Iatrogenically-induced neutropenia is the most important risk factor for invasive fungal infections [270, 830]. Three groups of patients comprise the majority of cases in this category [266]:

  • Cancer patients who have received cytotoxic chemotherapy
  • Bone marrow transplant recipients

And to a lesser extent,

  • Solid organ transplant recipients

In addition to neutropenia, these patients are commonly exposed to the (A):usual risk factors for invasive candidiasis such as receipt of multiple broad-spectrum antibiotics, use of central venous catheters, and colonization with Candida.

Neutropenic children and adults with persistent fever despite broad-spectrum antibiotics have a 20% risk of developing an invasive fungal infection [675, 1811]. The most common types of fungal infections include invasive candidiasis and invasive aspergillosis [62]. Infection due to other filamentous fungi (e.g., Fusarium spp. or the agents of zygomycosis) may also be seen.

The empirical use of an antifungal agent in neutropenic patients with unexplained fever starting on or after day five of an appropriate course of antibiotics reduces the likelihood of developing a clinically apparent invasive fungal infection [675, 1811]. Although the data supporting this strategy are limited, this strategy is widely accepted and implemented. Only (A):two randomized studies have been done to assess the effect of the empiric use of intravenous amphotericin B for persistent fever of unknown origin in neutropenic patients. Despite the limited statistical strength of these studies, they are the support for the standard recommendation of the empirical use of amphotericin B to treat possible fungal infections when confronting persistent fever of unknown origin in neutropenic patients [1083, 1915].



Empirical therapy for neutropenic patients is recommended in cases of persistent fever beginning on or after the fifth day of an appropriate antibacterial regimen. This treatment should continue until neutropenia is resolved. During this time, signs and symptoms that would specifically suggest a fungal infection should be carefully sought. The drug doses and regimens discussed below are for empirical therapy of fever of an unknown origin. If a specific etiology is identified, then more targeted therapy should be employed. This has particular implications with respect to infections due to Aspergillus or other filamentous fungi, as these organisms require maximal tolerated dosages of amphotericin B.

The workup of patients with fever in the setting of neutropenia includes at a minimum a complete physical examination, urine cultures, blood cultures, and sputum cultures. A chest radiograph should be obtained as well. If this radiograph is normal, patients at more than trivial risk for an invasive mould infection (e.g., on steroids, underlying disease is leukemia, or bone marrow transplant recipients) should have a high-quality CT scan of the chest performed. In addition, a CT scan of the paranasal sinuses should be performed if there are any symptoms referable to the upper airways. Based on these studies, findings that would suggest a specific referable etiology would include:

  1. Persistent or recurrent fever that coincides with recovery from neutropenia suggests hepatosplenic candidiasis
  2. Ulcerative lesions in the nasal area are suggestive of aspergillosis or N/A(L):zygomycosis
  3. Sinus consolidation by CT, facial tenderness, or facial swelling (these are also suggestive of aspergillosis or zygomycosis)
  4. Pulmonary infiltrates that are not clearly due to a bacterial process, as with sinusitis, are suggestive of aspergillosis or zygomycosis


Special resource: You may also want to refer to the Infectious Disease Society of America-Mycoses Study Group (IDSA-MSG) Practice Guidelines for this syndrome (Use of Antimicrobial Agents in Neutropenic Patients). It is available at the IDSA website.

A suitable empirical regimen for fever in the setting of neutropenia should address both yeast and mould infections. Currently, the following list of antifungal agents are available for this purpose. However, no studies have compared dosages or time of initiation, and very few studies have directly explored the relative merits of one drug versus another.

  1. Amphotericin B

    Recommended dose: 0.5-0.7 mg/kg/day

    Due to its broad and reliable spectrum, N/A(L):amphotericin B has been the traditional antifungal agent for this type of therapy [675, 1811].

    In the first major study of this entity, Pizzo et al. randomly assigned patients with persistent fever after 7 days of treatment with carbenicillin, cephalothin and gentamicin to one of three groups [1811]. The first group had all antimicrobials discontinued. Group 2 continued the same antibiotics, and group 3 had amphotericin B added to the same antibiotic regimen. Results can be summarized in the following table:

Group No. 1 2 3
Discontinue antibiotics Continue antibiotics Add Amphotericin B to antibiotics
N 16 16 18
Documented Infection 9 (56 %) 6 (38 %) 2 (11%)
Fungal Infections 1 5/6 (83 %) 1/2 (50%)
1 1 0
0 1 0
0 1 0
Aspergillosis (DA)
0 1 0
Candidiasis + DA
0 1 0
    Petrielidium boydii
0 0 1

When analyzing these data, it is important to take into consideration that these numbers come from 4 years of prospective and consecutive evaluation of all neutropenic febrile patients at the Pediatric Oncology Branch of the National Cancer Institute. Of the 652 episodes of fever and neutropenia (in 271 patients) that presented during that time (November 1975 to December 1979), only 50 patients (18%) met the criteria for the study of empirical antifungal therapy. However, to further support this strategy, Pizzo et. al also evaluated persistent fever and neutropenia among those patients who had a clinically or microbiologically defined infection at the onset of their febrile episode. Among these 22 patients, amphotericin B was added after a median of 9 days. Fifty percent of these patients defervesced within a median of 3 days after the addition of amphotericin B. In addition, none of the 20 patients treated empirically with amphotericin B within the first two weeks of starting an antibiotic regimen developed fungal infections.

Seven years later, the European Organization for Research on Treatment of Cancer (EORTC) published a similar study that was performed in a collaborative fashion during two consecutive trials of empiric antibiotic therapy for febrile neutropenia [674, 1206]. Patients that remained febrile and neutropenic after 4 days of broad spectrum antibiotics were randomly assigned to receive or not amphotericin B. Results are summarized in the following table:

Group No. 1 2 P
Continue antibiotics alone Add Amphotericin B to antibiotics
N 64 68
Fungal Infections 6 (9 %) 1 (1%) 0.1
      Candidemia 2 1
      Severe OPC 2
      Pulmonary Aspergillosis 1
      Disseminated mucomycosis 1
Fatal cases 4 0 0.05

Although neither trial yielded a statistically significant result, the overall results were sufficiently convincing that this strategy became the standard of care.

  • Liposomal Amphotericin B (Ambisome)

    Recommended dose: 3 mg/kg/day

    Almost two decades after the previously disscused studies, Walsh et al. published a study comparing a lipid-preparation of amphotericin B with the regular formulation [2357]. In this multicenter study, febrile neutropenic patients were randomized after five days of empirical antibacterial therapy to receive either N/A(L):liposomal amphotericin B (3.0 mg/kg/day) or conventional amphotericin B (0.6 mg/kg/day). While the global results showed that the two regimens were comparable in overall outcome, liposomal amphotericin B was favored in terms of safety and tolerance. In addition, there was a strong trend towards superiority in its effectiveness reducing breakthrough fungal infections.

Group No. 1 2 P
Liposomal Amphotericin B Conventional Amphotericin B
N 343 344
Proven Fungal Infections 11 (3.2 %) 27 (7.8%) 0.009
      Aspergillosis 6 12
      Candidemia 3 12
      Fusariosis 1 1
      Cryptococcosis 0 1
      Zygomycosis 1 0
      Others 0 1


  • Other lipid-associated preparations of amphotericin B

    Although not specifically studied in this setting, there is little reason to believe that N/A(L):ABLC or N/A(L):ABCD would not also be effective. The optimal dosage is not known, but 3 mg/kg/d would appear rational.


  • Itraconazole

    N/A(L):Itraconazole is to date the only azole with an antifungal spectrum broad enough to represent a suitable general-purpose alternative for this type of therapy [290]. Although the somewhat unpredictable bioavailability of its oral formulations (capsules and cyclodextrin solution) has limited its use in the past [1518, 1531], the new IV formulation resolves this problem. Choice of this agent should be made with care. If the patient has previously received an azole antifungal agent as prophylaxis, such therapy could have been selected for infection with an azole-resistant strain or species of Candida. In such a case, therapy with an amphotericin B-based preparation would be preferred.


  • Fluconazole

    Due to its limited activity against Aspergillus and the other filamentous fungi, fluconazole is not usually a good choice as therapy in this setting. In addition, the frequent use of fluconazole as a prophylactic agent further reduces its viability as a choice. However, these considerations become irrelevant if (a) the patient is at low risk for infection due to Aspergillus or other filamentous fungi, (b) the patient lacks any signs or symptoms suggesting that the current fever might be due to Aspergillus or other filamentous fungi, (c) local epidemiology suggests that the patient is at low risk for infection with azole-resistant isolates of Candida, and (d) the patient has not received an azole antifungal agent as prophylaxis. In this case, fluconazole at 400 mg/day has been used successfully in selected patients [1421, 2331, 2447], and could be considered as an alternative strategy.