Introduction

Invasive fungal infections (IFI) are associated with a high rate of morbidity and mortality in at-risk patients [1]. Candida and Aspergillus species remain the predominant pathogens with a shift towards resistant fungal pathogens being noted [2]. Reports on resistance prevalence on fungi and bacteria vary, depending on regions, populations, and health-care facilities evaluated [3]. In Austria, a continuous monitoring of antimicrobial resistance is required by law; the idea is to increase the awareness of health-care associated infections, to provide data which support national and international comparison, and at least to provide reference data. Hence, within the last 10 years, we performed a nationwide passive surveillance on antifungal drug resistance and notice a low rate of drug resistance in molds when compared to other countries [4].

Methods

Clinical samples were obtained from the University Hospital Innsbruck and other medical centres in Tyrol, Austria. Fungal positive, primarily sterile specimens, and body fluids (n = 3903) were evaluated with broncho-alveolar lavages (n = 1602), blood cultures (n = 1580), and tissue biopsies (n = 312) being the most important ones.

Fungal specification was determined either by the conventional methods, matrix-assisted laser desorption ionization–time of flight mass spectrometry, or ribosomal internal transcribed spacer DNA sequencing. In total, 2670 yeasts and 1565 molds were detected from routine sampling and were evaluated for their antifungal susceptibility. The Etest® (bioMérieux, Vienna, Austria) and the reference methods released by the European Committee of Antimicrobial Susceptibility Testing for yeasts and molds (EUCAST-AFST) [5, 6] were applied to evaluate the in vitro susceptibility against amphotericin B, anidulafungin, voriconazole, posaconazole, and isavuconazole; MIC results were read following 24 h of incubation, if not, otherwise, indicated. Quality control was performed as recommended in EUCAST documents using C. krusei ATCC 6258 and C. parapsilosis ATCC 22019. EUCAST-AFST was applied only in cases of imprecise Etest® results and confirmation of resistance. EUCAST antifungal clinical breakpoints (CBPs) and epidemiological cutoffs (ECVs) have been used to monitor the emergence of resistance; these values segregate wild-type isolates from isolates that are likely to carry a resistance mechanism, designated non-wild type. Resistance trends were calculated for species for which in vitro resistance increased > 0.5% within the last 10 years.

Results

Tables 1, 2 summarize the results of Etest® and/or EUCAST broth microdilution method for echinocandins, azoles, and amphotericin B; Table 3 displays major fungal resistance trends. Amphotericin B was active against a broad range of yeasts and molds tested; resistance is not on rise. Overall, anidulafungin displayed good activity against Candida species; resistance increased from 0.5 to 3.6% in C. glabrata, from 0.5 to 1.7% in C. parapsilosis, and from 0.5 to 0.9% in C. albicans isolates tested. Resistance rates of C. glabrata against azoles were consistent over the years.

Table 1 In vitro susceptibility of the various antifungal agents against molds (n = 1233)
Full size table
Table 2 In vitro susceptibility of the various antifungal agents against yeasts species (n = 2670)
Full size table
Table 3 Resistance trends (species for which in vitro resistance increased for > 0.5% within the last 10 years) applying EUCAST breakpoints
Full size table

The absence of ECVs and CBPs for rare species such as Candida lusitaniae and C. guilliermondii does not support a categorization in wild types and non-wild types; however, an increase of isolates with high MICs over time was not recorded.

ECVs were also recently published from EUCAST for various Aspergillus species, and we applied these criteria for A. fumigatus species complex, Aspergillus niger species complex, Aspergillus terreus species complex, and for Aspergillus flavus species complex isolates. Only one azole-resistant A. fumigatus isolate was detected; azole resistance in A. terreus was noticed in a few isolates, which were collected between 2007 and 2009. ECVs and CBPs are lacking for non-Aspergillus molds, and hence, we display MIC data only, see Table 1.

Discussion

This survey on antifungal resistance observed a low frequency of azole resistance in A. fumigatus, a moderate increase of echinocandin resistance in yeasts, and a stable amphotericin B activity against yeasts and molds. Posaconazole resistance in A. terreus was newly detected in a few isolates; however, these isolates were collected between 2007 and 2009. Some countries report an alarmingly increase of antifungal resistance development within the past, with azole-resistant A. fumigatus displaying the major threat [4]. In Tyrol, fungal in vitro susceptibility data have not changed tremendously; conspicuous trends were not observed. Susceptibility tests were performed by applying the Etest® method to determine the MICs of the respective antifungal agents; in uncertain cases, the EUCAST methodology was applied [5, 6]. This commercially available test was previously described as a simple and reliable method for AST that was comparable in performance to reference procedures [7].

Most interesting from our study is the fact that we did not observe azole resistance in A. fumigatus as a major issue; only a few isolates were identified within the last 10 years. This is contrary to the overall worldwide trend. Azole antifungal drugs are the first line of therapy against A. fumigatus, a common etiologic agent of aspergillosis [8]. Resistance to azole drugs has been associated with treatment failure and deaths in patients with aspergillosis [4]. However, azole resistance in A. fumigatus has been documented in many regions in the past decade. Why neighbouring countries and others display such increase of azole-resistant A. fumigatus is unknown yet; it is hypothesized that the environment displays major sources due to agricultural usage of azoles [4].

Applying the new EUCAST breakpoints categorized a few isolates of A. terreus species complex to be posaconazole resistant with the clinical importance being unclear; a CBP of > 0.25 µg/ml for posaconazole was applied, and most of the patients involved were treated with voriconazole and improved [2]. At our institution, infections due to A. terreus were prevalent over the last 3 decades; however, the timely use of targeted treatment with voriconazole and the widespread use of posaconazole/micafungin prophylaxis decreased invasive aspergillosis from 12 to 6% in patients with haematological malignancies [9]; fungal breakthrough infections due to A. terreus are lacking [2], and hence, the role of resistance against posaconazole in A. terreus needs further clinical and molecular-based evaluation.

C. albicans is the leading cause of candidemia worldwide being responsible for one-third of the cases of candidemia in the US and European countries [10]. The echinocandin resistance rates among the Candida species were low overall, but resistance among C. glabrata is on rise [10]. Such development supports the importance of antifungal stewardship programs, as the echinocandins are the drug of choice to treat Candida-infections.

Remarkable is the efficacy of amphotericin B against this broad panel of fungi without either losing activity or introducing resistance over time. Amphotericin B MIC data are stable, resistance development is lacking. Hence, this drug has an exceptional position within the antifungals available in terms of resistance development.

For the mucormycetes and other molds, ECVs and CBPs are lacking; whether MICs > 2 mg/ml for amphotericin B, posaconazole, and isavuconazole represent clinical failure is unclear and needs further investigation. However, the proportion of fungi displaying MICs > 2 mg/ml for amphotericin B and posaconazole was stable over the last decade. By contrast, higher isavuconazole MICs against the Mucorales were not uncommon, see Table 1. This finding is of great interest, as isavuconazole was recently licensed as the second-line treatment of mucormycoses on the European market. Hence, more real life data are needed to clinically define azole resistance in Mucorales.

Limitations of this study include that data acquisition over years is a subject to change. Hence, different experts may read test results a little differently; protocols and the arbitrary nature of sampling may have changed minimally.

Despite the low antifungal resistance rates among Candida and Aspergillus isolates, continuous monitoring of antifungal susceptibility patterns and continuous determination of an understanding of mechanisms of resistance seem prudent. Reports of breakthrough infections, the increasing prevalence of uncommon species refractory to clinically available antifungal agents, and emerging resistance mechanisms highlight the importance of local and global surveillance.