|Home | About | Journals | Submit | Contact Us | Français|
Posaconazole MICs for 50 Aspergillus fumigatus isolates with distinct genotypes were determined by three methods. MICs were ≥0.5 μg/ml for 5, 11, and 15 strains by the CLSI reference M38-A2, Etest (48-h), and oxygen consumption methods, respectively. The levels of categorical agreement between the results obtained by the CLSI method and those obtained by the oxygen consumption and Etest methods were 80 and 84%, respectively.
Invasive aspergillosis is the most common mold infection among immunocompromised patients, and Aspergillus fumigatus is the most common filamentous mold (12). The relationship between antifungal susceptibility and the potential for pathogenicity remains unknown, but strains with reduced susceptibilities to antifungal agents have been closely associated with high mortality rates in murine models (4, 11). The germination of A. fumigatus conidia to the hyphal form is a critical step in the progression to invasive disease and may represent a significant therapeutic target (8, 12). As a result of antifungal stress conditions, assessment of the oxygen consumption following fungal germination and growth has predicted patterns of A. fumigatus susceptibility to various antifungal agents (3).
In this study, we compared the susceptibilities of 50 A. fumigatus clinical and environmental isolates to posaconazole (Schering-Plough Farma) as determined by the reference Clinical and Laboratory Standards Institute (CLSI), oxygen consumption, and Etest methods. Clinical strains were isolated from lower respiratory tract and surgical wound specimens, while environmental strains were collected during routine ward air sampling. Genotyping of A. fumigatus isolates was performed by using a microsatellite single multiplex PCR based on tri-, tetra-, and pentanucleotide motifs (2), resulting in a distinct genotype for each isolate.
Reference posaconazole MICs were determined by the M38-A2 broth microdilution method and were defined by the absence of visual growth at 48 h (7). Quality control Candida parapsilosis ATCC 22019 and Candida krusei ATCC 6258 isolates were included each time testing was performed, and posaconazole MICs were within the expected limits (5). As reported previously (3), the evaluation of oxygen consumption was performed using 1.5-ml volumes of a conidial suspension in RPMI 1640 broth and two posaconazole concentrations (0.25 and 2 μg/ml). The tubes were agitated for 12 h at 37°C in a biological oxygen monitor (model 5300; YSI Inc.). Continuous monitoring and quantification of the available oxygen in the test suspensions were used for MIC determination (3). Etest susceptibility testing was performed on RPMI 1640 agar (Sigma) supplemented with 2% glucose (pH 7.0) according to the instructions of the Etest manufacturer (AB Biodisk); MICs were obtained after 24 and 48 h at 35°C.
Fungal strains were classified according to recently proposed epidemiological posaconazole cutoffs (for wild-type [WT] isolates, MIC ≤ 0.25 μg/ml; for non-WT isolates, MIC ≥ 0.5 μg/ml) (15). MICs at or below the susceptibility end point of 0.25 μg/ml are the most frequently observed posaconazole MICs, those for ≥90% of clinical isolates. Next, MIC pairs obtained by the reference method and the oxygen consumption or Etest method were compared, and they were considered to be in essential agreement when the difference was no more than three dilutions (e.g., 0.5, 1.0, and 2 μg/ml) (9). For the evaluation of categorical agreement, a discrepancy was considered to be a “very major error” when the MIC was above the cutoff end point by the reference method but below the epidemiological cutoff by either the oxygen consumption or Etest method (false susceptibility). When the result was above the cutoff end point by either one of the alternative methods and below the epidemiological cutoff by the reference method, the discrepancy was considered a “major error” (false resistance). Data were compared at a significance level of 0.05 by the analysis of variance test using the Bonferroni correction and by Student's t test for paired samples.
Posaconazole MICs for the 50 A. fumigatus strains ranged from 0.06 to 1 μg/ml by the M38-A2 method; MICs of 1 μg/ml were determined for three well-documented azole-resistant strains (two itraconazole resistant and one voriconazole resistant) (Table (Table1).1). The proportions of reference posaconazole MICs beyond the epidemiological end point have previously ranged between 5 and 20% for both environmental and clinical isolates (6, 9, 14-16). We have found similar results: 10% of our A. fumigatus strains were non-WT, with MICs for these strains exceeding the epidemiological cutoff (0.25 μg/ml). In our prior study, the oxygen consumption cutoff did not distinguish between posaconazole MICs of 0.25 and 1 μg/ml (3), because the cutoff was more suitable for the 1-μg/ml epidemiological end point proposed for itraconazole and voriconazole (15). After adjusting our oxygen consumption cutoff to 40%, we were able to identify isolates with posaconazole MICs beyond the epidemiological cutoff by using 0.25 μg/ml of posaconazole and 4 h of incubation. However, we observed a higher percentage of strains to be identified as non-WT according to posaconazole MICs by the oxygen consumption method (30%; 15 strains) than by the reference method; results were also higher by the Etest method (16% [8 strains] at 24 h and 22% [11 strains] at 48 h) (P < 0.05) (Table (Table1).1). Because of these findings, although the essential agreement among the three methodologies was excellent (90 to 96%), the categorical agreement was lower (80 to 88%) due to these false-resistant results, or major errors. On the other hand, the numbers of very major errors by both alternative methods were low (Table (Table2).2). Our levels of essential agreement were similar to or higher than those previously observed when reference and Etest results were compared (82 to 95%) (9, 14).
Based on phenotypic and molecular methods, Snelders et al. (17) have recently documented a higher incidence of A. fumigatus itraconazole-resistant strains than previously acknowledged in both clinical and environmental settings. It can be hypothesized that a high incidence of A. fumigatus strains with reduced susceptibilities to posaconazole also may be reported in the near future due to single-amino-acid substitutions (13). Furthermore, strains with both faster germination and greater growth fitness may alter their in vivo responses to antifungal agents, behaving as strains with reduced susceptibilities to these agents. Because susceptibility testing remains the most practical tool for the detection of either in vitro resistance or reduced susceptibility, more practical and/or rapid alternatives to the reference assay have been developed for testing A. fumigatus (1, 3, 10). Since essential agreement between results obtained by the reference and oxygen consumption methods was suitable, the oxygen consumption method has potential as an easier and more economic and rapid alternative for screening A. fumigatus isolates to determine posaconazole MICs either below or above the epidemiological cutoff after only 4 to 8 h. However, further examination of other posaconazole concentrations, as well as organism growth kinetics and resistance mechanisms, are warranted to better evaluate the suitability of this novel assay for testing A. fumigatus and other mold species.
We are grateful to Maria Luz Dias and Isabel Santos for their excellent technical assistance and Schering-Plough Farma (Cacém, Portugal) for providing posaconazole powder and Etest strips.
R.A. was cofinanced by the European Social Fund. IPATIMUP is partially supported by the Fundação para a Ciência e Tecnologia (FCT), the Programa Operacional Ciência e Inovação (POCI), and the Programa Operacional do Potencial Humano (POPH-QREN). We declare no conflict of interest.
Published ahead of print on 24 August 2009.