If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
PURPOSE: Several studies have documented the efficacy of amphotericin B as empiric antifungal therapy in cancer patients with prolonged fever and neutropenia. Amphotericin, however, is a toxic drug. Fluconazole has broad-spectrum antifungal activity with an excellent safety profile. Although prophylactic use of fluconazole is widespread, its efficacy as an empiric antifungal agent has not been extensively investigated.
PATIENTS AND METHODS: We randomly assigned 106 patients with absolute neutropenia (≤500 cells μL) and persistent fever of undetermined origin (>38°C) despite 1 week of broad-spectrum antibiotic therapy to receive either fluconazole 400 mg orally daily or amphotericin B 0.5 mg/kg/day. Patients with obvious invasive fungal infections were excluded, as were those with abnormal renal or hepatic function. Success was defined as defervescence with the initially assigned antifungal regimen without development of clinically evident invasive fungal infection.
RESULTS: Six patients were excluded from the analysis, mostly because they did not have severe neutropenia. Forty-eight patients received amphotericin B, and 52 received fluconazole. Baseline clinical characteristics and laboratory parameters as well as duration of neutropenia (7.7 versus 6.9 days), duration of fever (7.8 versus 8.1 days), and duration of hospitalization (10.4 versus 8.3 days) were similar between those receiving amphotericin and fluconazole. Treatment success rates and mortality rates were similar in the two groups: 22 (46%) patients in the amphotericin group and 29 (56%) patients in the fluconazole group responded successfully to therapy (P = 0.3), whereas 16 (33%) patients in the amphotericin group and 14 (27%) patients in the fluconazole group died during hospitalization (P = 0.5). Adverse events such as chills and fever (4 versus 1), bronchospasm (2 versus none), severe hypokalemia (25 versus 12) and nephrotoxicity (9 versus 3) were more frequently observed in patients receiving amphotericin. Adverse prognostic factors included prolonged duration of neutropenia and pneumonia.
CONCLUSIONS: These results suggest that fluconazole is an equally effective but less toxic alternative to amphotericin B as empiric antifungal therapy in cancer patients with prolonged fever and neutropenia.
Empiric use of systemic antibiotics has reduced the morbidity and mortality associated with febrile neutropenia (
). Patients with prolonged neutropenia, however, remain at a risk of persistent or recurrent fever. In patients without any other obvious cause, occult fungal infections are the most likely source of fever (
). Often this poor survival is attributable to delayed diagnosis. For example, only 25% to 40% of patients with disseminated candidiasis have positive blood cultures. Furthermore, inability to perform biopsies or other invasive diagnostic procedures in patients with bleeding diathesis and the lack of reliable serologic tests make early diagnosis difficult (
). Persistent or recurrent fever in a neutropenic patient being treated with broad-spectrum antibiotics may be the only clinical evidence of an invasive fungal infection.
Difficulty in diagnosis, and the poor outcome of those with established fungal infection, have led to the empiric use of antifungal agents in patients with prolonged fever and neutropenia. Empiric therapy with amphotericin B results in defervescence, a reduction in subsequent fungal infections, and a concomitant decrease in mortality from infectious causes (
). Favorable pharmacokinetics, a good safety profile, reliable absorption from the gastrointestinal tract, and availability in oral and parenteral formulations make fluconazole an attractive antifungal agent. It has proven efficacy against Cryptococcus neoformans and several candida species and has been used with variable success against blastomycosis, histoplasmosis, sporotrichosis, and coccidioidomycosis (
). It has, however, no substantial activity against aspergillus. Fluconazole has been used to treat patients with oro-pharyngeal and esophageal candidiasis, cryptococcal meningitis, and chronic disseminated candidiasis (
). Prophylactic use of fluconazole prevents colonization and development of superficial infection with candida in patients with leukemia and reduces the incidence of both systemic and superficial fungal infections in those undergoing bone marrow transplantation (
) have recently demonstrated that fluconazole is equally effective but less toxic than amphotericin B for prophylaxis in patients with leukemia. Serious side effects with the use of fluconazole are rare and limited to nausea, vomiting, anorexia, rash, and mild elevation of hepatic enzyme levels. We conducted a randomized trial to compare the efficacy and toxicity of fluconazole with amphotericin B as empiric antifungal agents in cancer patients with prolonged fever and neutropenia.
Patients and methods
This study was confined to adult cancer patients (≥16 years of age) with an absolute neutrophil count of ≤500/μL; persistent fever of 38°C or greater for 48 hours or more; at least 7 days of appropriate broad-spectrum antibiotic therapy (most often ceftazidime plus amikacin); and absence of any obvious source of fever, such as resistant bacterial infection, abscess, or infection with atypical microorganisms. Patients with known intolerance to the study drugs, evidence of systemic fungal infection, preexisting renal insufficiency (serum creatinine level ≥2.5 mg/dL), hepatic insufficiency (serum transaminase levels ≥4 times upper limit of normal or serum bilirubin level ≥3 mg/dL), or with recurrent fever of undetermined etiology were excluded. Pregnant or lactating women or patients who had used systemic antifungal drugs within the previous 14 days were also excluded. Informed consent was obtained from all patients, and the study was approved by the Institutional Human Subjects Committee. This study was carried out at two institutions.
Patients with fever and neutropenia despite prolonged (≥7 days) use of broad-spectrum antibacterial therapy and without any evident source of infection were randomly assigned to receive either fluconazole 400 mg orally per day or amphotericin B 0.5 mg/kg/day. In patients with severe mucositis or diarrhea, fluconazole was given intravenously in the same dosage. These drugs were planned to be given for a minimum of 4 days and, if successful, to be continued until neutropenia recovered or the patient remained afebrile for 4 days without any evidence of systemic fungal infection. Patients were maintained on the same dose of amphotericin B unless they developed fungemia or systemic fungal infection. In that case, amphotericin dose was increased to 1 or 1.5 mg/kg/day. Saline loading was not routinely done unless patients developed nephrotoxicity.
During the study period, patients were continued on broad-spectrum antibiotics, and changes were made as required. No other antifungal therapy was administered.
Pretreatment requirements and follow-up during the study period
A complete history was obtained and physical examination was performed in all patients with particular attention to potential causes of fever. Complete blood counts, serum transaminase, creatinine and electrolyte levels were obtained. Throat, urine, stool, and at least 2 sets of blood cultures, including fungal cultures, were obtained in all patients. Cultures from other sites that appeared to be the cause of fever and fine needle aspiration of suspected areas were also carried out. Chest radiographs were taken.
During therapy, temperature was determined every 4 hours and physical examination was performed at least once daily. Blood cultures were drawn, at least every other day, in patients who remained febrile. Similarly, repeat urine, throat, stool, and other cultures were taken at regular intervals. Laboratory tests were repeated on a daily basis or at least every other day. Patients were kept in strict reverse isolation, and fresh unpeeled fruits and vegetables were avoided.
Response to therapy and evaluation of adverse events
The endpoint of therapy in this study was defervescence (oral temperature of 37.6°C or less throughout the remaining period of neutropenia). Time to defervescence was documented in all patients. Every attempt was made to determine the underlying cause of fever and categorize it as clinical or microbiologic infection versus fever of undetermined origin. Response to therapy was described as successful when defervescence occurred with the initially assigned regimen without development of systemic fungal infection throughout the study period. Patients who had no change in fever pattern, developed signs and symptoms of systemic mycosis, developed serious side effects that required discontinuation of therapy, or died from any cause during the study period were considered as failures. Whenever possible, the single most likely cause of death was determined.
All adverse events were recorded, including time, severity, most likely cause, and outcome. Nephrotoxicity was defined as a rise in serum creatinine level of 0.5 mg/dL or more compared with the baseline value. Hepatic toxicity was defined as an increase in transaminase or alkaline phosphatase values more than twice the pretreatment level. Electrolyte imbalance was categorized as correctable when replacement of electrolytes allowed continuation of therapy, or refractory when therapy had to be discontinued. Other toxicities, such as drug rash, were individually determined. In patients developing nephrotoxicity, appropriate dose modifications were made.
Data were analyzed with chi-square and Fisher’s exact tests or with Student’s t test, as appropriate. All P values are two tailed. Multivariate associations were tested with logistic regression analyses. Time to defervescence was compared using Kaplan-Meier curves. Statistical significance was set at P <0.05.
Clinical characteristics of the study patients
One hundred and six patients were enrolled in the study. Six patients were excluded from the analysis, four patients because their neutrophil count was >500/μL. Two patients, on further investigation, had history of allergy to amphotericin. Forty-eight patients received amphotericin B, and 52 received fluconazole. Baseline clinical characteristics, such as sex, age, underlying disorder, and clinical signs and symptoms, and laboratory values, were similar in the two groups (Table 1).
Table 1Clinical Characteristics of the Evaluable Patients at Enrollment
At the time therapy was terminated, the duration of fever, neutropenia and antifungal therapy were similar in the two groups (Table 2). Patients treated with fluconazole had a slightly, but not significantly, shorter duration of hospitalization than those who received amphotericin. Overall, the cause of fever remained undetermined in almost half of the patients. Pneumonia with negative cultures was the predominant clinical infection. Several patients developed bacterial or fungal infections; however, there were no marked differences between the two groups (Table 3).
Table 2Clinical Characteristics of Patients at Termination of Therapy
Overall success was obtained in half of the patients (Table 4). Time to treatment failure was similar between the two groups (Figure 1). Death accounted for the majority of the failures, as 30% of the patients died during the study period, usually due to pneumonia and respiratory failure. Intracranial bleeding accounted for a small number of deaths. No significant differences were observed between the two groups.
Chills and fever were frequently observed in patients receiving amphotericin. These symptoms were successfully treated with antipyretics, antihistamines, and when necessary, meperidine. No patient was withdrawn from the study for this reason. Two patients developed immediate hypersensitivity reaction (flushing, hypotension, bronchospasm) to amphotericin and had to be withdrawn from the study. Severe electrolyte imbalance, nephrotoxicity, and hepatotoxicity were more often observed in patients receiving amphotericin (Table 5). Thirty-two (67%) patients receiving amphotericin and 19 (36%) patients receiving fluconazole experienced adverse side effects. The difference between the percentage of patients with adverse side effects in the two groups of 31% is highly significant (95% confidence interval [CI] 12%–49%; P <0.01).
Several clinical variables, including dyspnea at the time of randomization, persistent neutropenia, duration of antifungal therapy, radiologic evidence of pneumonia, and development of clinical or microbiologic cause of infection were associated with failure of therapy (Table 6). In a multivariable analysis, persistent neutropenia (odds ratio [OR] = 19; 95% CI: 5.2–68), and pneumonia (OR = 16; 95% CI: 3.5–71) were independent predictors of failure.
). In this setting, it causes defervescence and reduces the subsequent development of invasive mycoses, as well as mortality from infectious causes. Most patients treated with amphotericin, however, develop side effects (
), has not been extensively investigated as an empiric antifungal agent. Our study suggests that fluconazole may be an equally effective and less toxic alternative to amphotericin B in these patients. However, the total number of study patients, and particularly the number who developed clinically evident fungal infection, were small.
A recent study confirmed the efficacy of fluconazole in nonneutropenic patients with candidemia (
). Moreover, the prophylactic use of fluconazole in patients expected to have prolonged neutropenia reduces the incidence of both superficial and invasive candidiasis and decreases the number of deaths attributable to fungal infections (
). More widespread use of fluconazole can be expected to increase the magnitude of this problem. Restricting the use of fluconazole to those requiring empiric antifungal therapy may reduce the development of widespread resistance to this agent.
We attempted to identify factors that were associated with poor outcome. Patients with dyspnea at the time of randomization had poor prognosis, probably because of impending pneumonia; subsequent analysis confirmed that pneumonia was an adverse prognostic factor. Persistent neutropenia was also identified as an important prognostic factor, as previously reported (
). Successful outcome with antifungal therapy is associated with recovery of the neutrophil count. A different therapeutic approach may be necessary in patients with adverse prognostic factors.
One of the major concerns with the use of fluconazole, as compared to amphotericin, relates to its lack of activity against aspergillus. The incidence of aspergillosis varies widely between different institutions (
). In centers with a high incidence of aspergillus infection, extreme caution is required before undertaking such an approach. Another limitation of this trial was that we used an open-label study design, because differences in the nature of study drugs made blinding difficult. However, every effort was made to apply similar criteria in determining therapeutic failure.
In conclusion, our study suggests that fluconazole is an equally effective but less toxic alternative to amphotericin B for empiric antifungal therapy of cancer patients with prolonged febrile neutropenia. If confirmed by other investigators, this approach may be more cost-effective than prophylactic use of fluconazole with the additional benefit of reduced chances of colonization and infection with the resistant organisms.
We are grateful to Mr. Hamad Yousuf for typing the manuscript.
Concept of empiric therapy with antibiotic combinations.