Comparative Evaluation of Etest, EUCAST, and CLSI Methods for Amphotericin B, Voriconazole, and Posaconazole against Clinically Relevant Fusarium Species

TEXT

Opportunistic infections due to Fusarium species are increasingly reported due to the rising numbers of immunocompromised patients, although immunocompetent individuals can also be infected (1). The most common opportunists are members of the F. solani species complex (SC), followed by F. oxysporum SC and F. fujikuroi SC (28). First-line antifungal treatment options are amphotericin B and voriconazole (2), while posaconazole has been recommended as salvage therapy. Posaconazole shows some in vitro activity against Fusarium spp., depending on the species (2, 3). Although Fusarium spp. are resistant to many antifungals, we previously reported that this resistance was species specific, and species identification is essential for early antifungal treatment (2, 4).

Reference methods for antifungal susceptibility testing (AFST) and breakpoints (BPs) for Candida and Aspergillus spp. have been developed (5). However, species-specific BPs have not yet been established for Fusarium spp. Recently, Espinel-Ingroff et al. (6) established the epidemiological cutoff values (ECVs) for Fusarium spp. in order to differentiate wild-type from non-wild-type isolates. Although a reproducible method for AFST of Fusarium spp. has been described by the Clinical and Laboratory Standards Institute (CLSI) (7) and by the European Committee for Antimicrobial Susceptibility Testing (EUCAST) (8), both are based on broth microdilution (BMD) and are time-consuming. Furthermore, the limited AFST data and the lack of comparison of the two reference BMD methods for Fusarium spp. prompted us to study the agreement between the CLSI and EUCAST methods for testing amphotericin B and triazoles against Fusarium spp. In addition, the Etest has been suggested as an alternative approach for AFST of amphotericin B or triazoles for non-Aspergillus molds in the clinical laboratory (9), but data on Fusarium spp. are limited. The objective of this study was to assess whether Etest could accurately measure the MICs of amphotericin B, voriconazole, and posaconazole for clinical Fusarium isolates in comparison with the CLSI and EUCAST methods.

Twenty clinical Fusarium isolates belonging to the species F. acutatum, F. chlamydosporum, F. delphinoides, F. dimerum, F. equiseti, F. fujikuroi, F. solani SC 5 (FSSC5), F. incarnatum, F. keratoplasticum, F. lichenicola, F. napiforme, F. oxysporum, F. petroliphilum, F. proliferatum, F. sacchari, FSSC6, F. subglutinans, F. temperatum, F. thapsinum, and F. verticillioides were subcultured on Sabouraud glucose agar (SDA, Difco) with 0.02% chloramphenicol for 5 to 7 days at 35°C to 37°C. Molecular identification was performed using translation elongation factor-1α (TEF1α) and the RNA polymerase (rPB2), as previously described (10). Strains were identified using the GenBank BLAST, Fusarium MLST, and Fusarium ID databases.

Susceptibility testing was carried using three different methods: BMD as described in the CLSI document M38-A2 (7), EUCAST E.Def 9.3 (8), and Etest manufacturer's guide (bioMérieux SA, Marcy-l’Étoile, France). Drug concentration ranges in BMD for amphotericin B (Bristol-Myers Squibb, Woerden, The Netherlands), voriconazole (Pfizer Central, Sandwich, UK), and posaconazole (Merck, Whitehouse Station, NJ, USA) were 0.016 to 32 μg/ml. For the CLSI M38-A2 and the EUCAST E.Def 9.3, the MICs of amphotericin B, voriconazole, and posaconazole were determined with an inverted magnifying mirror after 48 h at 35°C as the lowest drug concentration with complete inhibition of growth. For the Etest, the inoculum concentration was adjusted to 0.5 McFarland standard (equivalent to 1 × 10⁶ to 5 × 10⁶ CFU/ml). Then, 0.5 ml of this suspension was inoculated onto plates containing RPMI 1640 agar with 2% glucose using a cotton swab. After a period of 15 min, the Etest strips were applied and incubated for 48 h at 35°C. The reading of the Etest MICs was performed as described and illustrated in the Etest guide, and the MIC was determined as the concentration at the intercept of the elliptical complete inhibition zone using four ATCC strains (Aspergillus flavus ATCC 204304, A. flavus ATCC 204305, Candida parapsilosis ATCC 22019, and Candida krusei ATCC 6258) as quality controls. To directly compare the Etest MICs with the EUCAST and CLSI MICs, the Etest MICs were converted to the nearest highest 2-fold dilution value that matched the CLSI and EUCAST 2-fold dilution scheme. MICs were transformed to log₂ values, and differences were assessed statistically with a paired t test. The median and range of the log₂ MIC differences between the methods were calculated. Microbiologically significant differences between the methods were assessed, calculating the agreement within 1 and 2 2-fold dilutions. Finally, clinically significant differences were assessed by calculating categorical agreement among the three methods using previously determined ECVs for Fusarium spp., namely 4 μg/ml for amphotericin B, 4 μg/ml for voriconazole, and 2 μg/ml for posaconazole.

Table 1 summarizes the in vitro susceptibilities of 20 isolates of Fusarium spp. to amphotericin B, voriconazole, and posaconazole. The data are presented as MICs, MIC ranges, median MICs, and MIC₉₀. CLSI MIC results of amphotericin B spanned a narrow range of 1 to 4 μg/ml. Overall, amphotericin B and voriconazole showed the most potent activity with all three methods. The median, range, and MIC₉₀ with Etest, EUCAST, and CLSI were 2, 0.25 to >32, and 16 μg/ml; 1, 0.25 to 8, and 8 μg/ml; and 1, 0.5 to 4, and 4 μg/ml for amphotericin B; and 2, 0.25 to >32, and 4 μg/ml; 2, 0.5 to 16, and 4 μg/ml; and 2, 0.5 to >16, and 4 μg/ml for voriconazole, respectively. Posaconazole exhibited high MICs against F. chlamydosporum, F. dimerum, F. incarnatum, F. napiforme, F. oxysporum, F. proliferatum, F. thapsinum, and all the reported species within F. solani species complexes, with median, range, and MIC₉₀ of 32, 0.25 to >32, and 32 μg/ml for Etest, 16, 0.5 to >16, and 16 μg/ml for the EUCAST method, and 8, 0.25 to >16, and 16 μg/ml for the CLSI method, respectively.

Table 2 shows the analysis of the essential agreement (EA) between 48-h CLSI, EUCAST, and Etest MICs for each drug tested. The levels of agreement (within ±2 dilutions) between the results of the CLSI and the EUCAST method were 100% for amphotericin B, voriconazole, and posaconazole. The levels of agreement (within ±2 dilutions) between the results of the EUCAST method and the Etest were 100% for posaconazole and 95% for both amphotericin B and voriconazole. The agreement between the CLSI and the Etest was 90% for amphotericin B, 95% for voriconazole, and 85% for posaconazole. Amphotericin B and posaconazole MICs with the Etest tended to be significantly higher (P = 0.007 to 0.097) (median dilution difference, 1 to 1.5) than those of the two reference methods, indicating that Etest MIC distribution is shifted to the right compared to the EUCAST and the CLSI MIC distributions. However, the most of the differences remained within ±2 dilutions. The comparison of MICs obtained with the three methods resulted in a considerable Pearson correlation coefficient (PCC) ranging from 0.71 to 0.97. The correlation among Etest versus CLSI and Etest versus EUCAST methodologies to posaconazole was higher (0.89 and 0.97) than amphotericin B (0.71 and 0.86) and voriconazole (0.77 and 0.76) (Table 2). Moreover, the PCCs were statistically significant (P < 0.0001), indicating a good correlation between the MICs obtained by the CLSI or EUCAST and the Etest. The categorical agreement (CA) for all three methods and drugs was >85%, with slightly higher levels of agreement found between the Etest and the EUCAST (90 to 100%) than the CLSI (85 to 95%).

Our results indicated a good agreement between the three methods. The EA and CA between the Etest and the two reference methods were high for all three drugs (≥85%). The MIC results for amphotericin B spanned a range of 0.25 to 8 μg/ml with the CLSI and the EUCAST method and a wider range of 0.25 to >32 with the Etest. Wide MIC ranges were also found with posaconazole and voriconazole and all three methodologies, indicating considerable interspecies variations. High MICs (>4 μg/ml) were found for amphotericin B with F. keratoplasticum, F. napiforme, and FSSC6 for voriconazole with F. petrophilum and FSSC6 and for posaconazole with most species except F. acutatum, F. delphinoides, F. equiseti, F. fujikuroi, F. sacchari, F. subglutinans, F. temperatum, and F. verticillioides. Our results were comparable to those of other studies using BMD testing (1, 2, 11–16). Posaconazole displayed high MIC results, with values ranging from 0.25 to over 32 μg/ml against Fusarium species, with little difference when BMD and Etest methods were compared, which is in agreement with previously reported data (2, 17).

Although posaconazole showed high MICs for most isolates in the present study, case reports have demonstrated successful treatment of fusariosis with this drug. In one case report, F. solani did not respond to treatment with natamycin and amphotericin B but responded to posaconazole with a MIC of 1 μg/ml, and the patient recovered completely (18). Tu et al. (19) describes 3 cases of keratitis due to F. solani successfully treated with posaconazole. In another publication, an HIV-positive patient had onychomycosis due to F. falciforme and was successfully treated with posaconazole with an MIC of 0.5 μg/ml (20). In addition, Fusarium peritonitis and keratitis were treated with posaconazole (21, 22). Thus, posaconazole may be effective against isolates with low MICs (≤1 μg/ml).

Recently, Espinel-Ingroff et al. (6) used the CLSI methodology to determine the epidemiological cutoff values (ECVs) for two Fusarium species complexes, namely, F. oxysporum and F. solani, and F. verticillioides of the fujikuroi SC for amphotericin B, itraconazole, voriconazole, and posaconazole. However, no ECVs for rare Fusarium species were determined in that study. A comparison of CLSI and EUCAST versus Etest was performed for filamentous fungi, including a few strains of Fusarium (9, 23–26). Two studies compared CLSI and Etest: Lamoth and Alexander (9) studied 34 clinical Fusarium isolates with 94% agreement for amphotericin B and 100% for voriconazole and posaconazole, respectively, whereas Debourgogne et al. (27) reported lower overall agreement in FSSC only, with 73% for amphotericin B and 92% for voriconazole. Our study included molecularly identified Fusarium spp., and we also found high levels of agreement.

In conclusion, Etest overall resulted in 1-dilution-higher MICs than the reference methods, with most differences being within 2 dilutions, which may lead to errors if same breakpoints will be applied. However, the categorical agreement was high (>85%) using previously published ECVs. Etest can be used for routine susceptibility testing of amphotericin B, voriconazole, and posaconazole for Fusarium species. Further work is warranted in order to establish clinical breakpoints for Fusarium.