Invasive Fungal Infections in the Neutropenic Cancer Patient

from Infections in Medicine ^®
Posted 09/05/2002

Andreas H. Groll, MD, Thomas J. Walsh, MD

Abstract and Introduction

Abstract

Invasive fungal infections are important causes of morbidity and mortality in cancer patients with prolonged neutropenia following dose-intensive chemotherapy or hematopoietic stem cell transplantation. Recent epidemiologic trends indicate a shift toward infections by Aspergillus species, non-albicans Candida species, and previously uncommon fungal pathogens that have decreased susceptibility to current antifungal agents. In the last decade, much progress has been made in establishing disease definitions and paradigms for antifungal intervention and in the design and conduct of interventional clinical trials. This article reviews current approaches to prevention and treatment of opportunistic fungal infections in neutropenic patients and discusses novel approaches to antifungal chemotherapy and supportive measures.

Introduction

Invasive fungal infections are important causes of morbidity and mortality in patients with hematologic malignancies and those who are undergoing hematopoietic stem cell transplantation (HSCT). The most significant risk factors in these settings are prolonged and profound neutropenia and therapy with high doses of corticosteroids. The overall frequency of invasive fungal infections in patients with acute leukemia and following allogeneic HSCT (the patient populations at highest risk) is between 10% and 25%; the overall case fatality rate exceeds 50% and is close to 100% in disseminated infections or persistent neutropenia. While Aspergillus and Candida species traditionally account for the majority of documented infections, recent epidemiologic trends indicate a shift toward infections by Aspergillus species, non-albicans Candida species, and previously uncommon fungi that often have little susceptibility to current antifungal agents.^[1-4]

For many years, the antifungal arsenal consisted of amphotericin B desoxycholate (D-AmB) and 5-fluorocytosine (5-FC). Therapeutic alternatives only emerged with the clinical development of fluconazole and itraconazole in the late 1980s. In the past 10 years, however, we have witnessed a significant expansion in antifungal drug research, which is reflected by the introduction of the lipid formulations of amphotericin B (amphotericin B colloidal dispersion, amphotericin B lipid complex, and liposomal amphotericin B [L-AmB]) and the development of novel echinocandin derivatives (anidulafungin, caspofungin, and micafungin) and improved antifungal triazoles (posaconazole, ravuconazole, and voriconazole).^[5,6]

Increased awareness among physicians, improved blood culture techniques, and the advent of high-resolution imaging techniques have had considerable impact on improving the clinical diagnosis of invasive fungal infections, and major progress has been made in harmonizing disease definitions, in defining paradigms for antifungal intervention, and in designing and implementing clinical trials.^[7,8] Despite these advances, however, invasive fungal infections remain difficult to diagnose and to manage, and there is a continuing and urgent need for improved diagnosis, treatment, and prevention.

Approaches to Prevention

Investigation of preventive approaches in patients with acute leukemia and/or following allogeneic HSCT is important, given the facts that our ability to diagnose fungal infections in these patients in a timely manner is limited and that once infection occurs there may be rapid clinical deterioration and high morbidity and mortality. Careful adherence to infection control measures, particularly meticulous hand washing, is the mainstay for the prevention of nosocomial Candida infections.^[9] High-efficiency particulate air filters and maintenance of positive air pressure in patient care areas can be effective for prevention of infection by airborne molds; construction activity in or near these areas requires an array of specific barrier measures to avoid massive exposure of high-risk patients to infective conidia.^[10]

Primary Chemoprophylaxis

Effective primary chemoprophylaxis of invasive opportunistic fungal infections has been shown for Candida species in the setting of marrow transplantation.^[11,12] A randomized, double-blind, placebo-controlled trial in mostly allogeneic marrow recipients has shown that fluconazole, given at 400 mg/d from the start of the conditioning regimen until day 75, can reduce the frequency of invasive Candida infections and lower mortality at day 110.^[12] Moreover, 8 years after completion of the study, there was persistent protection against invasive candidiasis and Candida-related death and a decreased frequency of severe, gut-related graft-versus-host disease (GVHD). There was also an overall survival benefit of 17% in fluconazole-treated patients; this benefit in survival was independent of the underlying condition and the occurrence of relapses.^[13]

In less risk-selected patients with hematologic malignancies who were undergoing remission-induction chemotherapy, both fluconazole (400 mg/d) and itraconazole cyclodextrin (2.5 mg/kg orally bid) have been shown to be effective in preventing systemic infection and death caused by Candida species.^[14,15] A potential drawback of prophylaxis with antifungal triazoles may be the selection of resistant Candida species: Emergence of fluconazole-resistant Candida glabrata and Candida krusei infections has been reported from individual centers,^[16,17] and in a large European multicenter survey, antifungal prophylaxis with fluconazole in patients with hematologic malignancies was significantly associated with infections by non-albicans Candida species.^[2]

However, a recently published study of 585 patients receiving fluconazole prophylaxis showed a low incidence of breakthrough candidemia and a low attributable mortality despite frequent colonization with fluconazole-resistant Candida species.^[18] Nevertheless, the selection and nosocomial spread of azole-resistant Candida isolates appear inevitable and remain a matter of continued concern.

Unfortunately, effective chemoprophylaxis against infections by Aspergillus species has not been demonstrated thus far.^[19] Apart from studies that compare prophylaxis with itraconazole or current investigational agents with prophylaxis with fluconazole, clinical trials are under way that investigate preventive approaches in patients with allogeneic HSCT during phases of aggressive immunosuppression for chronic GVHD.

Empiric Therapy

Among neutropenic patients with hematologic malignancies and patients who have had allogeneic HSCT, those with persistent or recurrent fever despite broad-spectrum antibiotic therapy are considered to be at high risk for invasive fungal infections. In this setting, broad-spectrum empiric antifungal therapy provides antifungal prophylaxis and early therapy for clinically occult invasive infections^[20,21] that may arise despite prophylaxis with antifungal azoles. In the United States, agents approved for this indication include D-AmB (0.6 mg/kg/d) and L-AmB (3 mg/kg/d).

In 2 large, randomized, multicenter trials, 1 of which included patients following allogeneic HSCT, L-AmB was found to be as effective as conventional amphotericin B and was associated with less infusion-related toxicity, less nephrotoxicity,^[22,23] and fewer proven breakthrough fungal infections.^[23] Efficacy equivalent to that of D-AmB has also been demonstrated for itraconazole^[24] and fluconazole^[25] in patients with hematologic malignancies not treated with allogeneic HSCT. However, since fluconazole is not active against filamentous fungi, its use in patients at highest risk for these infections is controversial at best.

More recently, a large, randomized, multicenter trial was completed that compared voriconazole, a novel broad-spectrum triazole, with L-AmB for empiric antifungal therapy.^[26] The results of this study showed comparable composite success rates but fewer proven and probable breakthrough infections and less infusion-related toxicity and nephrotoxicity in the voriconazole-treated cohort. However, patients who received voriconazole had more frequent episodes of transient visual disturbances and hallucinations. Trials are currently under way that investigate the role of other novel triazoles and of antifungal echinocandins for empiric therapy.

Approaches to Treatment

Invasive Candida Infections

Invasive Candida infections can be classified as candidemia or acute disseminated candidiasis with or without fungemia, and they arise from the entry of the organism into the bloodstream from colonized mucosal surfaces or catheters. Approximately 50% of patients who have hematologic malignancies or have undergone HSCT are colonized with Candida species at baseline. Without chemoprophylaxis, proven invasive candidiasis develops in 15% of these patients. This condition has a case fatality rate of 40% to 50% that may increase to 100% with deep tissue involvement.^[27] The introduction of fluconazole had a major impact on the epidemiology of Candida infections in high-risk cancer patients: Recent trends in the allogeneic HSCT population indicate an overall decrease in invasive candidiasis and the emergence of non-albicans Candida species as predominant invasive isolates.^[18]

Successful management of invasive Candida infections relies on early diagnosis and prompt institution of effective treatment. Current blood culture detection techniques such as the lysis-centrifugation and the BacT/Alert system can detect candidemia earlier and more frequently than conventional systems; however, candidemia is only one manifestation of invasive candidiasis, and most tissue-invasive infections are not reliably detected by blood culture techniques and may require invasive diagnostic procedures. For tissue-invasive Candida infections, ultrasonography, high-resolution CT, and MRI have become indispensable tools for detection, monitoring, and guidance of diagnostic procedures.^[28,29] Nonculture techniques, particularly nucleic acid amplification-based systems, may complement existing blood culture systems not only for early detection but also for determining resistance patterns to antifungal agents.^[30]

D-AmB, 1 mg/kg/d IV, is the initial standard therapy in neutropenic patients with blood cultures positive for a yeast-like organism^[5] (Table 1). If the organism is subsequently identified as C albicans, therapy with fluconazole, 400 to 800 mg/kg/d IV, may be considered, provided that the patient has uncomplicated fungemia and has not received systemic prophylaxis with antifungal azoles.^[31] Despite its merely fungistatic activity in vitro, experimental and clinical data support the usefulness of fluconazole for treatment of uncomplicated candidemia in neutropenic patients who are hemodynamically stable.^[5] Nevertheless, in patients who have undergone allogeneic HSCT, the role of fluconazole is limited because of its widespread use for antifungal prophylaxis.

In this setting, breakthrough infections with fluconazole-resistant Candida species, including C glabrata, C krusei, and fluconazole-resistant C albicans, are highly likely. Therefore, D-AmB remains the current agent of choice for most HSCT patients with blood cultures positive for a yeast-like organism. Indwelling vascular catheters should be removed when feasible,^[32] and ophthalmoscopy and abdominal imaging should be performed on recovery from neutropenia to head off loss of vision and to rule out chronic disseminated candidiasis.

If the patient displays signs of acute disseminated candidiasis with hemodynamic instability, persistent fungemia, and evidence of tissue invasion, therapy with high dosages of D-AmB (1.5 mg/kg/d IV) plus 5-FC (initial dosage, 100 mg/kg/d divided in 3 or 4 doses) is recommended^[5]; the use of fluconazole in patients who have acute disseminated candidiasis is controversial. Patients who cannot tolerate conventional amphotericin B or whose infections are refractory to it are candidates for one of the amphotericin B lipid formulations.^[33-35] In the near future, the echinocandins and second-generation triazoles will likely become very important additions to the chemotherapeutic arsenal against invasive candidiasis.^[6]

Patients in whom chronic disseminated candidiasis develops pose particular therapeutic challenges. However, this condition is not an absolute contraindication to further intensive chemotherapy or stem cell transplantation. In 2 small observational studies, the majority of patients (73% to 87%) showed continuing improvement with continuing antifungal therapy.^[36,37]

Invasive Aspergillosis

The cornerstones of management of invasive aspergillosis include an early presumptive diagnosis, the prompt institution of effective antifungal chemotherapy, attempts to secure definite microbiologic diagnosis, and surgical interventions as appropriate.

Invasive aspergillosis primarily affects the lungs and, to a lesser extent, the paranasal sinuses. It is initiated by exposure to airborne conidia of the organism. All around the globe, a steady increase of invasive aspergillosis has been observed^[1]; in some centers, the organism has become the most common cause of pneumonic death in allogeneic stem cell recipients.^[38] As exemplified by Aspergillus terreus,^[39] non-fumigatus Aspergillus species may be less susceptible to amphotericin B, underscoring the need for an exact microbiologic diagnosis and the development of predictive in vitro testing methods.

The advent of high-resolution CT technology has permitted improved detection of pulmonary infiltrates consistent with invasive pulmonary aspergillosis.^[40,41] However, although peripheral nodules, the halo sign, and cavitation are characteristic of pulmonary aspergillosis, these findings are not entirely specific, and in early phases, nonspecific air-space consolidation is common.^[42] Therefore, a microbiologic diagnosis by fiberoptic bronchoscopy with bronchoalveolar lavage or by biopsy procedures should be attempted whenever feasible.

Serial monitoring of galactomannan antigen and Aspergillus-specific nucleic acid sequences in blood^[43,44] also may contribute to the detection of invasive pulmonary aspergillosis, particularly in the neutropenic host, and warrants further investigation. Conceptually similar to the techniques used for diagnosing cytomegalovirus disease, these novel non-culture-based techniques may permit further refinement of the population at highest risk and initiation of therapy on a preemptive basis. Carefully designed clinical trials will be needed to determine the role of such preemptive strategies in comparison with fever-based empiric antifungal therapy and primary chemoprevention.^[7]

The historical standard for induction chemotherapy for both suspected and proven invasive aspergillosis consists of high-dose (1 to 1.5 mg/kg/d) D-AmB^[5] (Table 2). Data from open-label clinical trials have demonstrated that the lipid formulations of amphotericin B are overall less nephrotoxic and at least as effective as D-AmB, and that they can be effective when D-AmB is not.^[33,35,45] They are indicated when preexisting or arising nephrotoxicity or concomitant nephrotoxic agents preclude high-dose D-AmB therapy or when treatment with D-AmB appears to fail.^[46] Based on the concentration- and dose-dependent activity of amphotericin B in vitro and in animal models^[46] and the dosages in clinical studies that led to their approval, use of the highest approved dosages of the lipid formulations for treatment of suspected or documented infections is strongly advocated.

A recently completed open, randomized comparison of the investigational triazole voriconazole and D-AmB followed by other licensed antifungal agents for primary therapy for invasive aspergillosis demonstrated superior antifungal efficacy and improved survival at week 12 in the voriconazole arm.^[47] Provided that FDA approval is obtained for this use, the results of this pivotal study may ultimately lead to the replacement of D-AmB as the gold standard for induction therapy for invasive aspergillosis.

Despite response rates similar to those of D-AmB, the experience with itraconazole in either formulation for induction therapy in profoundly neutropenic patients with suspected or proven invasive aspergillosis is limited.^[48-50] However, itraconazole has an important role for consolidation therapy for invasive aspergillosis and for therapy for infections caused by certain dematiaceous molds.^[4]

Current data from open-label trials support the potential of extended-spectrum triazoles (posaconazole and ravuconazole) and echinocandins for treatment of invasive aspergillosis in immunocompromised patients.^[6] The echinocandin caspofungin was approved last year for therapy for invasive aspergillosis that is refractory to standard therapy or for patients who cannot tolerate standard therapy.^[51]

Surgery is essential for certain manifestations of invasive aspergillosis, including endocarditis and endophthalmitis, and should be strongly considered in progressive sinusitis, skin and soft tissue infections, CNS lesions, and other focal processes amenable to a surgical procedure.^[52] While the exact role of surgery in the management of invasive pulmonary aspergillosis and the optimal time points for intervention have not been defined, surgery may prevent local extension and hematogenous dissemination and may be curative.^[41]

Several case series suggest that surgery can be safely and effectively performed in patients who have localized infections, even when neutropenia is present.^[41,53] In a series of 36 patients with hematologic malignancies and proven or probable pulmonary aspergillosis, surgery combined with medical treatment was successful in 15 of 16 patients. In 4 patients, the intervention was performed for diagnostic purposes, and in 12 patients, for therapeutic purposes. In 8 of the latter patients, surgery was an emergency procedure based on repeated chest CT scans that showed contact of lesions with larger pulmonary arteries; 6 of these patients were neutropenic. Surgery was uneventful in all cases. Serial CT scans were an important part of this approach, and altogether, 26 (72%) of 36 patients responded to therapy.^[41]

Despite continuing systemic antifungal therapy, patients with invasive pulmonary aspergillosis who recover from neutropenia but require further intensive chemotherapy or allogeneic HSCT have a risk of recurrence or exacerbation of up to 30%.^[54,55] While there appears to be no clear role for surgical resection of residual lesions, patients should have had at least a partial response and receive continuing antifungal therapy at dosages recommended for treatment of primary aspergillosis.^[54,55] A retrospective analysis suggests that the type of antifungal therapy, surgical resection of residual lesions, and the achievement of a complete response to antifungal therapy before transplantation may not be of predictive value.^[55]

Emerging Fungal Pathogens

A wide variety of previously uncommon opportunistic fungi are increasingly encountered as causing life-threatening infections in neutropenic patients.^[4,56] These emerging pathogens include, among others, yeast-like organisms (such as Trichosporon beigelii and Blastoschizomyces capitatus), hyaline filamentous fungi (such as Fusarium species, Paecilomyces species, Pseudallescheria boydii, and Scedosporium prolificans), dematiaceous molds (such as Bipolaris, Exophiala, and Alternaria species), and the Zygomycetes.^[825]

Whereas yeast-like organisms follow the pattern of fungemia and dissemination known from Candida species, the emerging filamentous fungi cause infections that are virtually indistinguishable from those of Aspergillus species. However, some of the hyaline molds, most notably Fusarium species, Acremonium species, and Paecilomyces species, can cause fungemia and can disseminate via the bloodstream and cause numerous embolic skin lesions. Infections caused by the emerging pathogens are associated with extraordinarily high case fatality rates; several of these organisms, including but not limited to T beigelii, Paecilomyces lilacinus, Fusarium species, P boydii, and S prolificans, demonstrate little susceptibility to amphotericin B, and they may require therapy with antifungal triazoles^[4,5,56] (Table 1 and Table 2).

Dimorphic (Endemic) Molds

Histoplasma capsulatum, Blastomyces dermatitidis, and Coccidioides immitis can cause invasive infections in immunocompromised cancer patients.^[4] The clinical presentation is usually more acute, and disseminated disease is common. Initial therapy relies on the use of amphotericin B; alternatives in less severe cases and options for consolidation or maintenance chemotherapy currently include itraconazole (for histoplasmosis and blastomycosis) and fluconazole (for coccidioidomycosis)^[5]

Novel Antifungal Agents

Further elucidation of the structure-activity relationship of antifungal triazoles has resulted in a new generation of synthetic compounds that include posaconazole (SCH 56592), ravuconazole (BMS-207147), and voriconazole. These agents have enhanced potency and broad-spectrum antifungal activity that includes most clinically relevant opportunistic yeasts and molds. While their pharmacokinetics vary, no fundamental differences in potency, spectrum, or antifungal efficacy have been noted thus far among the 3 agents.

Data from phase 2 and early phase 3 clinical trials indicate highly promising clinical efficacy against oropharyngeal candidiasis, esophageal candidiasis, and invasive aspergillosis.^[6] Indeed, as already mentioned, recently published results from a randomized phase 3 trial suggest that voriconazole may replace D-AmB as the standard agent for primary therapy for invasive aspergillosis.^[47] A number of case reports also suggest the potential usefulness of these novel triazoles for treatment of infection with unusual hyaline and dematiaceous filamentous fungi.^[825]

The echinocandins are a novel class of semisynthetic antifungal lipopeptides that act by inhibiting the synthesis of (1,3)-D-glucan, an important component of the cell wall of many pathogenic fungi. Three echinocandin compounds are currently in advanced stages of clinical development or have recently been approved by the FDA: anidulafungin (LY303366), caspofungin, and micafungin (FK463). The current data indicate that these agents possess similar pharmacologic properties. They have broad-spectrum fungicidal activity in vitro against Candida species and are also active against Aspergillus species. Phase 2 and 3 clinical trials with all 3 echinocandins, performed in patients with esophageal candidiasis, have demonstrated potent clinical efficacy in conjunction with an excellent safety profile.^[6]

Published data on more invasive infections are currently limited to micafungin^[57] and caspofungin. Based on a clinical phase 2 trial in 63 patients with invasive aspergillosis who could not tolerate standard agents or whose infections were refractory to these agents,^[51] caspofungin was recently approved by the FDA for use in this situation. Because of mild and transient elevations of hepatic transaminase levels in single-dose interaction studies, the concomitant use of caspofungin and cyclosporine (but not tacrolimus) is currently not recommended.

A multilamellar liposomal formulation of nystatin (Nyotran) has been developed with pharmacokinetics markedly different from all 4 amphotericin B formulations. Clinical phase 2 trials have been completed and have documented the clinical efficacy and safety of liposomal nystatin in the treatment of refractory invasive candidiasis and aspergillosis.^[6] However, it is unclear at present whether this compound will be developed further.

Reconstituting Host Defenses

Restoration or amelioration of host defenses is paramount to successful management of opportunistic fungal infections. This may include the discontinuation or dosage reduction of corticosteroids, if feasible, and the administration of recombinant cytokines and cytokine-elicited granulocyte transfusions in profoundly neutropenic patients.

Recombinant hematopoietic cytokines, such as granulocyte colony-stimulating factor (G-CSF) and granulocyte-macrophage colony-stimulating factor, shorten the duration of neutropenia and reduce the period of greatest risk for invasive fungal infections. While the full impact of this potentially preventive modality on the incidence of invasive fungal infections is unclear, a body of preclinical in vitro and in vivo data has accumulated that shows that recombinant cytokines, effector cells, and antifungal drugs can work synergistically to oppose fungal growth.

In addition, there is growing evidence that T_H1-dependent immunity plays an important role in successful host defenses against invasive candidiasis and invasive aspergillosis. Cytokines and anticytokines that promote this pathway (namely, interferon-, interleukin [IL]-12 and anti-IL-4) may be protective in vivo and act in cooperation with antifungal drugs.^[58,59]

The administration of G-CSF to healthy donors before leukapheresis, improvements in collection techniques, and cytokine exposure of harvested and irradiated granulocytes are able to increase the dose and function of transfused granulocytes and are currently being investigated as adjunctive therapy for refractory infections in patients with persistent neutropenia.^[60] Novel avenues to cellular immunotherapy and prevention in the stem cell transplant setting include the cotransplantation of granulocyte/monocyte progenitor cells and the adoptive transfer of immunocompetent T cells in the stem cell transplant setting, and perhaps, development of T-cell vaccines.

Future Directions

Invasive fungal infections can be expected to remain a frequent and important complication of anticancer therapy. Indeed, the still investigational exploitation of a graft-versus-tumor effect and the induction of immunologic tolerance in solid organ recipients by means of nonmyeloablative allogeneic HSCT^[61] indicate that the number of patients at risk is only too likely to expand. Identification of high-risk populations, improved diagnostic tools, an expanded and refined drug arsenal, further elucidation of resistance mechanisms and in vitro/in vivo correlation, incorporation of pharmacodynamics, combination therapies, and immunotherapies offer hope for further substantial progress. Well-designed and carefully conducted trials are needed more than ever to translate this progress into clinical practice.

Andreas H. Groll, MD, Wilhelms-University, Münster, Germany; Thomas J. Walsh, MD, National Cancer Institute, Bethesda, Md

When this article was written, Dr. Groll was senior clinical fellow, immunocompromised host section, pediatric oncology branch, National Cancer Institute, Bethesda, Md. He is currently senior staff physician, center for bone marrow transplantation, department of pediatric hematology/oncology, Wilhelms-University Medical Center, Münster, Germany. Dr. Walsh is head, immunocompromised host section, pediatric oncology branch, National Cancer Institute.