Prophylaxis of early bacterial infections after autologous ... - Nature

Bone Marrow Transplantation (2005) 36, 59–65 & 2005 Nature Publishing Group All rights reserved 0268-3369/05 $30.00

www.nature.com/bmt

Prophylaxis of early bacterial infections after autologous peripheral blood stem cell transplantation (PBSCT): a matched-pair study comparing oral fluoroquinolones and intravenous piperacillin–tazobactam C Solano1, A Gutierrez1, F Martinez2, C Gimeno3, C Go´mez1, I Munõz3, F Faus4, R Goterris1, A Farga3 and D Navarro3 1

Department of Haematology and Medical Oncology, University Clinic Hospital, School of Medicine, University of Valencia, Valencia, Spain; 2Department of Statistics, University of Valencia, Valencia, Spain; 3Department of Microbiology, School of Medicine, University of Valencia, Valencia, Spain; and 4School of Nursing, University of Valencia, Valencia, Spain

Summary: The safety and efficacy of early bacterial prophylaxis with piperacillin–tazobactam were prospectively evaluated in 51 autologous peripheral blood stem cell transplantation (PBSCT) recipients. The results were compared with those obtained in 51 control patients receiving oral fluoroquinolones in a retrospective matched-pair control study. Overall, 76% of the study group and 98% of the control group developed at least one febrile episode during neutropenia (P ¼ 0.002). Time from neutropenia to the first febrile episode (FFE) was significantly longer in the study group than in the control group (P ¼ 0.04). Once a febrile episode appeared, the duration of fever was significantly longer in cases than in controls (median of 5 and 2 days respectively, Po0.001), and led to a more frequent use of empirical amphotericin B (AmB), not statistically significant (P ¼ 0.13). However, the total time of antibiotic administration was significantly greater in the control than in the study group (P ¼ 0.05). The duration of AmB treatment shows a trend toward a longer duration in the control than in study group (P ¼ 0.2). Overall, 86% of the Gram-positive bacteremia and 85% of the Gram-negative bacteria were susceptible to the tested antibiotics. Our study suggests that a subgroup of patients could benefit from prophylaxis with piperacillin– tazobactam without increasing toxicity or bacterial resistance. Bone Marrow Transplantation (2005) 36, 59–65. doi:10.1038/sj.bmt.1705005 Published online 23 May 2005 Keywords: peripheral blood stem cell transplantation; early bacterial infections; antibacterial prophylaxis; piperacillin–tazobactam

Early bacterial infection continues to be an important cause of morbidity and mortality after hematopoietic stem cell transplantation.1 In spite of the reduction in post transplant aplasia associated with the use of peripheral blood stem cells and myeloid growth factors, more than 80% of patients develop fever and infections that are the leading cause of transplant-related mortality (TRM).2–4 The most frequent approach to reducing the incidence of infection is the administration of oral antibiotic regimens, which aim to eliminate aerobic bacteria while maintaining colonization resistance by preserving the anaerobic flora.5–7 In neutropenic patients, antibacterial chemoprophylaxis with fluorinated quinolones has reduced the incidence of Gram negative bacteremia.8 However, quinolones have been associated with increased incidence of viridans streptococci bacteremia9–11 and quinolone-resistant aerobic Gram-negative bacilli (GNB) infections.12–15 Given the absence of well-demonstrated efficacy of oral prophylactic antibiotics, some centers have tried prophylactic intravenous antibiotics in patients with neutropenia of well-predicted duration, who usually need empiric antibacterial treatment despite oral antibacterial chemoprophylaxis.16–18 The aim of this prospective pilot study was to determine whether intravenous piperacillin–tazobactam is feasible as antibacterial prophylaxis, and whether it could reduce infection after autologous peripheral blood stem cell transplantation (PBSCT). After completing the pilot study, we compared the results with those observed in a group of patients who had received standard quinolone prophylaxis, using a matched-pair, case–control comparison design.

Patients and methods Patients and study drug regimen

Correspondence: Dr C Solano, Haematology and Medical Oncology Department, University Clinic Hospital, Blasco Iban˜ez 17, 46010 Valencia, Spain; E-mail: [email protected] Received 18 August 2004; accepted 28 February 2005 Published online 23 May 2005

Between August 1999 and November 2000, 51 consecutive patients with hematological malignancies or solid tumors undergoing autologous PBSCT in the Haematology and Medical Oncology Department of the University Clinic Hospital of Valencia, were enrolled in a prospective pilot study of intravenous piperacillin–tazobactam antibacterial prophylaxis (Study Group) after hematopoietic

Prophylactic piperacillin–tazobactam in stem cell transplant C Solano et al

60

transplantation. Inclusion criteria were the following: age 18–70 years, no major organ dysfunction, Karnofsky performance status X80%, no history of allergic reaction to betalactams, no fever or active infection, and at least 7 days free from antimicrobial therapy. Patients were asked to participate in the study and signed informed consent forms. These patients did not receive any oral antibacterial prophylaxis. On the day of hematopoietic transplantation (day 0), intravenous piperacillin/tazobactam (fixed-combination: 4/0.5 g every 8 h) was initiated and maintained until ANCX0.5 109/l. In the case of neutropenic fever, amikacin (15 mg/kg, i.v., every day) was included in the regimen.

Controls and matching variables A matched-pair control group was obtained by selecting one of the most recently transplanted control patients for each study patient from our historical PBSCT database of patients treated with standard quinolone antibacterial prophylaxis. Controls were individually matched to cases on a 1 : 1 ratio. For purposes of matching, every patient in the study group shared the following pretransplant characteristics with the corresponding patient receiving quinolone prophylaxis (control group): (1) age75 years; (2) sex; (3) Karnofsky performance status X80%; (4) underlying disease and phase of the disease. Matching was by neutropenic episode and not necessarily by individual patient. To avoid any potential selection bias, matching was blinded, and only the patient’s initials and pretreatment variables were known. Upon admission, usually on day 8, the control group started on norfloxacin (400 mg every 12 h p.o.). Quinolone prophylaxis was maintained until the first neutropenic fever appeared, when it was discontinued and broad-spectrum empiric antibiotic therapy consisting of ceftazidime (2 g, i.v., every 8 h) and amikacin (15 mg/kg, i.v., every day) was begun. Treatment was maintained until neutropenia resolved and the patient was afebrile for at least 5 days with no infectious symptoms. A new febrile episode was defined as the reappearance of a temperature of X381C after an afebrile period of X7 consecutive days, not related to transfusion or infusional reaction to drugs.

than 4 days despite broad spectrum empiric antibiotics, systemic antifungal therapy was initiated in both groups (study and control patients) with empirical conventional amphotericin B (AmB) (0.5 mg/kg, i.v.). In the case of proven or suspected fungal disease, antifungal therapy with AmB (1 mg/kg, i.v.) was initiated. In the case of infusional reaction, renal toxicity (creatinine X2 mg/dl) or clinical progression of the suspected or proven fungal disease, conventional AmB was substituted by a lipid-associated AmB.

Antibiotic susceptibility testing method Susceptibility testing was performed either by disk diffusion or broth micro dilution as described by the National Committee for Clinical Laboratory Standards (NCCLS).19

Conditioning and supportive treatment All patients with the same diagnosis received one of the following preparative regimens: cyclophosphamide (Cy) (120 mg/kg) and total body irradiation (TBI) (12 Gy, six fractions); busulfan (Bu) (16 mg/kg), and Cy (120 mg/kg) (BuCy) or melphalan (140 mg/m2) (BuMel); Cy (6 g/m2), thiotepa (500 mg/m2) and carboplatin (800 mg/m2) (STAMPV); BCNU (300 mg/m2), etoposide (800 mg/m2), cytosine-arabinoside (800 mg/m2), and melphalan (140 mg/ m2) (BEAM); carboplatin (1.6 g/m2), etoposide (1.8 g/m2) and Cy (6.4 g/m2) (CarboPEC). The conditioning regimen was followed by the infusion of cryopreserved peripheral blood apheresis product as the sole source of stem cells. Peripheral blood stem cell collection and processing was carried out by standard G-CSF-based mobilization (filgrastim, Amgen, Thousand Oaks, CA, USA) at a mean dose of 10 mg/kg/day by subcutaneous injection for 5 to 8 days. G-CSF (filgrastim, 5 m/kg/day by subcutaneous injection) starting on day þ 5 until the ANC was greater than 1 109/l for 3 consecutive days was administered post transplant. Single-donor or random platelet transfusions were administered for a platelet count of less than 10 109/l and RBC transfusion for a hemoglobin level lower than 8.0 g/dl.

General antimicrobial measures All patients were isolated in single rooms with a high efficiency filtered air (HEPA) system, received low microbial content food and were managed under strict control measures by hospital personnel to reduce nosocomial pathogen acquisition. On admission, surveillance cultures (throat and nose swabs, urine and stool cultures) were taken in order to determine the resident flora and resistance profile, and all patients received oral antiviral (acyclovir, 1200 mg/day) and antifungal prophylaxis consisting of daily fluconazole (200 mg/day). In the case of severe mucositis (4WHO grade 2), both drugs were administered intravenously. In the case of neutropenic fever without a microbiologically or clinically documented infection persisting more Bone Marrow Transplantation

Definitions Engraftment. Blood counts were performed daily until hospital discharge and thereafter, at least once a week for the first 3 months. Granulocyte engraftment was defined as the first of 3 consecutive days with an ANC X0.5 109/l. Platelet engraftment was defined as the first of 3 consecutive days with an unsupported count X20 109/l. Days of antibiotics/antifungal. Summation of the days of individual administration of each antibiotic/antifungal. TRM.

Death not related to disease progression.


61

Infection-related death (IRD). Death occurring in any time of the study having an infection as the primary cause.

Study design and parameters analyzed After completing the pilot study in 51 patients, we performed a pair-matched, case–control single center study. The main parameters analyzed between groups were: (1) incidence of febrile neutropenia (defined as an axilar temperature 4381C in a patient with therapy-induced neutropenia o0.5 109/l neutrophils); (2) incidence of bacteremia (defined as two or more positive blood cultures for coagulase-negative staphylococci (CNS) or at least one positive blood culture for other Gram-positive bacteria or GNB) during the first febrile episode (FFE) after the infusion of stem cells; (3) incidence of bacteremia by CNS, other Gram-positive organisms, Enterobacteriaceae and other GNB during the FFE; (4) antibiotic susceptibility to piperacillin–tazobactam, fluorinated quinolones and other broad-spectrum antibiotics; (5) duration of the FFE; (6) incidence of fever of unknown origin (FUO), defined as fever without a microbiologically or clinically documented infection; (7) evaluation of the admission time and days of antibiotic-antifungal administration; and (8) empiric or therapeutic use of AmB during the neutropenic period; and (9) incidence of IRD.

Statistical methods To compare patient characteristics between both groups we used the w2 test (with the Fisher exact test, where available) for categorical variables. For continuous variables we used the t-test for independent samples with a previous test for Table 1

equality of variances. When normality for the continuous variables could not be assumed using the corresponding normality test, we used the Mann–Whitney nonparametric test. The level of significance in all cases was set at Po0.05. All analyses were carried out using SPSS statistical software.

Results Patient characteristics Study and control group characteristics are listed in Table 1. There were no significant differences between both groups with respect to pretreatment characteristics, thus indicating the comparability of patients.

Infections Table 2 shows the results for all patients and for patients with febrile episodes. Overall, 76% of the study group (piptazo) and 98% of the control group (quinolone) developed at least one febrile episode during neutropenia (P ¼ 0.002). The pattern of appearance of the FFE was somewhat different in both groups. We observed a trend toward an earlier appearance of the FFE in the control group than the study group. In total, 60% of the control group patients had their FFE before the 5th day after the stem cell infusion, whereas 52% of the patients of the study group had their FFE after the 6th day post transplantation (P ¼ 0.08). Moreover, despite a similar duration of neutropenia in both groups, the time from neutropenia to fever was significantly longer in the study than in the control group (median 5 vs 4 days for the study and control groups, respectively) (P ¼ 0.04). However, once a febrile

Patient characteristics All patients

Study group (Pip-Tazo)

Control group (quinolones)

P

102 Jan-96/Nov-00 49 (18–68)

51 Aug-99/Nov-00 49 (23–68)

51 Jan-96/July-99 48 (18–65)

NS

Gender Male Female

35 (34%) 67 (66%)

16 (31%) 35 (69%)

19 (37%) 32 (63%)

Diagnosis AML/ALL NHL/HD/MM ST

10 (10%) 40 (39%) 52 (51%)

5 (10%) 20 (39%) 26 (51%)

5 (10%) 20 (39%) 26 (51%)

Phase of disease Early (1st CR, 1stCPh) Intermediate (X2ndCR) Advanced (PR, PD)

47 (46%) 37 (36%) 18 (18%)

22 (43%) 17 (33%) 12 (24%)

25 (49%) 20 (39%) 6 (12%)

Treatment administered TBI–Cy Bu–Cy/Bu–Mel BEAM STAMP–V/CarboPec

2 22 26 52

1 11 13 26

1 11 13 26

Number of patients Date of transplant Median age (range)

NS

NS

NS

NS (2%) (22%) (25%) (51%)

(2%) (22%) (25%) (51%)

(2%) (22%) (25%) (51%)

Abbreviations: AML ¼ acute myeloblastic leukemia; ALL ¼ acute lymphoblastic leukemia; NHL ¼ non-Hodgkin’s lymphoma; HD ¼ Hodgkin’s disease; MM ¼ multiple myeloma; ST ¼ solid tumor including breast carcinoma and germ cell carcinoma; CR ¼ complete remission; PR ¼ partial remission; CPh ¼ chronic phase; PD ¼ progressive disease; TBI ¼ total body irradiation; Cy ¼ cyclophosphamide; Bu ¼ busulphan; Mel ¼ melphalan; BEAM: includes BCNU, etoposide, Ara-C and melphalan; NS ¼ not significant.

Bone Marrow Transplantation


62 Table 2

Overall results All patients (n ¼ 102) 89 1 5 19 30 24 16 3 22 24 19 24

Patients with febrile neutropenia No. of febrile episodes during admission, median (range) Onset of FFE, median days after infusion (range) Day 0–3 Day 4–5 Day 6–7 Day 47 Duration of fever, median days (range) 1 day 2–3 days 4–5 days 45 days Duration of ANC o0.5 109/l, median days (range) Time from ANC o 0.5 109/l to fever, median days (range) Classification of the FFE FUO Clinically documented Microbiologically documented Bacteremia during the FFE (% of FFE) Number of microorganisms per bacteremia 1 2 Septic shock associated with the FFE All patients Days of admission, median (range) Days of antibiotics, median (range) Use of AmB Empirically For suspected/proven invasive fungal infections Prophylaxis due to previous fungal infection Early TRM Infection-related death FFE ¼ first febrile episode; ANC ¼ absolute AmB ¼ amphotericin B; NS ¼ not significant.

neutrophil

(87%) (0–3) (0–13) (21%) (34%) (27%) (18%) (1–16) (25%) (27%) (21%) (27%)

39 1 6 8 11 8 12 5 4 9 10 16

(76%) (0–3) (0–13) (21%) (28%) (21%) (31%) (1–16) (10%) (23%) (26%) (41%)

10 (5–30) 4 (0–15)

10 (5–30) 5 (0–15)

35 8 46 41

11 5 23 21

(39%) (9%) (52%) (46%)

Control group (Quinolone) (n ¼ 51) 50 1 5 11 19 16 4 2 18 15 9 8

(98%) (0–3) (0–9) (22%) (38%) (32%) (8%) (1–13) (36%) (30%) (18%) (16%)

(28%) (13%) (59%) (54%)

24 3 23 20

18 (90%) 2 (10%) 0

24 25 25 23 1 1 1

25 22 16 14 1 1 1

23 28 9 9

count;

(17–47) (0–115) (25%) (23%) (1%) (1%) (1%) 0 FUO ¼ fever

of

(17–47) (0–109) (31%) (27%) (2%) (2%) (2%) 0

unknown

o0.001

origin;

NS 0.04

(48%) (6%) (46%) (40%)

20 (95%) 1 (5%) 1 (3%)

P 0.002 NS NS

9 (6–29) 4 (0–9)

38 (93%) 3 (7%) 1 (1%)

episode appeared, the duration of fever was significantly longer in cases than in controls, with a median of 5 and 2 days (oP ¼ 0.001), respectively. There were no statistically significant differences in the classification of the FFE or in the type and location of the documented infections (data not shown). Positive blood cultures were documented during the FFE in 21 patients in the pip-tazo group and 20 in the control group (54 vs 40% of the patients who developed a FFE in each group, respectively (P ¼ 0.43)). The incidence of polymicrobial bacteremia was the same in both groups. Overall, there were no significant differences in hospital stay between cases and controls, with a median of 25 and 23 days, respectively (P ¼ 0.2). However, the analysis of the number of days on antibiotics revealed a statistically significant difference with medians of 28 days/antibiotic (cases) and 22 days/antibiotic (controls) (P ¼ 0.05). The use of AmB empirically, despite the fact that they were more frequent in the study group, did not achieve statistical significance (P ¼ 0.13). The analysis of the number of days of AmB administration showed a greater duration in the control group, with a median of 10 days (range 5–16) compared with 7 days (range 4 to 25) in the study group, although statistical significance was not achieved (P ¼ 0.2). Bone Marrow Transplantation

Study group (Pip-Tazo) (n ¼ 51)

NS NS

NS

(18–47) NS (0–115) 0.05 (18%) NS (18%) 0 0 0 0

TRM ¼ transplant

related

mortality;

One patient in the study group (2%) died on day 16 post transplant due to pulmonary and retroperitoneal hemorrhage associated with progressive disease without any microbiological documentation. Patients without febrile neutropenia, both in the study group and in the control group, had no significant differences in pretransplant characteristics compared with the whole group (data not shown). Table 3 shows the isolates observed in the microbiologically documented episodes according to treatment group. There were no differences in the rates of CNS, Staphylococcus aureus or viridans streptococci in either group or in the incidence of Enterobacteriaceae and other GNB.

Antibiotic susceptibilities of bacterial isolates Table 4 shows the antibiotic susceptibilities of isolated bacteria. Most of the isolates showed a high degree of susceptibility. Overall, 85% of all episodes of Grampositive bacteremia were susceptible to the antibiotics tested and there were no significant differences between the study and control groups. The overall rate of susceptible GNB was also high, but it was higher in the study group than in the control group especially with


63 Table 3

Pathogens isolated in patients with bacteremia during FFE, according to treatment group All patients (n ¼ 54) 45 37 4 2 2

Gram-positive bacteremia (% of isolates) CNS Streptococcus viridans Corynebacterium spp S. aureus Gram-negative bacteremia (% of isolates) Escherichia coli Campylobacter spp Enterobacter cloacae Salmonella spp

9 5 2 1 1

Study group patients (Pip-Tazo) (n ¼ 28)

(83%) (%) (%) (%) (%)

23 18 3 1 1

(%) (%) (%) (%) (%)

(82%) (%) (%) (%) (%)

5 (%) 3 (%) 0 1 (%) 1 (%)

Control group patients (Quinolone) (n ¼ 26) 22 19 1 1 1

(85%) (%) (%) (%) (%)

P NS

4 (%) 2 (%) 2 (%) 0 0

NS

CNS ¼ coagulase-negative staphylococci; NS ¼ not significant.

Table 4

Antibiotic susceptibilities of isolated bacteria (in % of total isolates) according to the type of therapy received Total

Gram-positive bacteria Fluoroquinolonesa Cephalotine Imipenem Piperacillin–Tazobactam Amikacin

38/44 38/45 35/45 39/44 42/45

Gram-negative bacteria Fluorquinolonesa Imipenem Piperacillin–Tazobactam Ceftazidime Amikacin

6/9 8/9 6/9 9/9 9/9

Study group (Pip-Tazo) (Quinolones)

(86%) (84%) (78%) (89%) (93%)

(67%) (89%) (67%) (100%) (100%)

22/23 21/23 18/23 20/23 21/23

5/5 5/5 3/5 5/5 5/5

(96%) (91%) (78%) (87%) (91%)

(100%) (100%) (60%) (100%) (100%)

Control group (Quinolones) 16/21 17/22 17/22 19/21 21/22

1/4 3/4 3/4 4/4 4/4

(76%) (77%) (77%) (91%) (95%)

P NS NS NS NS NS

(25%) (75%) (75%) (100%) (100%)

a

Susceptibilities to norfloxacin and ciprofloxacin did not differ in any isolate. NS ¼ not significant.

respect to susceptibility to fluoroquinolones, although the low total number of Gram-negative isolates precludes statistical analysis. In spite of the low number of patients, the quinolone group presented a higher rate of quinolone resistance, as previously described.20,21

Response to treatment and toxicity There was response (definitive disappearance of the responsible microbiological isolate) to the empiric antibiotic combination initiated after development of fever in the control group or the addition of amikacin to piperacillin–tazobactam in patients with microbiological documentation in 12 patients (52%) and 10 patients (43%), respectively, at 72 h of treatment; in 19 patients (83%) and 20 patients (87%), respectively, at 7 days; in 22 patients (96%) and 23 patients (100%), respectively, at the end of treatment. Given that patients treated with hematopoietic transplantation can exhibit significant side effects related with chemotherapy and support treatment, patients were evaluated daily using the WHO (World Health Organization) Toxicity Criteria. No significant differences were detected between the study and the control groups in any of the categories analyzed. Particularly, no cases of drug fever,

cytopenias, nephrotoxic or allergic reactions potentially related to the prophylaxis were identified. Four patients in the Study Group and five patients in the Control Group presented mild (WHO grade 1 or 2), transient transaminase elevations during treatment period were interpreted as possibly related to the prophylactic treatment because of temporal association without coadministration of other well-know hepatotoxic drugs.

Discussion Infection is one of the most common life-threatening complications of hematopoietic transplantation. Half of all infectious complications occur during the first 4–6 weeks following hematopoietic grafting, and most of them are bacterial in origin.1–4 In spite of prophylaxis with oral antibiotics (mainly fluoroquinolones) used by some groups, most of the patients (90–100%) who receive myeloablative hematopoietic transplantation develop fever22 and need initiating empirical antibiotic therapy that usually continues until neutropenia resolves. The use of empirical intravenous antibiotic therapy may actually prevent subsequent infections as reported in some studies by Bone Marrow Transplantation


64


1.0

0.8 Cumulative incidence of FFE

reducing the risk of a documented infection regardless of the level of neutropenia.1,23 This observation and the limited overall effectiveness of quinolone prophylaxis led some investigators to study the use of prophylactic intravenous antibiotic after hematopoietic transplantation, both in patients admitted to hospital16,17,24 and in outpatients receiving autologous transplantation.18,25–28 Avril et al16 prospectively studied 60 children treated for solid tumors with high-dose chemotherapy followed by bone marrow transplantation, who were randomly assigned to receive prophylaxis consisting of ceftazidime plus teicoplanin beginning before the onset of aplasia and fever (group A) or exactly the same antibiotic regimen but beginning at the onset of fever (group B). They observed a reduction in the incidence of septicemia (6.6 vs 24%) in group A, mainly due to the prevention of episodes of early onset. Similarly, the appearance of the first episode of fever was delayed in group A, and the overall duration was reduced. Kroschinsky et al17 analyzed in a pilot study 50 patients undergoing autologous or allogeneic stem cell transplantation, who received a prophylactic antibacterial treatment with intravenous piperacillin/tazobactam beginning on day of stem cell infusion. They were analyzed retrospectively for frequencies of FUO, documented infection, bacteremia and death because of infection. Data from microbiological monitoring and tolerability were evaluated. Compared to oral decontamination and chemoprophylaxis, an intravenous prophylactic with this regimen was considered effective and well-tolerated in prevention of bacterial infections and related complications. These preliminary studies suggested that the approach could be effective in preventing early bacterial infections related complications and cost-effective in this special setting of neutropenia of intermediate and well-predicted duration. Piperacillin–tazobactam is a broad-spectrum antibiotic with an excellent antibacterial activity and safety profile, which is frequently used to treat febrile neutropenia in the high-risk patient.22,29 Our results demonstrate that systemic bacterial prophylaxis with piperacillin–tazobactam after hematopoietic transplantation is feasible and, when retrospectively compared with controls, they suggest modification of the usual pattern of infection. The reduction in the frequency of neutropenic fever (Figure 1) together with the earlier presentation of the FFE and the shorter duration of fever once it appeared suggest an effect of the systemic prophylaxis on the earliest bacterial infections after transplant, regardless of the level of neutropenia. There were no significant differences in the classification of the FFE or in the location, microbiological documentation or incidence of infection-associated shock. The early TRM (100 days) was very low (2%) with no IRD. The susceptibility analysis detected an increased incidence of fluoroquinolone-resistant strains in the control group along with a possible crossresistance to other antibacterial agents of unrelated classes such as betalactams, as previously reported,20 although the low number of isolates precluded statistical analysis. In the study group, bacterial isolates did not show such a resistance pattern. In the study group a nonsignificant higher incidence of empirical use of AmB was observed, probably reflecting the

0.6

0.4

0.2

0.0 0

10

20

30

Days after transplantation

Figure 1 Cumulative incidence of the day of appearance of the FFE in the study group (dotted line) and control group (solid line) (Po0.001).

longer duration of the FFE once it appeared. However, the duration of this AmB treatment was shorter in the study group than in the control group. This finding, together with the absence of documented fungal infection, suggests that, at least in this group of patients, systemic prophylaxis did not represent a risk factor for significant invasive fungal infections, one potential risk of a prolonged use of broadspectrum antibiotics. The analysis of the most important infection-related parameters, and the subset of patients without fever, did not suggest a greater benefit in one set than another. In summary, although it cannot be considered a standard approach, our study suggests that systemic prophylaxis with piperacillin–tazobactam can be effective in preventing early bacterial infections and related complications in this special setting of neutropenia of intermediate and wellpredicted duration associated with hematopoietic stem cell transplantation in patients with solid tumors or hematological malignancies. The absence of detection of an increased incidence of beta-lactam or crossresistance in this study should not obviate concern over the emergence of resistant strains. On the contrary, close monitoring and antibiotic rotation should remain priority issues. The efficacy and absence of drawbacks shown in this particular setting is probably not applicable to other patient populations with neutropenia of long duration such as those being treated with induction or consolidation chemotherapy for acute leukemia. Whether this observation can be reproduced in prospective randomized studies, will determine the clinical and epidemiological relevance of this treatment in patients at high risk of severe infection.

References 1 Engels EA, Ellis CA, Supran SE et al. Early infection in bone marrow transplantation: quantitative study of clinical factors that affect risk. Clin Infect Dis 1999; 28: 256–266.


65 2 Kruger W, Russmann B, Kroger N et al. Early infections in patients undergoing bone marrow or blood stem cell transplantation – a 7 year single centre investigation of 409 cases. Bone Marrow Transplant 1999; 23: 589–597. 3 Barton T, Collis T, Stadtmauer E, Schuster M. Infectious complications the year after autologous bone marrow transplantation or peripheral stem cell transplantation for treatment of breast cancer. Clin Infect Dis 2001; 32: 391–395. 4 Saavedra S, Jarque I, Sanz GF et al. Infectious complications in patients undergoing unrelated donor bone marrow transplantation: experience from a single institution. Clin Microbiol Infect 2002; 8: 725–733. 5 Buzyn A, Tancrede C, Nitenberg G, Cordonnier C. Reflections on gut decontamination in hematology. Clin Microbiol Infect 1999; 5: 449–456. 6 Kerr KG. The prophylaxis of bacterial infections in neutropenic patients. J Antimicrobial Chemother 1999; 44: 587–591. 7 Sepkowitz KA. Antibiotic prophylaxis in patients receiving hematopoietic stem cell transplant. Bone Marrow Transplant 2002; 29: 367–371. 8 Cruciani M, Rampazzo R, Malena M et al. Prophylaxis with fluoroquinolones for bacterial infections in neutropenic patients: a meta-analysis. Clin Infect Dis 1996; 23: 795–805. 9 Horvathova Z, Spanik S, Sufliarsky J et al. Bacteremia due to methicillin-resistant staphylococci occurs more frequently in neutropenic patients who received antimicrobial prophylaxis and is associated with higher mortality in comparison to methicillin-sensitive bacteriemia. Int J Antimicrobial Agents 1998; 10: 55–58. 10 Elting LS, Bodey GP, Keefe BH. Septicemia and shock syndrome due to viridans streptococci: a case–control study of predisposing factors. Clin Infect Dis 1992; 14: 1201–1207. 11 Tunkel AR, Sepkowitz KA. Infections caused by viridans streptococci in patients with neutropenia. Clin Infect Dis 2002; 34: 1524–1529. 12 Yeh SP, Hsueh EJ, Yu MS et al. Oral ciprofloxacin as antibacterial prophylaxis after allogeneic bone marrow transplantation: a reappraisal. Bone Marrow Transplant 1999; 24: 1207–1211. 13 van Kraaij MG, Dekker AW, Peters E et al. Emergence and infectious complications of ciprofloxacin-resistant Escherichia coli in haematological cancer patients. Eur J Clin Microbiol Infect Dis 1998; 17: 591–592. 14 Oethinger M, Conrad S, Kaifel K et al. Molecular epidemiology of fluoroquinolone-resistant Escherichia coli bloodstream isolates from patients admitted to European cancer centers. Antimicrobial Agents Chemother 1996; 40: 387–392. 15 Baum HV, Franz U, Geiss HK. Prevalence of ciprofloxacinresistant Escherichia coli in hematologic-oncologic patients. Infection 2000; 28: 278–281. 16 Avril M, Hartmann O, Valteau-Couanet D et al. Antiinfective prophylaxis with ceftazidime and teicoplanin in children undergoing high-dose chemotherapy and bone marrow transplantation. Pediatr Hematol Oncol 1994; 11: 63–73.

17 Kroschinsky F, Wichmann G, Bornhauser M et al. Efficacy and tolerability of prophylactic treatment with intravenous piperacillin/tazobactam in patients undergoing hematopoietic stem cell transplantation. Transpl Infect Dis 2002; 4: 132–136. 18 Meisenberg B, Gollard R, Brehm T et al. Prophylactic antibiotics eliminate bacteremia and allow safe outpatient management following high-dose chemotherapy and autologous stem cell rescue. Support Care Cancer 1996; 4: 364–369. 19 National Committee for Clinical Laboratory Standards (NCCLS). Methods for Dilution Antimicrobial Susceptibility Tests for Bacteria that Grow Aerobically, Approved Standard, 5th edn. Clinical and Laboratory Standards Institute: Wayne, PA, 2000 (http://www.nccls.org). 20 Bauernfeind A, Abele-Horn M, Emmerling P, Jungwirth R. Multiclonal emergence of ciprofloxacin-resistant clinical isolates of Escherichia coli and Klebsiella pneumoniae. J Antimicrobial Chemother 1994; 34: 1074–1076. 21 Sanders CC, Sanders Jr WE, Goering RV, Werner V. Selection of multiple antibiotic resistance by quinolones, beta-lactams, and aminoglycosides with special reference to cross-resistance between unrelated drug classes. Antimicrobial Agents Chemother 1984; 26: 797–801. 22 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. 23 Karp JE, Dick JD, Angelopulos C et al. Empiric use of vancomycin during prolonged treatment-induced granulocytopenia. Randomized, double-blind, placebo-controlled clinical trial in patients with acute leukemia. Am J Med 1986; 81: 237–242. 24 Petersen F, Thornquist M, Buckner C et al. The effects of infection prevention regimens on early infectious complications in marrow transplant patients: a four arm randomized study. Infection 1988; 16: 199–208. 25 Jagannath S, Vesole DH, Zhang M et al. Feasibility and cost-effectiveness of outpatient autotransplants in multiple myeloma. Bone Marrow Transplant 1997; 20: 445–450. 26 Meisenberg BR, Ferran K, Hollenbach K et al. Reduced charges and costs associated with outpatient autologous stem cell transplantation. Bone Marrow Transplant 1998; 21: 927–932. 27 Chandrasekar PH, Abraham OC, Klein J et al. Low infectious morbidity after intensive chemotherapy and autologous peripheral blood progenitor cell transplantation in the outpatient setting for women with breast cancer. Clin Infect Dis 2001; 32: 546–551. 28 Rovira M. Outpatient management of autologous haematopoietic cell transplantation: the Barcelona experience. Presse Med 2004; 33: 479–481. 29 Sanz MA, Lopez J, Lahuerta JJ et al. Cefepime plus amikacin versus piperacillin–tazobactam plus amikacin for initial antibiotic therapy in haematology patients with febrile neutropenia: results of an open, randomized, multicentre trial. J Antimicrobial Chemother 2002; 50: 79–88.
