The changing face of epidemiology of invasive fungal disease in Europe

Authors

  • Cornelia Lass-Flörl

    1. Department of Hygiene, Microbiology and Social Medicine, Division of Hygiene and Medical Microbiology, Innsbruck Medical University, Innsbruck, Austria
    Search for more papers by this author

Cornelia Lass-Flörl, Fritz Pregl Str 3/III, Department of Hygiene, Microbiology and Social Medicine, Division of Hygiene and Medical Microbiology, Innsbruck Medical University, Innsbruck, Austria.
Tel.: +43 512 9003 70725. Fax: +43 512 9003 73700.
E-mail: cornelia.lass-floerl@i-med.ac.at

Summary

Invasive fungal diseases (IFDs) are an increasingly common complication in critically ill patients in Europe and are frequently fatal. Because of changes in treatment strategies and the increased use of antifungal prophylaxis, the epidemiology of IFDs has changed substantially in recent years and infections due to Candida species are no longer the majority in many institutions. In contrast, the emergence of non-Candida IFDs such as aspergillosis, zygomycosis and fusariosis has increased. European surveys indicate that Candida albicans is responsible for more than half the cases of invasive candidaemia; however, the occurrence of non-albicans-related IFDs appears to be increasing. Rates of IFD-related mortality in Europe depend on the pathogen, geographical location and underlying patient characteristics, with rates ranging from 28 to 59% for Candida infections and from 38 to 80% for invasive aspergillosis. Early initiation of antifungal therapy is critical for improving outcomes; however, this is complicated by the difficulty in diagnosing IFDs rapidly and accurately. The introduction of new extended-spectrum azole antifungal agents (e.g. voriconazole, posaconazole) and echinocandins (e.g. micafungin, caspofungin, anidulafungin) has increased the number of therapeutic options for early therapy. Choice between agents should be based on a variety of factors, including spectrum of activity, adverse events, drug interactions, route of administration, clinical efficacy of individual agents and local epidemiology.

Introduction

Invasive fungal diseases (IFDs) are an increasingly encountered threat among critically ill patients and are a significant cause of morbidity1–4 and mortality. The incidence and severity of IFD are dependent on a variety of factors including increased use of immunosuppressive agents, antineoplastic agents, broad-spectrum antibiotics, prosthetic devices and grafts and hyperalimentation.1 In addition, improvements in medical care have resulted in critically ill patients surviving longer, rendering them vulnerable to IFD. Populations at risk for IFD include haematopoietic stem cell transplant (HSCT) and solid organ transplant (SOT) recipients; patients with haematological malignancy; patients with HIV/AIDS; and intensive care unit (ICU), surgical and burn patients.1,4–7

Although a wide variety of pathogens can be associated with IFD, Candida species have historically been the most common causative organisms. However, the epidemiology of IFD in Europe has shifted in recent years as Aspergillus species and other moulds have become increasingly important pathogens. This article will review the epidemiology of IFD in Europe, risk factors associated with specific fungi and associated mortality rates.

Epidemiology of invasive fungal disease in Europe

A majority of large studies evaluating IFD epidemiology have been conducted in academic centres in North America.8 In contrast, the European data presented in this review are mainly derived from single-institution reports or multiple sites within countries rather than from multi-national reports.

Until the 1990s, candidiasis was the most common IFD in Europe among immunocompromised patients. Since then, there has been a notable shift in the epidemiology, as evidenced by an increased incidence of invasive aspergillosis (IA).9,10 For example, an autopsy study at a German hospital found that, although the 11 IFDs identified at the hospital between 1973 and 1980 were caused by Candida species, Aspergillus species were responsible for 62% of IFDs (16/26) diagnosed at the hospital between 1992 and 2001 and Candida spp. were identified in 35% of the cases.9 Other reports have also found that Aspergillus now exceeds Candida as the major cause of IFD in some European populations. Data from the SEIFEM (Sorveglianza Epidemiologica Infezioni Fungine nelle Emopatie Maligne) 2004 study indicate that 64% of IFDs in patients with haematological malignancy who were admitted to haematology wards in Italy between 1999 and 2003 were caused by mould infections.11 More than 90% of mould infections were caused by Aspergillus species, with the Zygomycetes (4%) and Fusarium species (4%) the next most frequently implicated moulds.11 In contrast, yeasts were the cause of 36% of IFDs and Candida species were responsible for most (91%) yeast infections.11 The same study reported similar results among patients who underwent HSCT: 91 (75%) of 121 IFDs were caused by moulds and 30 (25%) of 121 were caused by yeasts.12 Another trend in the epidemiology is that infections caused by non-albicans Candida species (e.g. C. krusei, Candida glabrata, Candida tropicalis, Candida parapsilosis) have been increasingly seen, especially among patients with haematological malignancy.3,8,13–15 Other, previously uncommon fungi, such as the Zygomycetes and Fusarium species, are also emerging as aetiological pathogens; frequency varies by geographical location.

The reasons for the changing epidemiology of IFD are not entirely known, but probably result from a number of factors, such as the increased use of fluconazole prophylaxis in HSCT recipients and changes in treatment strategies for various at-risk populations. Changes in treatment strategies include the use of T-cell–depleting agents (e.g. alemtuzumab) and increased use of allogeneic stem cells in transplant recipients.10,16

Candida

Candidiasis can manifest as a wide range of clinical symptoms, from mucocutaneous overgrowth to blood stream infections (candidaemia) and metastatic infections.17,18Candida is a common inhabitant of the normal human flora (e.g. skin, gastrointestinal tract, genitourinary tract) and is also found in the environment (especially on surfaces).18 Although the source of most cases of Candida infection (particularly C. albicans) is endogenous, infections can also originate from exogenous sources.13,19

Epidemiological studies from Europe and the United States showed that incidence rates of candidaemia increased from the 1970s through the early 1990s. Some data suggest that incidence rates then stabilised during the 1990s, although a recent comprehensive Danish study on fungaemia reported that incidence of candidaemia was increasing.20 A French study of over 49 000 ICU patients also reported a gradual increase in systemic candidiasis from 1.2 to 3.1% and 4.6% in 1995, 1996 and 1997 respectively,21 whereas an Austrian study reported a significant increase in the number of cases of candidaemia between 2001 and 2006.22 In the European Confederation of Medical Mycology (ECMM) survey conducted from September 1997 to December 1999, rates of 0.20 to 0.38 per 1000 admissions and 0.31 to 0.44 per 10 000 patient-days were reported by participating countries.8 These rates are generally lower than those reported in the United States.3,8 This is probably due to numerous factors including differences in patient demographics and comorbidities, as well as differences in medical practices, such as the use of long-term vascular catheters and antibacterial and antifungal usage patterns.3Table 1 summarises rates of candidaemia in various European countries according to hospital type and patient population. The incidence of candidaemia per 1000 admissions in European countries ranged from 0.17 (nationwide in Finland and in general hospitals in France) to 20 (ICU patients in Germany). The rate of fungaemia was less than 5 per 100 000 population in all European countries except Denmark, which had a rate of 11 per 100 000.20 The reason for the remarkably high rate of fungaemia in Denmark is not clear.

Table 1.   Incidence of candidaemia in Europe
Country/period of observationNo. of hospitals/type of hospitals (patient population)Rate/1000 admissionsRate/10 000 patient-daysRate/100 000 population
  1. ICU, intensive care unit.

  2. aPer 1000 patients staying more than 10 days in an ICU.

  3. bEuropean Confederation of Medical Mycology (ECMM) survey.15

Belgium
1992–200059ICU2.5  
Denmark
2003–20046034/overall hospitals0.49 11
2004–20062041/overall hospitals0.5 11
Finland
1995–199961Nationwide0.170.41.9
France
199562General hospitals0.170.17 
199562University hospitals0.380.52 
199562Cancer centres0.711.2 
19956225/overall hospitals0.290.35 
1995–19972132/ ICUs2  
1997–1999b15250.2  
1998–2001635/overall hospitals0.28  
Germany/Austria
1989–1990648/ICUs20a  
1997–1999 b155 0.31 
2001–200622University hospital0.27–0.77  
Holland/Netherlands
1987–1995655/overall hospitals 0.31 
1989–199366University hospital 5.5 
Hungary
1996–200067University hospital0.2–0.4  
Iceland
1980–199968Nationwide  1.4–4.9
Italy
1986–199769Cancer centre15.5  
1992–199770University hospitals 1.14 
1997–1999b15,7135/general hospitals0.380.44 
Norway
1991–199672Nationwide0.170.262.4
Portugal
200473University hospital2.0–2.7  
Spain
1991–19927428/ICU2.0  
1995–199975Teaching hospital0.81  
1997–19987627 (neonates)5.7  
1997–1999b15,7719  3.5
2002–20032614/overall hospitals0.530.734.3
Sweden
1997–1999b15,7860.32  
Switzerland
1989–200079University hospital 0.31 
1991–2000806/university and university-affiliated hospitals 0.50 
20008118/overall hospitals 0.49 
United Kingdom
1997–1999b156 0.30 
200381171/overall hospitals 2.53

Patient populations at risk for invasive candidiasis include patients in the ICU,23–25 patients with solid tumour or haematological malignancy and those undergoing surgery.8 Mucous membrane colonisation appears to be a prerequisite for subsequent development of IFD due to Candida species and the occurrence of colonisation appears to be significantly increased among critically ill patients receiving broad-spectrum antibiotics.8,24 Other risk factors for invasive candidiasis include the use of intravascular devices, catheters and parenteral nutrition; neutropenia; surgical procedures; renal failure; use of steroids and H2 blockers and longer duration of ICU stay.2,8,18,26

Data from the ECMM and other surveys indicate that C. albicans is responsible for more than half the cases of candidaemia.8,15,18,26 However, non-albicans Candida species such as Candida krusei, Candida glabrata, Candida tropicalis and Candida parapsilosis are emerging as causative agents.3,8,14,15,18 In the ECMM survey, incidence rates for non-albicans candidaemia infections were 14% each for C. glabrata and C. parapsilosis, 7% for C. tropicalis and 2% for C. krusei).8 A German study reported incidence rates of 19% for C. glabrata and 8% each for C. parapsilosis and C. tropicalis,27 whereas other surveys have reported incidence rates as high as 30% for C. parapsilosis and 11% for C. krusei.8 The ECMM found that the infective organism varies according to patient population (Fig. 1).8 In particular, non-albicans infections were more common in patients with haematological malignancy.8 A prospective, multicentre surveillance study by the Invasive Fungal Infection Group (IFIG) of the European Organization for Research and Treatment of Cancer (EORTC) reported that C. albicans was responsible for 70% of cases in patients with solid tumours and for only 36% among patients with haematological diseases, whereas C. glabrata, C. tropicalis, C. parapsilosis, C. krusei and other non-albicans Candida species were each responsible for 12 to 14% of the cases among patients with haematological malignancies.28 The ECMM also found an age-related effect on epidemiology, with a reduction in the incidence of C. parapsilosis and an increase in the incidence of C. glabrata with age.8

Figure 1.

 Frequency of Candida species from the European Confederation of Medical Mycology (ECMM) according to underling patient population.8 Reprinted from Tortorano AM, Kibbler C, Peman J, Bernhardt H, Klingspor L, Grillot, R. (2006) Candidaemia in Europe: epidemiology and resistance. Int J Antimicrob Agents 27, 359–366, with permission from Elsevier.

European surveys have found that there is little antifungal resistance to Candida species, with the exception of C. krusei, which has inherent resistance.8,29 Nevertheless, the use of prophylactic fluconazole appears to be a risk factor for at least some non-albicans species (e.g. C. krusei, C. glabrata).3,14,18 Studies from the YEASTS group in France have reported reduced posaconazole activity in C. glabrata isolates with reduced susceptibility to voriconazole and fluconazole,30 as well as some cross-resistance between the echinocandins caspofungin and micafungin for C. parapsilosis and C. guilliermondii.31 A semi-national fungaemia surveillance programme in Denmark reported a statistically significant increase (= 0.0001) from 2004 to 2006 (20% for 2004, 28% for 2005, 35% for 2006) for the percentage of fungal isolates showing decreased susceptibility to fluconazole and/or itraconazole (defined as MIC ≥4 mg l−1 and ≥0.25 mg l−1 respectively).20

Reported mortality rates from Candida fungaemia range from 28 to 59% in European surveys and depend on species and geographical location.8,26 Of 249 cancer patients studied by the IFIG of the EORTC, 39% died within 30 days of the onset of candidaemia, with the highest mortality rate associated with C. glabrata infections.28 In a French study of ICU patients, those with candidiasis were at significantly greater risk of death (= 0.001) with a mortality rate of 60% (compared to an overall mortality rate of 16%).21 In a Spanish study, treatment with antifungals and by catheter removal were protective factors against early death, whereas the presence of haematological malignancy was associated with increased risk.32 Other risk factors include older age, severity of the underlying disease, persistent infection, allogeneic bone marrow transplantation, septic shock and lack of antifungal prophylaxis.18

Aspergillus

The increasing incidence of IA varies according to location and patient population (Table 2); rates of up to 7% are reported in Europe.12,33Aspergillus fumigatus is a leading cause of aspergillosis;11,34 however, non-fumigatus Aspergillus species such as Aspergillus flavus and Aspergillus terreus are becoming more common.35,36

Table 2.   Incidence of aspergillosis in Europe
Country/period of observationNo. of hospitals /type of hospitalPatient groupNo. of admissionsIncidence, %Rate /10 000 patient days
  1. HSCT, haematopoietic stem cell transplant; ICU, intensive care unit; AIDS, acquired immune deficiency syndrome.

Austria
2000–2003421/university hospitalAllogeneic HSCT10423 
Autologous HSCT628 
Belgium
2000–2003331/university hospitalCritically ill18506.9 
199951/university hospitalICU2222.7 
France
2000–2002341/teaching hospitalOverall  0.44
Haematology  5.8
Italy
1999–20031211/tertiary care or university hospitalsAllogeneic HSCT12496.3 
Autologous HSCT19790.4 
1999–20031118/tertiary care or university hospitalsHaematological malignancy118022.6 
1986–1997824/infectious disease departmentsAIDS26142.1 
Spain
1996–20003915Allogeneic HSCT3959.4 
Finland
1990–2001836/transplant centresAutologous HSCT11880.8 
Poland
1990–2006841/paediatric hospitalPaediatric liver transplant recipients2773.6 

Patient populations at risk for IA include those with haematological malignancy, pulmonary disease [e.g. chronic obstructive pulmonary disease (COPD) and asthma], SOT recipients and patients with solid tumours.34,37 In patients undergoing HSCT, a bimodal occurrence of IA has been observed after HSCT, with late-onset IA occurring most frequently.38,39 Specific risk factors for IA include neutropenia (polymorphonuclear neutrophil count <0.5 g l−1), antibiotic therapy, corticosteroid therapy, cytotoxic chemotherapy, other immunosuppressants, Aspergillus colonisation, cytomegalovirus (CMV) or Pneumocystis jirovecii infection, stem cell source, graft-versus-host disease (GvHD), human leucocyte antigen mismatch in HSCT and use of monoclonal antibodies and nucleoside analogues.16,37,40–42 Disease presentation may differ between neutropenic and non-neutropenic patients; non-neutropenic IA is characterised by fewer symptoms (e.g. fever, cough, chest pain), more cases of pneumonia and a higher fatality rate than neutropenic IA.37

Mortality rates for IA are high in Europe and vary according to patient population, ranging from 38% in patients with acute myelogenous leukaemia (AML) to 94% in non-neutropenic patients.11,12,34,35,37 The epidemiology of IA is poorly documented in COPD patients; however, very high mortality rates (approaching 100%) have been reported.43,44 Disseminated disease, coinfection with P. jirovecii and/or CMV and intercurrent bacterial pneumonia have been identified as factors associated with a poor prognosis.37 However, there is some evidence for improved prognosis for acute leukaemia patients with IA; in an Italian study, there was a significant reduction in the attributable mortality rate of IA [from 60% (12 of 20 cases ending in death) during 1987–1988 to 32% (24 of 76 cases ending in death) during 2002–2003].45

Zygomycetes

Zygomycosis has emerged as an increasingly important IFD that is characterised by a high mortality rate.46 Pathogens from the order Mucorales (e.g. Rhizopus, Mucor, Rhizomucor) are most frequently implicated in human disease, although organisms from the order Entomophthorales can also be involved.10,46 Infections due to Mucorales (i.e. mucormycoses) are more aggressive, producing acute onset, rapidly progressive and commonly fatal angioinvasive disease.46 Entomophthorales typically produce indolent and chronically progressive disease, although they can produce clinical syndromes that are indistinguishable from those produced by Mucorales.46 The most common sites of involvement are the sinuses, lungs, skin, central nervous system and gastrointestinal tract.47

Neutropenic and immunosuppressed patients, such as those with haematological malignancy, those with HIV infection, those undergoing HSCT or those receiving corticosteroids, are at increased risk for zygomycosis. Other risk factors include metabolic conditions (e.g. diabetic ketoacidosis, uncontrolled diabetes, chronic metabolic acidosis), deferoxamine therapy, iron overload, skin or soft tissue breakdown (e.g. burn wounds), traumatic inoculation, surgical wounds, drug use, prematurity, malnourishment, sinusitis and prophylaxis with voriconazole.46–48 Among patients with haematological malignancy in the SEIFEM-2004 survey, the zygomycosis-related mortality rate was 64%.11 Risk factors for death include disseminated disease, renal failure and infection with Cunninghamella species; type 1 diabetes and absence of an underlying condition were associated with a reduced risk for death.47

Other fungi

Fusarium species are emerging pathogens and their incidence is expected to increase among neutropenic patients with haematological disease.49 The most common manifestations of fusariosis in immunocompromised patients include skin lesions, sinopulmonary infection, brain abscess, peritonitis and disseminated disease.50 In the SEIFEM-2004 survey, Fusarium species were responsible for 0.1% of infections; the majority was in patients with AML.11 In a survey of 14 Italian haematology departments, Fusarium infection was documented in 1.7% of patients with microbiologically documented filamentous fungi infection.49 In this study, the overall incidence of fusariosis was 0.06 (0.08 in patients with AML)49 among adult patients with leukaemia over a 10-year period. Risk factors for Fusarium infection include HSCT or SOT, haematological cancer (including acute leukaemia), burns, use of corticosteroids, neutropenia, severe immunosuppression, lymphopenia and GvHD.51 Mortality rates are high, ranging from 50% to almost 90% in immunocompromised patients.11,36,51–53 Risk factors for death include corticosteroid therapy and persistent neutropenia (<500 neutrophils mm−3).51

Scedosporium species are environmental pathogens that cause local and disseminated infection in immunocompetent and immunocompromised patients both.54 However, disseminated infection is found predominantly among immunocompromised patients, such as those with leukaemia or HSCT or SOT recipients and after near-drowning episodes.51,54,55Scedosporium is ubiquitous, with a worldwide distribution. Two species are medically important: Scedosporium apiospermum and Scedosporium prolificans.54 Cases of disseminated Scedosporium infection have been reported in Europe.55,56 Because Scedosporium is intrinsically resistant to many antifungal agents, most cases have a poor outcome.51,55,56

A number of endemic IFDs have heretofore been restricted to certain geographical locations. These include histoplasmosis, blastomycosis, paracoccidioidomycosis, coccidioidomycosis and penicilliosis. The possibility should be considered that travellers returning to Europe may have contracted infections from these pathogens.57

Local epidemiology and choice of antifungal treatment

Because delays in treatment are associated with increased mortality rate, earlier initiation of treatment strategies and prophylaxis may be appropriate.57 When initiating antifungal therapy, clinicians must weigh the risks associated with early intervention against the high morbidity and mortality rates associated with IFD in immunocompromised patients. A number of antifungal agents may be used for prevention, treatment of IFD or both, including several new extended-spectrum agents approved in recent years. Choice of a specific agent should be based on a number of factors, including antifungal activity, adverse events, drug interactions, route of administration, clinical efficacy of individual agents and local epidemiology.

Conclusions

A shift has occurred in the epidemiology of IFDs in Europe. Where invasive candidiasis was once the predominant type of IFD, invasive mould infections have become increasingly important, including those caused by unusual pathogens. Moulds have become the leading cause of IFD in some populations. Aspergillus species are the most frequent mould pathogens, but the number of infections caused by previously rare pathogens, such as the Zygomycetes and Fusarium species, is increasing. Various patient populations are at risk for IFD, particularly those who undergo allogeneic or autologous HSCT and treatment for haematological malignancy. The reasons for the shift in the epidemiology of IFDs are multifactorial, but are a result, at least in part, of the increased use of fluconazole as prophylaxis. These pathogens are being monitored by a European group, Fungiscope, which is collecting data related to cases of rare filamentous fungi.58

The management of IFD is complicated by the difficulty in diagnosis, which can result in the delay of appropriate treatment and in subsequent adverse outcomes. Because of the difficulty in diagnosis and high mortality rate associated with IFD, earlier treatment options and prophylaxis with extended-spectrum agents should be considered, along with appropriate adjunct surgery, to improve patient outcomes in appropriate high-risk patient groups.

Ancillary