Abstract

Introduction

The incidence of invasive fungal infections is rising [1-3], and although they are not yet as much of a threat as bacterial pathogens, their increase is significant enough for us to realize that a sinister shift in epidemiology has been occurring, especially in oncology patients and transplant recipients. In the United States from 1980 through 1997, the annual number of deaths in which an invasive mycosis was listed on the death certificate increased from 1,557 to 6,534—a 320% increase over 17 years [4]. Despite the expanded armamentarium of antifungal agents now available, invasive fungal infections continue to produce significant morbidity and mortality. This is because, with an increasing number of solid organ transplants being performed (over 23,000 organs were transplanted in the United States in 2008 alone—double the amount transplanted 10 years previously [5]), coupled with the use of newer and more potent chemotherapeutic agents and regimens, the pool of immunocompromised patients is increasing dramatically. These patient populations are extremely susceptible to invasive fungal infections because of their highly compromised immune status and the external pressures from antibiotic usage, which explains why the burden of invasive fungal infections is unfortunately rising commensurately with these medical advances. This article highlights recent invasive fungal infection epidemiologic pattern shifts and discusses the important factors to consider when approaching the unique patient populations affected by such infections.

The evolving epidemiology

Three epidemiological trends have emerged recently. First, in hematopoietic stem cell transplant (HSCT) recipients and in neutropenic patients with acute leukemia, an increased incidence of invasive mold infections—in particular, invasive aspergillosis—has occurred. This has necessitated changes in the first choice for antifungal therapy from less broad-spectrum agents covering Candida spp. predominantly to ones that are active against Aspergillus and other molds. Second, with the increasing incidence of non-albicans Candida spp., the utility of fluconazole as first-line initial therapy for invasive candidiasis has waned due to reduced susceptibility of the yeasts to this drug. Unfortunately, resistance to even the newer echinocandin class of antifungals is now also being observed. Finally, similar to the pattern of resistance arising because of antibiotic pressure, antimicrobial pressure exerted by the use of broad-spectrum antifungal agents such as voriconazole has precipitated the emergence of invasive mold infections caused by the Zygomycetes class of fungi. These trends are explored in more detail below.

The Transplant-Associated Infections Surveillance Network (TRANSNET), a consortium of 23 transplant centers in the United States, prospectively studied the epidemiology of invasive fungal infections in both solid-organ transplant and HSCT recipients over a 5-year period from 2001 to 2006 [6,7]. The most common invasive fungal infections in the solid-organ transplant recipients were candidiasis (53% of all invasive fungal infections found) followed by invasive aspergillosis (19%), cryptococcosis (8%), non-Aspergillus molds (8%), endemic fungi (5%), and zygomycosis (2%) [6]. In contrast, when looking at incidence rates in HSCT recipients only, invasive aspergillosis (43%) was more common than invasive candidiasis (28%) [7]. The rising prevalence of invasive aspergillosis and mold infections has also been observed in a single-center series of autopsy-proven patients with hematologic malignancies [8]. Comparing the prevalence from 1989 through to 2003, invasive aspergillosis and zygomycosis rose from 16% to 19% and from 1% to 3% respectively, while prevalence of invasive candidiasis fell from 13% to 8%. The shift toward more invasive mold infections is significant because the overall 1-year survival is lower in invasive aspergillosis compared to invasive candidiasis in both HSCT (25.4% versus 33.6%) [7] and solid-organ transplant patients (59% versus 66%) [6].

The second trend observed is the increasing incidence of non-albicans Candida spp.. Candida tropicalis, Candida parapsilosis, and Candida glabrata have been isolated much more frequently as causes of invasive candidiasis worldwide, although C. albicans remains the most common [9]. In addition, the SENTRY Antimicrobial Surveillance Program, which is an international surveillance of blood stream infections, also described an increasing resistance to the azole class of antifungals among nosocomial isolates of Candida spp. [10]. Fluconazole resistance in C. glabrata, C. parapsilosis, and C. tropicalis isolates was found in 7.7%, 5.8%, and 3.3%, respectively. For C. tropicalis in particular, 2.2% of the isolates were resistant to both fluconazole and voriconazole. Interestingly for C. parapsilosis, no echinocandin resistance was demonstrated. However, resistance to both azoles and echinocandins was prominent in C. glabrata, ranging from 3.2% resistance to micafungin to 7.7% with fluconazole. Resistance was even seen in the newer azoles, such as posaconazole (5.1%), and another, newer echinocandin, anidulafungin (3.8%). The most worrisome though, was the discovery of the propensity of C. glabrata to mutate to a multidrug-resistant phenotype in vivo in a single patient. The findings of resistance to fluconazole have been highlighted in the recommendations of the Infectious Diseases Society of America (IDSA), who recommend that infection with C. parapsilosis be treated with an azole (or with a lipid formulation of amphotericin B, alternatively) while infection with C. glabrata be treated with an echinocandin [11]. There is, however, some concern that the former recommendation may inadvertently result in increased azole resistance [10]. Indeed, it now appears crucial that microbiology laboratories perform antifungal susceptibility testing more readily for infections involving the non-albican Candida spp. to better guide the clinician in the optimal choice of antifungal agent.

The third trend relates to the emergence of zygomycosis and warrants further discussion. Following the landmark randomized, unblinded trial where voriconazole showed a superior 12-week survival rate compared with amphotericin B when used as primary therapy for invasive aspergillosis [12], voriconazole is now the drug of choice for the treatment of invasive aspergillosis as recommended by the IDSA [13]. It is of note though, that with the increasing use of voriconazole—as prophylaxis, empirical, preemptive, and targeted treatment for invasive aspergillosis in patients with hematologic malignancy—concerns have arisen about an increase in incidence of zygomycosis, an invasive mold infection resistant to voriconazole. In a series of 19 immunocompromised patients (including solid-organ transplant recipients, patients with hematologic malignancy or HIV-AIDS, and those using steroids) diagnosed with invasive zygomycosis, the increase in the number of cases was associated with the introduction of voriconazole and caspofungin and an increase in the number of HSCTs performed [14]. In a case-control study involving 14 leukemic and 13 HSCT patients, voriconazole prophylaxis was identified as an independent risk factor, increasing the risk for invasive zygomycosis by more than 10-fold [15]. Singh and colleagues performed a prospective, matched case-control study of 50 solid-organ transplant recipients and found that voriconazole and/or caspofungin use increased the risk of zygomycosis by 4.4 times [16].

The risk groups and risk factors for invasive fungal infection

Patients with acute leukemia, HSCT recipients, and solid-organ transplant recipients represent the three most common cohorts of patients at risk of invasive fungal infections. In approaching these immunocompromised patients, it is important to recognize that the probable etiologic pathogens may be predicted based on an awareness of the underlying immunologic defect (e.g., neutrophil, T cell- or B cell-mediated) and the duration and severity of the defect(s). Extrinsic factors (radiation, drugs, or surgery, resulting in breaches in the mucocutaneous defensive surfaces of the body), intensity of immunosuppression (dose, duration, and temporal sequence of immunosuppressive regimen) and environmental exposures (community or nosocomial) may also play significant roles and influence which pathogen may produce an invasive fungal infection. The summary risk of infection depends on the overall net state of immunosuppression and the epidemiologic exposures encountered [17].

In patients with acute leukemia, the risk of invasive fungal infection is related to the status and type of leukemia (i.e., newly diagnosed de novo acute leukemia, post-remission status, relapsed leukemia, or acute leukemia refractory to treatment, as well as acute myelogenous leukemia versus acute lymphocytic or hairy cell leukemia), depth and duration of neutropenia, and the types of chemotherapeutic agents used to treat the leukemia. Neutrophils are rapidly destructive to fungal hyphae in particular, through toxic oxygen radicals. Protracted neutropenia coupled with breaches in skin (e.g., from a central intravascular line) and bowel (e.g., from mucotoxic chemotherapy) predispose patients to candidemia and invasive aspergillosis.

For the HSCT recipients, the influence of the host, the graft (with regard to the duration of neutropenia prior to engraftment), and complications of the procedures performed (e.g., use of central venous intravascular catheters) fluctuate throughout the post-transplant course, creating a dynamic timeline. Host factors (e.g., older age) and transplant variables (e.g., human leukocyte antigen mismatch) tend to influence invasive fungal infection risk early on. In particular, the use of immunosuppressives during the pre-engraftment phase to prevent graft-versus-host disease may result in severe neutrophil depletion and hence the propensity to develop invasive candidiasis and invasive aspergillosis. However, post-engraftment complications of the transplantation (e.g., chronic graft-versus-host disease and Cytomegalovirus disease, which produce profound mononuclear cell dysfunction) lead to alveolar macrophage phagocytic defects and the propensity to develop late-onset invasive aspergillosis [18] and other fungal infections such as zygomycosis. Some biologic factors like malnutrition, iron overload, diabetes mellitus, and cytopenias moderate the risk somewhat throughout the post-transplant course. One-year cumulative incidences for invasive fungal infection were highest for mismatched-related (8.1%) and matched-unrelated allogeneic (7.7%) HSCTs, followed by matched-related HSCT (5.8%) and autologous HSCT with the lowest incidence rate (1.2%) [7].

In solid-organ transplant recipients, based on the types of organ transplanted, 1-year cumulative incidences for invasive fungal infection were highest for small bowel (11.6%) and lowest for kidney transplant recipients (1.3%) with a trend toward a slight increase in cumulative incidence from 2002 to 2005 [6]. Rejection and exogenous immunosuppressive agents, especially high-dose steroids and antilymphocyte globulin use, escalate the risk for acquiring invasive fungal infections. Medical and surgical factors both play crucial roles [19]. For example, prolonged ischemia time has been associated with the development of invasive aspergillosis in lung transplant recipients [20]. For liver transplant recipients, fulminant hepatic failure, retransplantation, and renal failure requiring dialysis are risk factors for invasive fungal infection [21]. Diabetes mellitus and need for prolonged hemodialysis before transplant are unique risks for renal transplant recipients [22]. In pancreas transplant recipients, older donor age, enteric (versus bladder) drainage, pancreas transplant after kidney transplant (versus pancreas only), post-transplant pancreatitis, retransplantation, and preoperative peritoneal dialysis have been identified as risk factors for invasive fungal infection (candidiasis) [23].

Timing of invasive fungal infections

The time to development of invasive fungal infection after transplantation depends on the transplant type, the use and duration of use of antifungal prophylaxis, and the fungal pathogen. In the HSCT population, the timeline is split into three time periods from the time of transplantation: early onset (≤1 month [pre-engraftment phase]), late onset (1–6 months [post-engraftment phase]) and very late onset (>6 months) (Figure 1). In the TRANSNET report, the median time after transplantation to onset of invasive fungal infection was 61, 99, 123, and 135 days for candidiasis, aspergillosis, fusariosis, and zygomycosis cases, respectively [7]. Another multicenter report of invasive fungal infections occurring between 2004 and 2007 from the Prospective Antifungal Therapy (PATH) Alliance registry found that invasive candidiasis tended to occur earlier after autologous HSCT (median 28 days) compared with allogeneic HSCT (median 108 days) [24]. The earlier onset of invasive candidiasis is usually related to neutropenia and mucositis whereas the later onset is more often seen in allogeneic HSCT recipients due to the development of graft-versus-host disease and the need for chronic central venous catheter usage. Invasive aspergillosis occurs more frequently but is encountered later in allogeneic HSCT recipients compared with autologous HSCT, and this late occurrence confers a higher mortality [7]. A study that examined the records of 5,589 patients who underwent HSCT from 1985 through 1999 found that the majority of scedosporiosis cases tended to occur within the first 40 days after transplant and, unfortunately, all those with scedosporiosis died within 1 month of infection [25].

Open in a separate window

Figure 1.

Timing of invasive fungal infections after organ transplantation

Solid-organ transplantation (SOT) figures are based on the US Transplant-Associated Infection Surveillance Network (TRANSNET) data from 2001 to 2006 [6]. Hematopoietic stem cell transplantation (HSCT) figures are based on US TRANSNET data from 2001 to 2006 [7].

In the solid-organ transplant population, the timeline has traditionally been divided into three time periods not dissimilar to the HSCT population [17]. However, recent data has demonstrated that invasive fungal infections are occurring later [6]. The median time to onset is 103 days for invasive candidiasis, 184 days for invasive aspergillosis, 312 days for zygomycosis, 343 days for endemic fungal infections, 467 days for non-Aspergillus molds, and 575 days for cryptococcosis. The majority of infections occur more than 90 days after transplantation. The occurrence of early infections (≤90 days) is dominated by invasive candidiasis and invasive aspergillosis. Significant numbers occur beyond 3 years after transplantation.

Diagnostics

In a case series involving patients with hematologic malignancies, although there was a high prevalence of invasive fungal infection (31%) at autopsy, with 77% of the patients’ deaths related to infection, only one-third of them were diagnosed by the accepted criteria before the patients died [8]. This highlighted the inadequacies of current diagnostic methods for invasive fungal infections.

Although traditional microbiological culture-based testing for fungi produces low yields and creates time delays, it is inexpensive and provides valuable material on which to perform drug-sensitivity testing, which has become increasingly relevant due to the changing patterns of resistance. Another factor to consider is that histological evaluation of tissue specimens for fungi may be too invasive for most patients and may not be specific enough for some fungal species.

New methods to detect invasive fungal infections early and with high sensitivity and specificity are needed. Galactomannan, a component of the cell wall of Aspergillus spp., is released during early stages of growth thereby allowing the tests for its presence to be fairly sensitive (71%) and specific (89%) [26]. It can be detected in serum and also bronchoalveolar lavage fluid thus providing a surrogate marker for the diagnosis of invasive aspergillosis. The recent development of Aspergillus polymerase chain reaction for the rapid diagnosis of infections holds promise for increasing the sensitivity and specificity further, but standardization issues of the assays remain unresolved [27].

Testing of beta-glucan (a major cell wall constituent characteristic of most fungi), although approved in some jurisdictions for detecting invasive candidiasis and other fungal infections (except zygomycosis), has not received widespread acceptance due to difficulties in performance techniques and lack of specificity [28]. Nevertheless, with the increasing reliance on serum or bronchoalveolar lavage galactomannan testing and serum beta-glucan assays, centers are now performing fewer biopsies and obtaining fewer histopathology specimens [8], commonly employed to confirm a proven case of invasive fungal infection according to the European Organization for Research and Treatment of Cancer/Mycoses Study Group (EORTC/MSG) criteria [29].

Computed tomography (CT) scans of the chest are widely used as diagnostic tools when dealing with pulmonary or hepatic invasive fungal infections in immunocompromised patients. Fungal infections such as invasive aspergillosis present with nodules surrounded by haziness (halo sign) and may be associated with cavitatory lesions. Identifying these lesions early via CT scans has helped to cut the time it takes to diagnose invasive mold infections and allow earlier initiation of appropriate antifungal therapy. Similarly, defects in the liver architecture noted on CT scanning of the abdomen provide an early sign of an invasive fungal infection involving the liver. The current EORTC/MSG criteria for diagnosis of invasive fungal infections involve a combination of host factors, histopathology, imaging, and microbiological cultures or markers to clinch diagnosis [29]. These criteria enabled some standardization of the diagnosis of invasive fungal infections.

Management

The management of invasive fungal infections begins with a high index of suspicion and prompt diagnosis. Clearly the early initiation of antifungal therapy is a major contributor to enhanced survival. Reducing immunosuppression whenever possible with prompt initiation of appropriate antifungal therapy (monotherapy or combination) (Table 1) remains the mainstay for the treatment of invasive fungal infections. However, rapid reduction of immunosuppressive therapy in conjunction with initiation of antifungal therapy in solid-organ transplant recipients may lead to the development of immune reconstitution inflammatory syndrome, the clinical manifestations of which mimic worsening disease [36]. Reversal of neutropenia with granulocyte colony-stimulating factor or granulocyte infusions may be used to hasten recovery from neutropenia.

Surgery may be required depending on the type of invasive fungal infection. Recommendations for surgery may be considered in patients with invasive aspergillosis who have a solitary lung lesion before chemotherapy or HSCT, those with hemoptysis from a lung lesion, those with disease that invades the chest wall, or situations where the infection involves the pericardium or great vessels [13]. For zygomycosis, surgery is often required in addition to antifungal therapy.

Conclusion

The epidemiologic changes in the occurrence of invasive fungal infections in a rapidly expanding population of immunocompromised patients present a real challenge to physicians managing them. The persistently high morbidity and mortality associated with these infections continues despite the increasing spectrum of antifungal pharmacotherapy. Perhaps the best approach to counter the ubiquitous fungi that cause invasive fungal infections may be to develop less damaging methods of immune suppression. If we can shorten the duration of neutropenia and improve our immune suppression regimens to better achieve immune tolerance of the allograft, we might gain an upper hand in our fight against these sophisticated eukaryotes.

Abbreviations

Notes

The electronic version of this article is the complete one and can be found at: http://f1000.com/reports/m/3/14

Notes

References