero MJ, Sanchez Avalos JC. [Prophylaxis with fluoroquinolones in patients with neutropenia]. Medicina (B Aires). 1993;53:401–407. Karp JE, Merz WG, Hendricksen C, et al. Oral norfloxacin for prevention of gram-negative bacterial infections in patients with acute leukaemia and granulocytopenia. A randomized, double-blind, placebo-controlled trial. Ann Intern Med. 1987; 106:1–7. Lee DG, Choi SM, Choi JH, et al. Selective bowel decontamination for the prevention of infection in acute myelogenous leukaemia: a prospective randomized trial. Korean J Intern Med. 2002;17:38–44. Moreau P, Milplied N, Richard P, Bulbois CE, Richet H, Harousseau JL. Prospective randomized study comparing prophylactic ciprofloxacin (C) and amoxicillin þ clavulanic acid (A-CA) versus prophylactic vancomycin (V) versus nonprophylactic conventional empirical antimicrobial therapy in neutropenic patients. Bone Marrow Transplant. 1995;15:S129. Abstract. Nenova IS, Ananostev NH, Goranov SE, Mateva NG, Haidushka IA. Fluoroquinolone prophylaxis for bacterial infections in neutropenic patients with hematologic malignancies. Folia Med (Plovdiv). 2001;43:40–55. Ruiz I, Fuentes I, Capdevila JA, et al. Ofloxacin prophylaxis in preventing bacterial infection after autologous peripheral blood stem cell transplantation (A-PBSC). In: Abstracts of the 41st Interscience Conference on Antimicrobial Agents and Chemotherapy. Washington, DC: American Society of Microbiology; 2001:447. Abstract L-781. Sampi K, Maseki N, Hattori M. [A comparison of nystatin with norfloxacin for prevention of infection after consolidation therapy in patients with acute leukaemia or autologous bone marrow transplantation: a randomized study]. Gan To Kagaku Ryoho. 1992;19:823–826. Talbot GH, Cassileth PA, Paradiso L, Correa-Coronas R, Bond L. Oral enoxacin for infection prevention in adults with acute nonlymphocytic leukaemia.The Enoxacin Prophylaxis Study Group. Antimicrob Agents Chemother. 1993; 37:474–482. Thomas X, Troncy J, Belhabri A, et al. [Effectiveness of combined vancomycin and pefloxacine in gastrointestinal decontamination for preventing infections after chemotherapyinduced bone marrow aplasia. A randomized double-blind study]. Presse Med. 2000;29:1745–1751. Tsutani H, Imamura S, Ueda T, et al. Prophylactic use of ofloxacin in granulocytopenic patients with hematological

15.

16.

17.

18.

19.

20. 21.

22.

23.

24.

25.

26.

27.

28. 29. 30.

31.

32.

malignancies during post-remission chemotherapy. Intern Med. 1992;31:319–324. Yamada T, Dan K, Nomura T. Prevention of bacterial and fungal infections in acute leukaemia patients: a new and potent combination of oral norfloxacin and amphotericin B. Intern Med. 1993;32:710–715. Casali A, Verri C, Paoletti G, et al. Chemoprophylaxis of bacterial infections in granulocytopenic cancer patients using norfloxacin. Chemioterapia. 1988;7:327–329. Lew MA, Kehoe K, Ritz J, et al. Prophylaxis of bacterial infections with ciprofloxacin in patients undergoing bone marrow transplantation. Transplantation. 1991;51:630–636. Maiche AG, Muhonen T. Granulocyte colony-stimulating factor (G-CSF) with or without a quinolone in the prevention of infection in cancer patients. Eur J Cancer. 1993;29A: 1403–1405. Carlson JW, Fowler JM, Mitchell SK, Carson LF, Mayer AR, Copeland LJ. Chemoprophylaxis with ciprofloxacin in ovarian cancer patients receiving paclitaxel: a randomized trial. Gynecol Oncol. 1997;65:325–329. Hartlapp JH. Antimicrobial prophylaxis in immunocompromised patients. Drugs. 1987;34(Suppl 1):131–133. Schroeder M, Schadeck-Gressel C, Selbach J, Westerhausen M. Antibiotic prophylaxis with gyrase inhibitors during cytostatically induced granulocytopenias in patients with solid tumors: a double-blind prospective randomized study. Onkologie. 1992;15:476–479. Tjan-Heijnen VC, Postmus PE, Ardizzoni A, et al. Reduction of chemotherapy-induced febrile leucopenia by prophylactic use of ciprofloxacin and roxithromycin in small-cell lung cancer patients: an EORTC double-blind placebo-controlled Phase III study. Ann Oncol. 2001;12:1359–1368. Kern WV, Steib-Bauert M, de With K, et al. Fluoroquinolone consumption and resistance in haematology-oncology patients— ecological analysis in two university hospitals 1999–2002. J Antimicrob Chemother. 2005;55:57–60. Cometta A, Calandra T, Bille J, Glauser MP. Escherichia coli resistant to fluoroquinolones in patients with cancer and neutropenia. N Engl J Med. 1994;330:1240–1241. Kern WV, Andriof E, Oethinger M, Kern P, Hacker J, Marre R. Emergence of fluoroquinolone-resistant Escherichia coli at a cancer center. Antimicrob Agents Chemother. 1994;38:681– 687. Baum HV, Franz U, Geiss HK. Prevalence of ciprofloxacinresistant Escherichia coli in hematologic-oncologic patients. Infection. 2000;28:278–281. Razonable RR, Litzow MR, Khaliq Y, Piper KE, Rouse MS, Patel R. Bacteremia due to viridans group streptococci with diminished susceptibility to levofloxacin among neutropenic patients receiving levofloxacin prophylaxis. Clin Infect Dis. 2002;34:1469–1474. Leibovici L, Shraga I, Andreassen S. How do you choose antibiotic treatment? BMJ. 1999;318:1614–1618. National Cancer Institute. SEER Statistics. Available from URL: http://seer.cancer.gov/statfacts [accessed August 24, 2006]. Griebling TL. Urologic diseases in America project: trends in resource use for urinary tract infections in men. J Urol. 2005; 173:1288–1294. Griebling TL. Urologic diseases in America project: trends in resource use for urinary tract infections in women. J Urol. 2005;173:1281–1287. Vidal L, Paul M, Ben Dor I, Soares-Weiser K, Leibovici L. Oral versus intravenous antibiotic treatment for febrile neutropenia in cancer patients: a systematic review and meta-analysis of randomized trials. J Antimicrob Chemother. 2004;54:29–37.

Antibiotic Prophylaxis in Neutropenia/Leibovici et al. 33. Kern WV, Klose K, Jellen-Ritter AS, et al. Fluoroquinolone resistance of Escherichia coli at a cancer center: epidemiologic evolution and effects of discontinuing prophylactic fluoroquinolone use in neutropenic patients with leukaemia. Eur J Clin Microbiol Infect Dis. 2005;24:111–118. 34. Reuter S, Kern WV, Sigge A, et al. Impact of fluoroquinolone prophylaxis on reduced infection-related mortality among patients with neutropenia and hematologic malignancies. Clin Infect Dis. 2005;40:1087–1093. 35. Martino R, Subira M, Altes A, et al. Effect of discontinuing prophylaxis with norfloxacin in patients with hematologic malignancies and severe neutropenia. A matched case-control study of the effect on infectious morbidity. Acta Haematol. 1998;99:206–211. 36. Delarive P, Baumgartner JD, Glauser MP, Cometta A. Evaluation de la prophylaxie antibiotique chez les patients neutropeniques avec hemopathie maligne. Schweiz Med Wochenschr. 2000;130:1837–1844. 37. Gomez L, Garau J, Estrada C, et al. Ciprofloxacin prophylaxis in patients with acute leukaemia and granulocytopenia in an area with a high prevalence of ciprofloxacin-resistant Escherichia coli. Cancer. 2003;97:419–424. 38. Gracia Escudero A, Martin Gonzalez M, Gimenez Garrido F, et al. Estudio prospectivo y controlado, no aleatorizado de tres cohortes sobre la efectividad de dos pautas antibioticas, levofloxacino y contrimoxazol, en profilaxis antibacteriana de pacientes neutropenicos. Med Clin (Barcelona). 2003;120: 321–325. 39. Prentice HG, Hann IM, Nazareth B, Paterson P, Bhamra A, Kibbler CC. Oral ciprofloxacin plus colistin: prophylaxis

40.

41.

42.

43.

44.

45.

46.

1751

against bacterial infection in neutropenic patients. A strategy for the prevention of emergence of antimicrobial resistance. Br J Haematol. 2001;115:46–52. Fadda G, Nicoletti G, Schito GC, Tempera G. Antimicrobial susceptibility patterns of contemporary pathogens from uncomplicated urinary tract infections isolated in a multicenter Italian survey: possible impact on guidelines. J Chemother. 2005;17:251–257. Luzzaro F, Vigano EF, Fossati D, et al. Prevalence and drug susceptibility of pathogens causing bloodstream infections in northern Italy: a two-year study in 16 hospitals. Eur J Clin Microbiol Infect Dis. 2002;21:849–855. Kuijper EJ, van den Berg RJ, Debast S, et al. Clostridium difficile ribotype 027, toxinotype III, the Netherlands. Emerg Infect Dis. 2006;12:827–830. Pepin J, Saheb N, Coulombe MA, et al. Emergence of fluoroquinolones as the predominant risk factor for Clostridium difficile associated diarrhea: a cohort study during an epidemic in Quebec. Clin Infect Dis. 2005;41:1254–1260. Loo VG, Poirier L, Miller MA, et al. A predominantly clonal multi-institutional outbreak of Clostridium difficile-associated diarrhea with high morbidity and mortality. N Engl J Med. 2005;353:2442–2449. McDonald LC, Killgore GE, Thompson A, et al. An epidemic, toxin gene-variant strain of Clostridium difficile. N Engl J Med 2005;353:2433–2441. Weiss K. Poor infection control, not fluoroquinolones, likely to be primary cause of Clostridium difficile-associated diarrhea outbreaks in Quebec. Clin Infect Dis. 2006;42:725– 727.