¹Vanderbilt University, Nashville, TN, USA

²Baylor University Medical Center, Dallas, TX, USA

³University of Louisville, Louisville, KY, USA

⁴Jewish Hospital, Cincinnati, OH, USA

⁵Cleveland Clinic, Cleveland, OH, USA

⁶West Virginia University, Morgantown, WV, USA

⁷University of Texas Health Science Center at San Antonio, San Antonio, TX, USA

Correspondence to: Dr S N Wolff, Bone Marrow Transplant Program, 2617 TVC, Vanderbilt University, Nashville, TN 37232-5505, USA

systemic fungal infections are a major problem in bone marrow transplant recipients who have prolonged neutropenia or who receive high-dose corticosteroids. prophylaxis with fluconazole or low-dose amphotericin b reduces, but does not eliminate these infections. to determine which prophylactic agent is better, we performed a prospective randomized study. patients undergoing allogeneic (related or unrelated) or autologous marrow or peripheral stem cell transplantation were randomized to receive fluconazole (400 mg/day p.o. or i.v.) or amphotericin b (0.2 mg/kg/day i.v.) beginning 1 day prior to stem cell transplantation and continuing until recovery of neutrophils to >500/l. patients were removed from their study drug for drug-associated toxicity, invasive fungal infection or suspected fungal infection (defined as the presence of fever >38°C without positive culture while on broad-spectrum anti-bacterial antibiotics). Proven or suspected fungal infections were treated with high-dose amphotericin B (0.5-0.7 mg/kg/day). Patients were randomized at each institution and stratified for the type of transplant. The primary end-point of the study was prevention of documented fungal infection; secondary endpoints included fungal colonization, drug toxicity, duration of hospitalization, duration of fever, duration of neutropenia, duration and total dose of high-dose amphotericin B and overall survival to hospital discharge. From July 1992 to October 1994, a total of 355 patients entered into the trial with 159 patients randomized to amphotericin B and 196 to Fluconazole. Patient groups were comparable for diagnosis, age, sex, prior antibiotic or antifungal therapy, use of corticosteroids prior to transplantation and total duration of neutropenia. Amphotericin B was significantly more toxic than Fluconazole especially in related allogeneic transplantation where 19% of patients developed toxicity vs 0% of Fluconazole recipients (p < 0.05). approximately 44% of all patients were removed from prophylaxis for presumed fungal infection. proven fungal infections occurred in 4.1% and 7.5% of fluconazole and amphotericin-treated patients, respectively. proven fungal infections occurred in 9.1% and 14.3% of related allogeneic marrow recipients receiving fluconazole or amphotericin b, respectively, and 2.1% and 5.6% of autologous marrow recipients receiving fluconazole or amphotericin b, respectively (P > 0.05). In this prospective trial, low-dose amphotericin B prophylaxis was as effective as Fluconazole prophylaxis, but Fluconazole was significantly better tolerated. Bone Marrow Transplantation (2000) 25 , 853-859.

fungal infection prophylaxis; Fluconazole; amphotericin B

Intense cytotoxic therapy combined with hematopoietic stem cell rescue is now front-line or salvage treatment for a variety of malignancies. However, this form of treatment is associated with an abundance of severe side-effects, either directly from the cytotoxic therapy or arising from the ensuing myelosuppression.^1,2 Foremost among these toxicities is infection associated with neutropenia.^3,4,5,6 These infectious consequences remain despite reduction in the duration of neutropenia with the use of hematopoietic growth factors such as G-CSF or GM-CSF.

Opportunistic fungal infections were uncommon prior to the initiation of dose-intensive therapy.⁷ Two clinical situations predominate - fungemia and dissemination with yeast forms and pulmonary infection with hyphal organisms.^2,8 The contribution of neutropenia to fungal infections is noted by approximately a 15% incidence after therapy of acute leukemia or marrow transplantation.^9,10,11 Fungal infections require long-term treatment and even with prompt initiation of therapy, lead to substantial mortality.¹² Overall, fungal infections account for more than 50% of infection-related mortality during neutropenia.¹³

The standard of therapy for serious fungal infections is amphotericin B, although its administration can be limited by side-effects.¹⁴ Lipid-associated or liposomal amphotericin B provides some protection against renal toxicity and infusion-related chills and fever.^{15,16,17,18,19} The triazoles are a newer class of anti-fungal agents with a much improved toxicity profile compared with amphotericin B.²⁰

A number of prophylactic approaches have been implemented to protect neutropenic patients from fungal infections.^{21,22,23,24,25,26,27} One study used low-dose amphotericin B (0.1 mg/kg/day) to prevent fungal infections after autologous marrow transplantation and showed a reduction in colonization but not invasive infection.²⁸ Aerosolized amphotericin B, suggested to be useful for prophylaxis, was not confirmed as effective in a randomized evaluation.^29,30 A recent study using 2 mg/kg/day of liposomal amphotericin three times a week demonstrated little toxicity, but did not show a significant reduction in fungal infections.³¹ Three large-scale series have compared Fluconazole and oral placebo, either in patients with acute leukemia or those undergoing bone marrow transplantation.^9,10,11 These trials demonstrated a substantial reduction in the incidence of colonization and/or invasive fungal infection and showed that Fluconazole is an appropriate agent for fungal prophylaxis in patients with prolonged neutropenia. However, it is still possible that aggressive empiric use of anti-fungals would be as successful as prophylaxis. Additionally, amphotericin B is an antifungal with a greater spectrum of activity including Aspergillus species and yeast forms, such as Candida kruseii, that are resistant to Fluconazole.³²

Here, we describe a trial conducted by The North American Marrow Transplant Group (NAMTG) comparing Fluconazole with amphotericin B for antifungal prophylaxis in patients undergoing dose-intensive therapy. Our study indicates that Fluconazole and amphotericin B have similar efficacy but Fluconazole is less toxic.

Methods

The study was a prospective, randomized, stratified and non-blinded trial comparing antifungal prophylaxis with amphotericin B with Fluconazole. Patients were stratified into three groups; matched related allogeneic marrow (allo) transplants, matched unrelated allogeneic (MUD) transplants and autologous marrow or stem cell (PBSC) transplants. These strata were chosen due to the anticipated difference in fungal infection rate with each type of transplant. Patients were randomized at each study center.

The primary endpoint of the study was the incidence of proven fungal infection. Secondary endpoints were suspected fungal infection, duration of fever, duration of hospitalization, duration of absolute neutropenia <500/l, fungal colonization of skin and gastrointestinal tract, use of intensified empiric amphotericin B therapy for suspected fungal infection, toxicities of anti-fungal prophylaxis, and patient survival to the time of hospital discharge. Follow-up of patients after hospital discharge was not undertaken. These endpoints were established prospectively.

The trial was conducted by the North American Marrow Transplant Group, a multi-center collaborative group. The following institutions entered patients into the trial: Baylor University Medical Center, Dallas, Texas; Cleveland Clinic, Cleveland, Ohio; Jewish Hospital, Cincinnati, Ohio; University of Louisville, Louisville, Kentucky; University of Texas Health Science Center at San Antonio, San Antonio, Texas; Vanderbilt University, Nashville, Tennessee; and West Virginia University, Morgantown, West Virginia.

Patients eligible for entry into the trials were required to be 18 years old, give written informed consent approved by the treating institution's IRB, be undergoing dose-intensive therapy and hematopoietic stem cell (marrow or peripheral blood) transplantation with an anticipated duration of neutropenia of at least 10-14 days, have no clinical evidence of infection at the time of study entry, have no known allergy or intolerance to the study drugs, have no laboratory evidence of significant hepatic or renal dysfunction (defined as a SGOT or SGPT 4 ´ upper limit of normal, total bilirubin 3 ´ upper limit of normal and serum creatinine 2 ´ upper limit of normal) and not be actively receiving antibiotics or non-topical anti-fungal therapy with either Fluconazole or amphotericin B. All patients were required to receive prophylactic antibiotics active against gram-negative bacteria until resolution of neutropenia. The specific antibiotic for gram-negative prophylaxis, addition of additional antibiotics for fever and the use of prophylactic antibiotics to prevent gram-positive infection was defined by standard care at each treatment center.

Fluconazole was administered at a dose of 400 mg/day orally or intravenously for patients unable to tolerate oral administration. Amphotericin B was administered intravenously at a daily dose of 0.2 mg/kg (based on ideal body weight) with a maximum dose of 20 mg. Both drugs were begun at the conclusion of the cytotoxic therapy (1 day prior to stem cell infusion) and continued until recovery of neutrophils (> 500/l), unless a severe adverse reaction due to the prophylactic drug occurred.

Prophylaxis was stopped and high-dose amphotericin B (0.5-0.7 mg/kg/day) was initiated with documentation of any proven invasive fungal infection. A proven fungal infection required positive culture from a normally sterile source or histologic evidence of invasive fungal elements in tissue section. Empiric high-dose amphotericin B replaced prophylactic treatment for suspected fungal infection. Suspected fungal infection was defined as fever >38°C persisting for 3-5 days without positive blood cultures. There were thus three reasons for failure of prophylaxis: (1) adverse event, (2) proved fungal infection and (3) suspected fungal infection.

Toxicity was graded using the NAMTG marrow transplant toxicity system which is similar to the toxicity scale of Bearman et al.³³ This system grades toxicities from 0-4 using the following generalization: grade 0 = none; grade 1 = mild; grade 2 = severe; grade 3 = life-threatening; and grade 4 = fatal. In our study, the major toxicity noted was renal and the definitive scale for this organ system is:Renal grade 0: BUN 20 mg/dl or creatinine 1.4 mg/dl, creatinine clearance >90% of baseline;Renal grade 1: BUN 21-40 mg/dl or creatinine = 1.5-2.0 mg/dl creatinine clearance 76-90% of baseline;Renal grade 2: BUN 41-75 mg/dl or creatinine = 2.1 to 4.0 mg/dl, creatinine clearance 50-75% of baseline;Renal grade 3: creatinine clearance <50% of baseline, BUN >75 mg/dl or creatinine >4.0 mg/dl, symptomatic renal failure requiring dialysis, nephrotic syndrome;Renal grade 4: fatal renal toxicity.

Statistical design and analysis

The trial was designed to detect a 10% difference in fungal infection incidence (primary endpoint) with an error of 0.05 and error of 0.10 (power of 90%) assuming the incidence of infection in the inferior group was 15%. A total of 187 patients were scheduled to enter each treatment arm. Proportions were compared by the method of Simon and required confirmation by chi-square analysis. Group comparison for continuous variables were compared with t-tests or Wilcoxon rank sum test.

Results

From 21 July 1992 to 16 October 1994 a total of 355 patients were entered into the trial. The distribution of patients at each treatment center and strata is shown in Table 1. Imbalance at one treatment center (Baylor University Medical Center) was due to a misinterpretation of the autologous randomization codes. Patient diagnoses (data not shown) were of equal distribution (P > 0.05) between the two prophylactic groups. Additional demographic parameters are shown in Table 2. None of these comparisons are statistically different except creatinine (P = 0.05) among related allogeneic recipients receiving amphotericin B (median 0.91 mg/dl, range 0.5-1.4) compared with Fluconazole-treated patients (median 0.77 mg/dl, range 0.5-1.2). However, all creatinine values (as required by the eligibility criteria) were normal (1.4 mg/dl). Therefore, all treatment groups were comparable for pre-treatment features that could influence development of fungal infection or toxicity.

The most critical feature for the development of fungal infection is the duration of neutropenia. Table 3 indicates that the distribution of neutropenia was similar between treatments for all groups and strata.

Overall, 47.4% (95% confidence interval 40.6-54.4%) and 57.9% (95% confidence interval 50.1-65.3%) of all patients receiving prophylaxis with Fluconazole and amphotericin B respectively, stopped prophylaxis early. These frequencies did not differ significantly either globally or by type of marrow transplant. Duration of prophylaxis is also shown in Table 3 and did not differ between amphotericin B or Fluconazole or by transplant strata.

Table 4 reviews the reasons noted for discontinuation of prophylactic anti-fungal therapy. As shown, 43.9% (95% confidence interval 38.9-49.1%) of all patients had their prophylaxis discontinued because of persistent fever while on broad-spectrum anti-bacterials. Documented fungal infection developed during prophylaxis in 2.3% (95% confidence interval 1.1-4.4%) including 2.6% and 1.9% (P > 0.05) of patients receiving Fluconazole or amphotericin B, respectively. Neither the rate of documented fungal infection nor of suspected fungal infection differed significantly between Fluconazole or amphotericin B groups globally or by strata as shown in Table 4.

However, drug toxicity leading to discontinuation of prophylaxis was statistically more frequent in amphotericin B recipients. Twenty such patients (12.6%, 95% confidence interval of 8.3-18.6%) developed toxicity with 13 having renal toxicity. Renal toxicity was categorized as grade 2 in nine patients and grade 3 in four. Three patients required dialysis while receiving amphotericin B. The baseline creatinine values of the 13 patients who developed renal toxicity were normal with a median of 0.75 mg/dl (range 0.5-1.1) and not different from the Fluconazole-treated patients (median 0.82 mg/dl, range 0.4-1.6). The other seven toxicities were six episodes of infusion-related fever and chills and one episode of hepatic dysfunction. By contrast, only one Fluconazole recipient (0.5%, 95% confidence interval 0.1-2.8%) undergoing autologous transplantation developed drug toxicity (skin rash) requiring termination of prophylaxis. In analysis of the individual strata, 19%, 42.9% and 8.3% of patients undergoing allogeneic related, allogeneic unrelated and autologous transplants, respectively, developed toxicity while receiving prophylactic amphotericin B. The incidence of renal toxicity was 16.3% in all allogeneic recipients and 4.5% in autologous recipients (P < 0.05).

Overall, 5.6% (95% confidence intervals of 3.7-8.5%) of the patients entered into the study ultimately developed a documented fungal infection including 4.1% and 7.5% (P > 0.05) of patients receiving Fluconazole or amphotericin B, respectively (Table 5). These included infections noted during and after discontinuation of prophylaxis. The frequency of infection did not differ between prophylactic groups or by type of marrow transplantation. Fungal infection in recipients of amphotericin B prophylaxis was caused by Candida albicans in eight patients, Candida glabrata in one patient, Candida parapsilosis in two patients and Aspergillus sp. in one patient. In recipients of Fluconazole prophylaxis, fungal infection was caused by Candida albicans in two patients, Candida glabrata in three patients, Candida parapsilosis in one patient, Candida krusei in two patients and Aspergillus sp. in two patients (two patients had dual infections with Candida glabrata combined with either Aspergillus sp. or Candida krusei. These infections were predominantly blood stream (15) with the remaining pulmonary (four) and genitourinary (one).

One hundred and ten patients receiving amphotericin B and 147 patients receiving Fluconazole underwent surveillance cultures of throat and rectum during prophylaxis. Forty-eight (43.6%, 95% confidence interval 34.7-53.0%) and 40 (27.2%, 95% confidence interval 20.7-34.9%) of the patients receiving amphotericin B and Fluconazole, respectively, developed a positive surveillance culture during prophylaxis (P < 0.01).

Death during therapy occurred in 19 patients (11.9%, 95% confidence interval 7.8-17.9%) receiving amphotericin B and in 24 patients (12.2%, 95% confidence interval 8.3-17.6%) receiving Fluconazole (P > 0.05). Causes of death included progressive neoplastic disease, toxicity, infection and graft-versus-host disease. Two patients in the amphotericin B arm (1.3%, 95% confidence interval 0.3-4.5%) and five patients in the Fluconazole arm (2.6%, 95% confidence interval 1.1-5.8%) died directly from a fungal infection (P > 0.05). Data on the duration of therapeutic amphotericin B, total dose of therapeutic amphotericin B, duration of fever and total duration of hospitalization are shown in Table 5. Comparison of these values between recipients of amphotericin B or Fluconazole did not show any significant differences.

Discussion

The results of this trial are consistent with other reports showing successful prophylaxis of serious fungal infections after dose-intensive therapy for cancer. Utilizing either Fluconazole or amphotericin B, we observed an overall invasive fungal infection rate of 5.6%. Since we did not have an untreated or placebo control group, this number must be compared to historically-derived data. The incidence of fungal infection seen in this trial is quite similar to the rates (2.8-7.0%) seen in the Fluconazole treatment arms of three large randomized trials.^9,10,11 In their placebo control arms, these trials had a fungal infection rate of 8.0-18%.

The ultimate aim of our study was to compare the anti-fungal efficacy of Fluconazole and amphotericin B. The results showed therapeutic equivalence. However, amphotericin B was more toxic especially in allogeneic transplant patients due to a higher rate of renal dysfunction. Considering this toxicity, further evaluation of additional allogeneic transplant, to increase the statistical power for this group of patients, would not seem warranted. The increased toxicity was likely due to the nephrotoxic effect of combining amphotericin B with the immunosuppressive agents cyclosporine or tacrolimus. Patients who developed renal toxicity had normal serum creatinine values (median 0.75 mg/dl, range 0.5-1.1) prior to initiation of prophylaxis.

It still remains unclear whether prophylactic approaches lead directly to improved survival since careful clinical observation and rapid institution of empiric anti-fungal therapy will enable most patients to recover, especially with concurrent recovery of neutropenia.¹⁰ The study by Goodman et al did not show improved overall survival although death due to invasive fungal infection was reduced by prophylaxis. On the other hand, the study by Slavin et al¹¹ in which Fluconazole was continued into the post-engraftment period did show improved survival. Our study, without an untreated control arm, cannot resolve this issue. Negative aspects of prophylaxis, such as the substantial emergence of resistant organisms were not noted in our study.

The choice of amphotericin B dosing was an important aspect of this trial.³⁴ Since our study did not demonstrate the superiority of amphotericin B, it is possible that we used a 'sub-optimal' dose. Supporting this hypothesis is the lack of success of a lipid formulation for prophylaxis when used at doses less than recommended for therapeutic administration.³¹ It is also possible that we could have achieved similar efficacy and avoided toxicity with a lower amphotericin B dose.

The prophylactic use of lipid formulations of amphotericin B might overcome the dose-limiting toxicity of conventional amphotericin B. However, these formulations were not readily available during the design of our trial. In addition, a comparative trial would have to be massive to detect an improvement in the 4.1% rate of fungal infections seen with Fluconazole prophylaxis.

Fungal infections developed in a total of 5.6% of our patients. Most of these infections were due to yeast forms and 19 of the 22 isolates were due to Candida species (10 albicans, two krusei, three parapsilosis and four glabrata). The remainder were due to Aspergillus sp. infections. Thus, most of these infections were likely to be sensitive to Fluconazole and Candida krusei was only isolated in two patients on the Fluconazole arm. Candida krusei is an uncommon isolate even at centers using Fluconazole routinely, and one major outbreak was likely due to nosocomial spread from a point source.³²

Seventy percent of the patients in our trial underwent surveillance cultures, an adequate number for valid observation. Fluconazole was superior in the prevention of colonization with yeasts as detected by oral and rectal surveillance cultures. Since this was not associated with a significant reduction in the actual number of invasive infections, its clinical significance is uncertain. Our results are similar to previous reports using Fluconazole.³⁵

This trial was not designed to evaluate the prevention of fungal infections in long-term follow-up. Aspergillosis remains the most problematic fungal infection after engraftment, at least in allogeneic transplant recipients. We had hoped that prophylactic amphotericin B would offer some protection against Aspergillus as suggested in another study.³⁶ However, during the study only three infections with Aspergillus were noted, divided between prophylactic arms. This may have been due to the aggressive use of empiric high-dose (0.5-0.7 mg/kg/day) of amphotericin B for persistently febrile patients.

The requirement for the initiation of therapeutic dose amphotericin B was planned using a conventional definition of presumed fungal infection, ie 3-5 days of unexplained fever occurring while patients were receiving broad-spectrum anti-bacterial antibiotics. This practice was developed prior to the widespread use of anti-fungal prophylaxis. In our study, only 3.4% of patients were proven to have a fungal infection after discontinuation of prophylaxis. Thus, it is possible that in patients receiving prophylaxis, the introduction of therapeutic doses of amphotericin B could be delayed for longer durations of fever (eg >7-10 days) or for more definitive criteria (eg pulmonary infiltrates without known bacterial cause).³⁷

In summary, we have demonstrated that low-dose amphotericin B (0.2 mg/kg/day) and Fluconazole (400 mg/day) are equivalent for prophylaxis of fungal infection after dose-intensive cytotoxic therapy followed by stem cell transplantation. However, especially in allogeneic transplant recipients, the greater safety of Fluconazole makes it the preferred agent.

1 Khardori N, Bodey, GP. Infections in hematologic malignancies. Hematology 1991; 14: 363-424,

2 Verfaillie C, Weisdorf D, Haake R et al. Candida infections in bone marrow transplant recipients. Bone Marrow Transplant 1991; 8: 177-184, MEDLINE

3 Klastersky J. Febrile neutropenia. Curr Opin Oncol 1993; 5: 625-632, MEDLINE

4 Koll BS, Brown AE. Changing patterns of infections in the immunocompromised patient with cancer. Hematol/Oncol Clin North Am 1993; 7: 753-769,

5 Heurlin N, Bergstrom SE, Winiarski J et al. Fungal pneumonia: the predominant lung infection causing death in children undergoing bone marrow transplantation. Acta Paediatr 1996; 85: 168-172, MEDLINE

6 Gamis AS, Gudnason T, Giebink GS, Ramsay NKC. Disseminated infection with fusarium in recipients of bone marrow transplants. Rev Infect Dis 1991; 13: 1077-1088, MEDLINE

7 Uderzo C, Angelo PD, Rizzari C et al. Central venous catheter-related complications after bone marrow transplantation in children with hematological malignancies. Bone Marrow Transplant 1991; 9: 113-117,

8 Guiot HF, Fibbe WE, van't Wout JW. Risk factors for fungal infection in patients with malignant hematologic disorders: implications for empirical therapy and prophylaxis. Clin Infect Dis 1994; 18: 525-532, MEDLINE

9 Goodman JL, Winston DJ, Greenfield RA et al. A controlled trial of fluconazole to prevent fungal infections in patients undergoing bone marrow transplantation. New Engl J Med 1992; 326: 845-851, MEDLINE

10 Winston DJ, Chandrasekar PH, Lazarus HM et al. Fluconazole prophylaxis of fungal infections in patients with acute leukemia. Results of a randomized placebo-controlled, double-blind, multicenter trial. Ann Intern Med 1993; 118: 495-503, MEDLINE

11 Slavin MA, Osborne B, Adams R et al. Efficacy and safety of fluconazole prophylaxis for fungal infections after marrow transplantation - a prospective, randomized, double-blind study. J Infect Dis 1995; 171: 1545-1552, MEDLINE

12 Kruger W, Sobottka I, Stockschlader M et al. Fatal outcome of disseminated candidosis after allogeneic bone marrow transplantation under treatment with liposomal and conventional amphotericin-B. A report of 4 cases with determination of the Mic values. Scand J Infect Dis 1996; 28: 313-316, MEDLINE

13 Jehu V. Managing fungal and viral infections in the immunocompromised host. Rec Res Cancer Res 1988; 108: 61-70,

14 Winston DJ. Prophylaxis and treatment of infection in the bone marrow transplant recipient. Curr Clin Topics Infect Dis 1993; 13: 293-321,

15 Walsh TJ, Finberg RW, Arndt C et al. Liposomal amphotericin B for empirical therapy in patients with persistent fever and neutropenia. National Institute of Allergy and Infectious Diseases Mycoses Study Group. New Engl J Med 1999; 340: 764-771, MEDLINE

16 Pasic S, Flannagan L, Cant AJ. Liposomal amphotericin (AmBisome) is safe in bone marrow transplantation for primary immunodeficiency. Bone Marrow Transplant 1997; 19: 1229-1232, MEDLINE

17 Chopra R, Fielding A, Goldstone AH. Successful treatment of fungal infections in neutropenic patients with liposomal amphotericin (AmBisome) - a report on 40 cases from a single centre. Leuk Lymphoma 1992; 7: (Suppl1) 73-77, MEDLINE

18 Kruger W, Stockschlader M, Russmann B et al. Experience with liposomal Amphotericin-B in 60 patients undergoing high-dose therapy and bone marrow or peripheral blood stem cell transplantation. Br J Haematol 1995; 91: 684-690, MEDLINE

19 Kruger W, Stockschlader M, Sobottka I et al. Antimycotic therapy with liposomal amphotericin-B for patients undergoing bone marrow or peripheral blood stem cell transplantation. Leuk Lymphoma 1997; 24: 491-499, MEDLINE

20 Fluconazole. Med Lett 1990; 32: 50-52,

21 Gotzsche PC, Johansen HK. Meta-analysis of prophylactic or empirical antifungal treatment vs placebo or no treatment in patients with cancer complicated by neutropenia. Br Med J 1997; 314: 1238-1244,

22 Ehninger G, Schuler HK, Sarnow E. Fluconazole in the prophylaxis of fungal infection after bone marrow transplantation. Mycoses 1996; 39: 259-263, MEDLINE

23 O'Donnell MR, Schmidt GM, Tegtmeier BR et al. Prediction of systemic fungal infection in allogeneic marrow recipients: impact of amphotericin prophylaxis in high-risk patients. J Clin Oncol 1994; 12: 827-834, MEDLINE

24 Powles RL, Milliken S. The prophylaxis of fungal infections. J Antimicrob Chemother 1991; 28: (Suppl. A) 97-103, MEDLINE

25 Quabeck K, Muller KD, Beelen DW et al. Prophylaxis and treatment of fungal infections with fluconazole in bone marrow transplant patients. Mycoses 1992; 35: 221-224, MEDLINE

26 Tollemar J, Ringden O, Andersson S et al. Randomized double-blind study of liposomal amphotericin B (AmBisome) prophylaxis of invasive fungal infections in bone marrow transplant recipients. Bone Marrow Transplant 1993; 12: 577-582, MEDLINE

27 Ringden O, Andstrom EE, Remberger M et al. Prophylaxis and therapy using liposomal amphotericin B (AmBisome) for invasive fungal infections in children undergoing organ or allogeneic bone-marrow transplantation. Pediatr Transplant 1997; 1: 124-129, MEDLINE

28 Perfect JR, Klotman ME, Gilbert CC et al. Prophylactic intravenous Amphotericin B in neutropenic autologous bone marrow transplant recipients. J Infect Dis 1992; 165: 891-897, MEDLINE

29 Schwartz S, Behre G, Heinemann V et al. Aerosolized amphotericin B inhalations as prophylaxis of invasive aspergillus infections during prolonged neutropenia: results of a prospective randomized multicenter trial. Blood 1999; 93: 3654-3661, MEDLINE

30 Hertenstein B, Kern WV, Schmeiser T et al. Low incidence of invasive fungal infections after bone marrow transplantation in patients receiving amphotericin B inhalations during neutropenia. Ann Hematol 1994; 68: 21-26, MEDLINE

31 Kelsey SM, Goldman JM, McCann S et al. Liposomal amphotericin (AmBisome) in the prophylaxis of fungal infections in neutropenic patients: a randomised, double-blind, placebo-controlled study. Bone Marrow Transplant 1999; 23: 163-168, MEDLINE

32 Wingard JR, Merz WG, Rinaldi MG et al. Increase in Candida krusei infection among patients with bone marrow transplantation and neutropenia treated prophylactically with fluconazole. New Engl J Med 1991; 325: 1274-1277, MEDLINE

33 Bearman SI, Appelbaum FR, Buckner CD et al. Regimen-related toxicity in patients undergoing bone marrow transplantation. J Clin Oncol 1988; 6: 1562-1568, MEDLINE

34 De Laurenzi A, Matteocci A, Lanti A et al. Amphotericin B prophylaxis against invasive fungal infections in neutropenic patients: a single center experience from 1980 to 1995. Infection 1996; 24: 361-366, MEDLINE

35 Chandrasekar PH, Gatny CM. The effect of fluconazole prophylaxis on fungal colonization in neutropenic cancer patients. Bone Marrow Transplantation Team. J Antimicrob Chemother 1994; 33: 309-318, MEDLINE

36 Rousey SR, Russler S, Gottlieb M, Ash RC. Low-dose amphotericin B prophylaxis against invasive aspergillus infections in allogeneic marrow transplantation. Am J Med 1991; 91: 484-492, MEDLINE

37 Chandrasekar PH, Gatny CM. Effect of fluconazole prophylaxis on fever and use of amphotericin in neutropenic cancer patients. Bone Marrow Transplantation Team. Chemotherapy 1994; 40: 136-143, MEDLINE

Table 1  Distribution of patients according to treatment center

Table 2  Pre-treatment demographics

Table 3  Overall results

Table 4  Reasons noted for stopping prophylaxis

Table 5  Infection-related outcomes