TEXT

Top Abstract Text References

Three recent advancements are having an important impact upon patient outcome for those who have mycotic infections. First, standardization of in vitro susceptibility testing (7) provides consistent and reproducible data that may predict clinical response when used in conjunction with individual patient risk factors. Second, the development and introduction of new antifungal agents, such as the triazole voriconazole, provide more options for treating a broad spectrum of etiologic agents in a wide range of clinical settings. Third, molecular genetics have allowed us to begin to understand phylogenetic relationships among pathogenic fungi and their environmental and phytopathogenic relatives.

When published susceptibility testing data is reviewed (2-5), a trend showing etiologic agent phylogenetic relatedness and response to antifungal drugs begins to emerge. The present study was undertaken to evaluate the in vitro sensitivity of a number of filamentous ascomycetes to amphotericin B, itraconazole, and voriconazole. The fungi evaluated were selected on the basis of their relatedness and similarity in the spectrum of mycoses that they cause as a means to determine if this hypothesis is correct.

Three hundred eighty-one isolates representing two classes, five orders, nine families, 30 genera, and 51 species (some of the genera and species having unknown phylogenetic relatedness) (Table 1) were maintained in the University of Texas Medical Branch (UTMB) culture collection until needed for testing. Isolates originated from humans and animals having hyalohyphomycosis, phaeohyphomycosis, chromoblastomycosis, or mycetoma.

Isolates maintained in the collection (8) were revived and subcultured on potato glucose agar in test tubes. After 24 to 72 h of incubation at 30°C, surface growth was aseptically removed, added to sterile saline, and adjusted to give a concentration equal to a McFarland 0.5 turbidity standard. For the quality control (QC) yeasts Candida albicans UTMB 3502 (ATCC 90028), Candida krusei UTMB 3849 (ATCC 6258), and Candida parapsilosis UTMB 3848 (ATCC 22019) and the filamentous ascomycetes, this process resulted in approximately 1 × 10⁶ to 5 × 10⁶ CFU/ml. The final inoculum concentration in the drug dilutions was 1 × 10⁴ to 5 × 10⁴ CFU/ml. Growth controls and purity plate counts to verify the inoculum concentration, as well as the QC isolates, were run with each set of tests.

By using a broth macrodilution method based upon National Committee for Clinical Laboratory Standards approved standard M27-A (7), voriconazole (Pfizer Central Research, Sandwich, United Kingdom), amphotericin B (USPC), and itraconazole (USPC) were evaluated against the test isolates. The drugs were dissolved in dimethylformamide, dimethyl sulfoxide, and polyethylene glycol, respectively. Each drug was prepared in 10-fold concentrations, with voriconazole and itraconazole in RPMI 1640 (American Biorganices, Inc., Niagara Falls, N.Y.) and amphotericin B in Bacto Antibiotic Medium 3 (Difco Laboratories, Detroit, Mich.). Tenfold drug concentrations prepared in twofold serial dilutions (0.1-ml drug volume), maintained at 70°C until needed, were thawed, and then 0.9 ml of inoculum was added to each tube to achieve the final drug concentrations.

Following incubation at 35°C, the tests were read at 24, 48, and 72 h. A few isolates required an incubation temperature of 30°C for adequate growth. The tests were considered valid when the controls and QC isolates had the expected values. The MICs of voriconazole and itraconazole were the drug dilutions that resulted in 80% or greater reduction in turbidity in comparison to that of the drug-free growth control. Amphotericin B MICs were the drug dilutions that resulted in an optically clear solution. The lowest and highest drug dilutions tested were considered end points for the dilution series if they were determined to be the final MICs.

The geometric mean MICs of voriconazole and itraconazole were essentially the same (Table 1). Members of the Microascaceae, Pseudallescheria boydii (5), Scedosporium prolificans, and Scopulariopsis brumptii, were more sensitive to voriconazole than to amphotericin B and itraconazole. The geometric mean MICs of voriconazole, amphotericin B, and itraconazole for P. boydii were 0.22, 13.8, and 1.57 µg/ml, respectively (unpublished data and reference 7). Overall, this clade within the Microascales is more resistant than other fungal groups to the three antifungal drugs tested (Table 2), except for P. boydii, which is sensitive to voriconazole (7). This correlates well with the poor response of patients with infections caused by these fungi (6) to treatment with either amphotericin B or itraconazole.

If an amphotericin B MIC of greater than 1 µg/ml reflects likely isolate resistance as observed for Candida spp. (7), then the clades Microascaceae, Chaetomiaceae, and Mycosphaerellaceae are resistant. In contrast, members of these clades except for the Microascaceae (excluding P. boydii) should be sensitive (7) to the triazoles tested. When the sensitivity of fungi of unknown phylogeny is considered (Table 1), the drug MICs for these fungi are also low. While the voriconazole MICs are consistently lower for this heterogeneous group, higher values for amphotericin B and itraconazole were recorded for Phaeoacremonium parasiticium, Phialemonium curvatum, and Phialemonium obovatum.

The majority of antifungal drugs in current use target various components of the ergosterol biosynthetic pathway. Amphotericin B, in contrast to the azoles, binds to ergosterol in the fungal cell membrane, where it disrupts glycolysis and respiration by changing the membrane's permeability properties (6). The triazoles itraconazole and voriconazole interfere with the ergosterol biosynthetic pathway by hindering the C-14 demethylation of lanosterol. C-14 demethylase, a microsomal cytochrome P-450 enzyme, catalyzes the removal of the C-14 methyl group from either lanosterol (yeasts) or 24-methylene dihydrolanosterol (molds) (10). This might help to explain why fluconazole is effective against yeasts but not molds. Likewise, phylogenetically associated differences and similarities involving the multistep oxidative removal of the C-14 methyl group in the Microascaceae clade, in contrast to the other clades, might explain why fungi like P. boydii are resistant to therapy. Our preliminary data indicates that as a group, this clade should be responsive to the drugs we tested. Voriconazole was obviously more effective than the other two drugs.

Weete and Gandhi (10) have recently reviewed the biochemistry and molecular biology of fungal sterols. They have shown that the ergosterol biosynthetic pathway can be thought of in terms of an evolutionary process. Our data indicates that in vitro antifungal susceptibility testing data could be related to clade-specific variations. Whether the results reflect differences in the biosynthetic pathway or resistance mechanisms such as efflux pumps needs to be determined.

Additional molecular data will be required before the relatedness of the other fungi in category III (Table 1) can be elucidated. Under the appropriate conditions, isolates of Madurella grisea will form pycnidia identical to those of Pyrenochaeta romeroi (1). Knowing that Pyrenochaeta and Phoma species are associated with ascomycetes classified in the Lophiostomataceae, we have placed M. grisea among the Pleosporales. The MIC for the strain we tested is compatible with this conclusion. Madurella mycetomatis and M. grisea appear to be a heterogeneous group (11). Phialides can be seen in some isolates (9) that are like those of Myrioconium spp. The teleomorphs of Myrioconium spp. are formed by species of Myriosclerotinia, a genus classified in the apothecial ascomycetous family Sclerotiniaceae. The phylogenic positions of the remaining fungi are unknown.

The MIC data indicates that voriconazole is active against the majority of filamentous ascomycetes tested. This activity can be extended to numerous additional fungi (2-5). The trend of in vitro antifungal susceptibility testing data to parallel phylogenetically based clades can have an important impact upon selection of antifungal agents, especially as new opportunistic pathogens whose sensitivity to antifungal drugs is unknown are recovered. The low MICs for the isolates in the Microascaceae indicate that voriconazole should be considered for patients with infections caused by members of this clade.