Potency of Anidulafungin Compared to Nine Other Antifungal Agents Tested against Candida spp., Cryptococcus spp., and Aspergillus spp.: Results from the Global SENTRY Antimicrobial Surveillance Program (2008)

Opportunistic fungal infections are increasing in incidence (18) and are associated with high rates of morbidity and mortality (1, 11, 13). The rise in prevalence of individuals with short-term neutropenia (cancer patients undergoing chemotherapy regimens), long-term immunosuppression (organ transplant patients), immune system disorders (patients with HIV/AIDS), or central venous catheters has coincided with the increased occurrence of problematic opportunistic fungal infections (11). At this time, only a limited number of azole and echinocandin antifungal agents are available for therapeutic intervention against these infections.

Anidulafungin (9, 14-17) is a novel semisynthetic agent that targets cell wall structural integrity via noncompetitive inhibition of β-1,3-d-glucan synthesis, resulting in cell rupture and death. Excellent broad-spectrum in vitro and in vivo activities against a variety of fungal pathogens have been demonstrated (16). We present here contemporary data (2008) from the global SENTRY Antimicrobial Surveillance Program comparing the activity of anidulafungin to those of nine additional antifungal agents by use of reference methods (5-7).

A collection of 1,201 clinical yeasts from bloodstream infections (BSI) and 79 molds from pneumonias (lower respiratory tract infections [LRTI]) in the United States, Europe, Latin America, and the Asia-Pacific region (APAC) was processed by Clinical and Laboratory Standards Institute (CLSI) methods and included (in rank order) Candida albicans (587 isolates), C. glabrata (218), C. parapsilosis (196), C. tropicalis (126), C. krusei (24), C. lusitaniae (19), C. dubliniensis (12), C. guilliermondii (4), C. kefyr (4), C. famata (3), C. rugosa (2), C. haemulonii (1), C. inconspicua (1), C. lambica (1), C. norvegensis (1), C. pelliculosa (1), and C. sake (1). The collection also included Cryptococcus neoformans (43 isolates), Aspergillus fumigatus (60), and 19 other molds (data not shown: Aspergillus flavus [3], Aspergillus niger [3], Fusarium spp. [4], Penicillium spp. [3], Rhizopus spp. [2], Bipolaris sp. [1], and Mucor sp. [1], as well as 2 molds not identified to the species level). Laboratories were instructed to submit unique BSI and LRTI isolates obtained in consecutive order, allowing prevalence of the fungal isolates in participating centers to be determined.

All fungal isolates were identified at the participant's medical center by established laboratory methods in use at each institution and confirmed at the central reference laboratory (JMI Laboratories, North Liberty, IA) using Vitek (bioMerieux, Hazelwood, MO) and conventional reference procedures (12, 19). All yeasts were tested by broth microdilution using the CLSI M27-A3 (5) standardized reference method. Preparation of inocula for molds followed procedures described in the CLSI M38-A2 reference method for filamentous fungi (7). Quality control (QC) isolates C. krusei ATCC 6258 and C. parapsilosis ATCC 22019 were used, and all QC results were within published ranges (6).

Anidulafungin and voriconazole (Pfizer, Inc., New York, NY), amphotericin B, fluconazole, itraconazole, ketoconazole, and flucytosine (Sigma Chemical Co., St. Louis, MO), caspofungin (Merck Research Laboratories, Rahway, NJ), micafungin (Astellas Toyama Co., Ltd., Toyama, Japan), and posaconazole (Schering-Plough Research Institute, Kenilworth, NJ) were obtained as standard powders and prepared according to CLSI guidelines (5-7). The final concentration ranges (in μg/ml) were as follows: for anidulafungin, 0.001 to 32; for caspofungin and micafungin, 0.008 to 16; for amphotericin B, 0.12 to 8; for flucytosine and fluconazole, 0.5 to 64; for itraconazole, 0.015 to 2; and for posaconazole, voriconazole, and ketoconazole, 0.06 to 8. Antifungal dilution testing ranges were selected for maximal capture of MIC₅₀ and MIC₉₀ wild-type and mutant populations, including expanded ranges for newer and investigational agents to detect organism populations exhibiting potential resistance to these compounds. MIC values (yeasts and molds) and 90% minimal effective concentrations (MEC₉₀) (echinocandins, molds only) were determined as described in the CLSI reference methods (5, 7).

Table 1 displays the in vitro activities of 10 antifungal agents tested against yeast BSI isolates collected from the 2008 SENTRY Program. Anidulafungin was the most active agent against (MIC₉₀ in μg/ml) C. albicans (0.06), C. glabrata (0.12), C. tropicalis (0.06), and C. krusei (0.12) and was less potent against C. parapsilosis (MIC₉₀, 2 μg/ml) and C. guilliermondii (data not shown). The echinocandin potency against A. fumigatus was greatest for anidulafungin (MEC₉₀, 0.002 μg/ml) and caspofungin (MEC₉₀, 0.008 μg/ml) (Table 1). The results demonstrate the expanded utility of these agents against the most common mold species identified in lower respiratory tract infections.

The most active agents against Cryptococcus neoformans were the azoles voriconazole and ketoconazole (MIC₉₀, ≤0.06 μg/ml), itraconazole and posaconazole (MIC₉₀, 0.12 μg/ml), and fluconazole (MIC₉₀, 4 μg/ml). Susceptibility rates (MIC, ≤2 μg/ml) for the three echinocandins (Table 2) ranged from 98.4 to 99.9%, and these agents inhibited nearly all yeasts except C. neoformans. Yeast MIC values when tested against the echinocandins did not vary significantly for the four most common Candida spp. among the monitored geographic regions of this surveillance (Table 3) . However, some C. glabrata isolates displayed non-wild-type elevated MIC values for one or more echinocandins (MIC, ≥0.5 μg/ml), specifically, caspofungin (1 to >16 μg/ml), micafungin (0.25 to 8 μg/ml), and anidulafungin (1 to 4 μg/ml).

Elevated MIC values of echinocandin compounds have been associated with mutations within two highly conserved regions of fks1 and fks2 that encode the subunits of β-1,3-d-glucan synthase (GS), the target in the fungal cell wall (3). Six C. glabrata isolates were selected for fks1 hot spot 1 (HS1) and fks2 HS1 sequencing, since mutations in these regions have commonly been associated with elevated echinocandin MIC values and/or reduced susceptibility of GS to these compounds (8, 10). These strains were isolated in the United States (five strains, from Indiana, Ohio, and Washington) and Germany (one strain). DNA extraction was performed using a QIAamp DNA mini kit (Qiagen, Hilden, Germany). Singleplex PCRs were set up with generic or specific (C. glabrata) fks1 HS1 or fks2 HS1 primers (4). PCR amplicons were sequenced on both strands. The nucleotide sequence-deduced amino acid sequences were analyzed using the Lasergene software package (DNA STAR, Madison, WI). Sequences were then compared to other available sequences through Internet sources (http://www.ncbi.nlm.nih.gov/blast/ ).

Amino acid substitutions in the serine residue of position 645 in the fks1 and fks2 regions have been detected in several Candida species clinical isolates obtained from therapeutic failures or patients showing poor response to treatment with echinocandin compounds (8). Our results showed that three of the six C. glabrata strains harbored mutations encoding the S645P fks1 HS1 alteration, corroborating prior observations (8, 10), and that the three remaining isolates exhibited fks2 HS1 alterations (S645F, 1 strain; S645P, 2 strains).

The SENTRY Program surveillance of echinocandins and established antifungal agents demonstrates that the echinocandins continue to provide the most potent activity against yeasts isolated from BSI and A. fumigatus implicated in LRTI. Candida spp. (C. parapsilosis, C. guilliermondii, and some C. glabrata isolates) with less susceptible echinocandin profiles were detected with MIC values at or near the CLSI breakpoint of 2 μg/ml. However, recent findings by Arendrup et al. (2) have illustrated the challenges in using susceptibility testing methods for differentiating wild-type populations from fks HS mutants. In the SENTRY Program, follow-up sequencing of fks1 HS1 and fks2 HS1 regions confirmed strains with amino acid substitutions and reduced susceptibility to these agents. The SENTRY Program findings demonstrate the need for continued international surveillance to detect emerging resistance patterns among the classes of antifungal agents currently in clinical use. Correlation of higher or non-wild-type MIC values and genetic studies is critical in the recognition and elucidation of resistance mechanisms as well as the selection of appropriate antifungal interventions.

• Copyright © 2010 American Society for Microbiology