In Vitro Exposure to Increasing Micafungin Concentrations Easily Promotes Echinocandin Resistance in Candida glabrata Isolates

ABSTRACT

We assessed the in vitro susceptibility of five echinocandin-susceptible Candida glabrata isolates after exposure to micafungin. The direct exposure to plates at different micafungin concentrations resulted in the inhibition of growth at 0.062 μg/ml. The progressive exposure was performed on plates using 0.031 μg/ml of micafungin and sequential propagation on plates containing the next 2-fold concentration; the MICs of micafungin and anidulafungin increased sequentially, and all the isolates became echinocandin resistant, showing fks2 mutations.

TEXT

Echinocandins are currently recommended as the first-line treatment for invasive candidiasis (1–3). Although the resistance to echinocandins reported in Spain remains low (4), emerging Candida glabrata echinocandin-resistant isolates have been reported elsewhere (5, 6). We hypothesized that, as in the case of Aspergillus fumigatus (7), the in vitro exposure to increasing concentrations of echinocandins could promote the generation of mutations conferring resistance.

(This study was partially presented at the 26th European Congress of Clinical Microbiology and Infectious Diseases in Amsterdam, The Netherlands, 2016 [8].)

C. glabrata isolates from five patients with candidemia admitted to Ramón y Cajal Hospital (Madrid) were initially tested for antifungal susceptibility to micafungin (Astellas Pharma, Inc., Tokyo, Japan), anidulafungin, and fluconazole (Pfizer Pharmaceutical Group, New York, NY, USA) according to the EUCAST EDef 7.2 microdilution procedure (MIC_initial) (9–12).

The isolates were grown on chromogenic agar plates and incubated at 35°C for 24 h to check for purity. A loopful of cultured isolates was suspended in 10 ml of yeast extract-peptone-dextrose (YPD) broth (Becton Dickinson, Madrid, Spain) and incubated at 30°C overnight with vigorous shaking (150 to 160 × g) in an orbital shaker. Yeast cells were collected by centrifugation (3,000 × g for 5 min), and the pellet was resuspended in phosphate-buffered saline (PBS) (Oxoid, Hampshire, UK) and centrifuged again for washing. Washed suspensions were adjusted to 2 ×10⁹ to 4 ×10⁹ (mean, 2.94 ± 0.89 × 10⁹) CFU/ml using a Neubauer chamber and stroked on micafungin-containing plates under two different sets of conditions (direct exposure and progressive exposure). Sabouraud dextrose agar plates were prepared using eight different micafungin concentrations (0.015, 0.031, 0.062, 0.125, 0.25, 0.5, 1, and 2 μg/ml) for the exposure experiments, all of which were set up in duplicate to ensure reproducibility. In the direct micafungin exposure experiments, adjusted inocula (100 μl) were directly transferred to plates containing the eight different micafungin concentrations and the plates were incubated at 35°C and visually inspected for growth after 24 h. The MICs of micafungin and anidulafungin against the isolates growing on the plates containing the highest micafungin concentration were determined. In the progressive micafungin exposure experiments, adjusted inocula (100 μl) were stroked on plates containing micafungin at 0.031 μg/ml and the plates were incubated at 35°C for 24 h. If growth was observed, a loopful of cultured isolates was spread on the plate with a concentration of micafungin that was 2-fold greater. These steps were repeated up to the concentration of 2 μg/ml of micafungin. Each sequential suspension was used to study the MIC of micafungin and anidulafungin at each subsequent propagation step (MIC_subsequent) and final propagation step (MIC_final), and echinocandin and fluconazole MICs were determined using the plates containing 2 μg/ml of micafungin with the aid of EUCAST. Geometric means of MIC_initial, MIC_subsequent, and MIC_final of micafungin and anidulafungin were compared using the Wilcoxon signed-rank test (PASW Statistics 18.0; SPSS Inc., Chicago, IL). The comparisons were considered statistically significant with a P value of <0.05.

fks1 and fks2 genes from the isolates used to study the MIC_initial, MIC_subsequent, and MIC_final were amplified (13, 14). The stability of phenotypic and genotypic resistance was studied.

The isolates were genetically unrelated, fluconazole intermediate, echinocandin susceptible (MIC_initial), and wild type (Table 1). The direct micafungin exposure allowed all isolates to grow on plates containing micafungin at concentrations up to 0.031 μg/ml, but the echinocandin MICs studied in these isolates were identical to MIC_initial. The exposure to progressively increasing concentrations of micafungin allowed isolates to grow on all micafungin-containing plates. Both MIC_subsequent and MIC_final were significantly higher than the MIC_initial (Table 1). Overall, a trend toward higher echinocandin MICs was observed with increasing micafungin concentrations in the plates but with stable MICs of fluconazole. All isolates grown on plates with micafungin at 0.062 μg/ml or 0.125 μg/ml became resistant to anidulafungin and/or micafungin, but two relevant observations were made (Table 2). First, at 0.125 μg/ml, four isolates were micafungin and anidulafungin resistant and fks2 mutations were found; the remaining isolate (CG3) was resistant only to anidulafungin but had the wild-type fks2 gene. However, the MIC of anidulafungin against the CG3 isolate rose progressively and a deletion at F658 in fks2 was found when the isolate was grown on 0.25 μg/ml micafungin plates, although micafungin resistance was detected only in the last plate. Second, the CG1 isolate was anidulafungin and micafungin resistant on plates containing micafungin at 0.062 μg/ml, and two mutations (D666Y and S663P) in the fks2 gene were found; however, only the S663P substitution was found when isolates were grown on the plates with the subsequent concentration of micafungin. The fks substitutions and the phenotypic resistance were stable and reproducible after several propagations on micafungin-free agar plates. Genotyping showed that isolates were identical before and after micafungin progressive exposure, thus excluding any potential contamination of the isolates during the propagation steps.

TABLE 1.

Susceptibilities of C. glabrata isolates to micafungin and anidulafungin before and after progressive exposure to micafungin

Exposurea	MIC (μg/ml)
	Micafungin		Anidulafungin
	Range	Geometric mean	Range	Geometric mean
Initial	0.015	0.015	0.062	0.062
Progressive
Subsequent	0.015–2	0.18b	0.031–4	0.56b
Final	0.062–4	1.15c	1–4	2.30c

^aInitial, before progressive exposure to micafungin; Subsequent and Final, after exposure to micafungin.

^bData represent results of comparisons between the geometric means of the initial MIC and the subsequent MIC, for which significant differences were determined (P < 0.05).

^cData represent results of comparisons between the geometric means of the initial MIC and the final MIC, for which significant differences were determined (P < 0.05).

TABLE 2.

Antifungal activity of micafungin and anidulafungin against the five isolates before and after progressive micafungin exposure^a

Exposure	Result(s) for indicated isolate
	CG1		CG2		CG3		CG4		CG5
	EUCAST MIC (μg/ml) (MYC/AND)	fks2 mutation	EUCAST MIC (μg/ml) (MYC/AND)	fks2 mutation	EUCAST MIC (μg/ml) (MYC/AND)	fks2 mutation	EUCAST MIC (μg/ml) (MYC/AND)	fks2 mutation	EUCAST MIC (μg/ml) (MYC/AND)	fks2 mutation
Initial	0.015/0.062	WT	0.015/0.062	WT	0.015/0.062	WT	0.015/0.062	WT	0.015/0.062	WT
Subsequent (MYC concn in plates, μg/ml)
0.031	0.015/0.031	WT	0.015/0.062	WT	0.015/0.062	WT	0.015/0.031	WT	0.015/0.031	WT
0.062	0.5/2	S663P D666Y	0.015/0.062	WT	0.015/0.125	WT	0.015/0.062	WT	0.015/0.062	WT
0.125	1/2	S663P	0.125/0.5	S663Y	0.031/0.25	WT	2/2	Del F658	2/2	Del F658
0.25	1/2	S663P	0.125/1	S663Y	0.031/0.5	Del F658	2/2	Del F658	2/2	Del F658
0.5	2/2	S663P	0.25/2	S663Y	0.031/1	Del F658	2/2	Del F658	2/2	Del F658
1	2/4	S663P	0.5/2	S663Y	0.031/1	Del F658	2/2	Del F658	2/4	Del F658
2	4/4	S663P	1/2	S663Y	0.062/1	Del F658	2/2	Del F658	4/4	Del F658
Subsequent geometric mean of MICs	0.74/1.34		0.12/0.55		0.03/0.37		0.49/0.67		0.55/0.82

^aMYC, micafungin; AND, anidulafungin; Del, deletion. Boldface characters represent MICs above the clinical breakpoints and FKS point mutations.

The emergence of echinocandin resistance could be caused by the predisposition of this pathogen to easily acquire mutations in response to drug pressure due to its haploid nature and to alterations caused by mismatched repair genes (15). The five isolates studied became echinocandin resistant when grown even on plates containing low concentrations of micafungin. Micafungin penetrates slightly in the peritoneal fluid; the reported peritoneal fluid/plasma ratio (area under the concentration-time curve from 0 to 24 h [AUC_0–24]) is 0.3 (16). Shields et al. hypothesized that the abdomen of patients with previous exposure to echinocandins can be a hidden reservoir for mutant-resistant C. glabrata isolates (6). This suggests that the exposure to low echinocandin concentrations may promote the generation of mutant isolates that may potentially cause invasive infections. In our in vitro study, resistance was obtained at concentrations close to the MIC.

No resistant isolates were obtained with direct exposure. On the other hand, mutations were found in the progressive exposure experiment with the same isolates, suggesting that increasing micafungin concentrations may be effective in terms of selecting and enriching underrepresented mutant populations. In this sense, the CG1 isolate illustrates the phenomenon of coexistence of several populations, as two mutations were detected when the isolate was grown on plates containing low micafungin concentrations. The D666Y mutation confers weak resistance, while the S663P mutation confers strong resistance (17). However, only the S663P mutation was found when the isolate was grown on plates containing higher micafungin concentrations. CG3 became resistant to anidulafungin, and this isolate also showed micafungin resistance in the last step of progressive exposure. Some isolates can be resistant to anidulafungin and susceptible to micafungin (18). This supports the use of anidulafungin as a surrogate marker to predict echinocandin resistance. Furthermore, the deletion at position F658 does not correlate with a predictable pattern of susceptibility to candins.

Echinocandin resistance in C. glabrata has been associated with a loss of fitness in isolates with mutations at S663P in the fks2 gene (17). The fitness cost for the isolates carrying a mutation could explain the relatively low spread of resistant isolates reported to date. Further studies should be done on this topic.

This study had limitations. Despite of the low number of isolates analyzed, this proof of concept would be enhanced by the inclusion of a large number of clinical isolates. This was an in vitro study, and its impact in clinical practice is unknown, although our observations help improve understanding of the previous clinical reports on the presence of C. glabrata mutant isolates sourced from anatomical sites with low echinocandin concentrations. In conclusion, we found that a progressive exposure to increasing concentrations of micafungin can easily promote resistance to echinocandins in C. glabrata clinical isolates.