Comparative Effects of Micafungin, Caspofungin, and Anidulafungin against a Difficult-To-Treat Fungal Opportunistic Pathogen, Candida glabrata

Abstract

The aim of this study was to compare the in vitro and in vivo activities of micafungin, caspofungin, and anidulafungin against Candida glabrata. The MICs against 28 clinical isolates showed that the overall susceptibilities to caspofungin and to micafungin were not statistically different in the absence of human serum, whereas the isolates were less susceptible to micafungin than to caspofungin in its presence. Minimum fungicidal concentrations, as well as time-kill experiments, showed that caspofungin was more active than anidulafungin, while micafungin was superior to either caspofungin or anidulafungin without serum; its addition rendered caspofungin and micafungin equally effective. A murine model of systemic candidiasis against a C. glabrata-susceptible isolate was performed to study the effects of all three echinocandins, and kidney burden counts showed that caspofungin, micafungin, and anidulafungin were active starting from 0.25, 1, and 5 mg/kg of body weight/day, respectively. Two echinocandin-resistant strains of C. glabrata were selected: C. glabrata 30, a laboratory strain harboring the mutation Fks2p-P667T, and C. glabrata 51, a clinical isolate harboring the mutation Fks2p-D666G. Micafungin activity was shown to be as effective as or more effective than that of caspofungin or anidulafungin in terms of MICs. In vivo studies against these resistant strains showed that micafungin was active starting from 1 mg/kg/day, while caspofungin was effective only when administrated at higher doses of 5 or 10 mg/kg/day. Although a trend toward colony reduction was observed with the highest doses of anidulafungin, a significant statistical difference was never reached.

INTRODUCTION

Candida glabrata has recently been reported to be the second most common cause of invasive candidiasis, and there are increasing amounts of data showing its important role in determining either superficial or deep-seated infections (4, 18). Systemic infections due to C. glabrata are characterized by a high mortality rate, and they are difficult to treat due to reduced susceptibility of the species to azole drugs, especially to fluconazole (26). According to the published guidelines, amphotericin B can be used to treat infections due to C. glabrata, especially in profoundly immunocompromised hosts (22). Fortunately, the species also appears to be highly susceptible to the echinocandins (i.e., caspofungin, anidulafungin, and micafungin), making these agents valuable alternatives as first-line therapy against this Candida species (22). Interestingly, patients suffering from systemic candidiasis due to C. glabrata showed a trend, although not a significant one, to a better clinical outcome when treated with micafungin (87%) rather than with caspofungin (67%) (23).

The three echinocandin antifungal agents anidulafungin, caspofungin, and micafungin have a unique mechanism of action, inhibiting β-1,3-d-glucan synthase, an enzyme that is necessary for the synthesis of β-1,3-d-glucan polymers, essential components of the cell walls of several fungi (8).

Although resistance to echinocandins remains relatively low, a growing number of reports have been published about breakthrough infections involving high-MIC isolates in patients receiving echinocandin therapy (29, 31).

The aim of this study was to compare the effects of all three FDA-approved echinocandins against C. glabrata. In particular, the potential fungicidal activities of these compounds were investigated both in vitro and in a neutropenic murine model of systemic infection due to echinocandin-susceptible and -resistant strains of C. glabrata.

MATERIALS AND METHODS

In vitro studies. (i) Isolates.

A total of 30 isolates of C. glabrata were studied: 21 were isolated from blood samples, while the other 7 were recovered from the gastrointestinal, respiratory, and urinary tracts. Each strain represented a unique isolate from a patient. Additionally, two echinocandin-resistant strains were evaluated: C. glabrata 51, a clinical isolate (provided by M. Sanguinetti) that harbors a mutation in FKS2 (A1997G; Fks2p-D666G), as well a laboratory strain, C. glabrata 30, which harbors another mutation in FKS2 (C1999A; Fks2p-P667T). The latter strain was selected in vitro by plating C. glabrata 11 (caspofungin MIC, 0.03 μg/ml) on 20 μg/ml caspofungin-containing yeast extract-peptone-dextrose (YPD) agar plates. Yeast isolates were identified at the species level by conventional morphological and biochemical methods and were stored at −70°C in 10% glycerol. Before the initiation of the study, yeast isolates were subcultured on antimicrobial-agent-free medium to ensure viability and purity.

(ii) Antifungal drugs.

Amphotericin B was used as pure powder (Sigma Chemical, Milan, Italy) for in vitro studies and as a commercial preparation (Fungizone; Bristol-Myers Squibb S.p.A.) for in vivo studies. Anidulafungin (Pfizer, Inc.), caspofungin (Merck & Co., Inc.), and micafungin (Astellas Pharma Inc.) were supplied as pure powders and used for the in vitro and in vivo experiments.

For in vitro studies, stock solutions of amphotericin B and anidulafungin were prepared in dimethyl sulfoxide (DMSO) (Sigma), while caspofungin and micafungin were prepared in sterile water. Further dilutions were prepared in test medium. For in vivo studies, solutions of amphotericin B, anidulafungin, caspofungin, and micafungin were prepared following the manufacturers' instructions.

(iii) Broth dilution.

Antifungal susceptibility testing was performed by the broth microdilution method in accordance with the Clinical and Laboratory Standards Institute (CLSI) document M27-A3 (7). The final concentrations of drugs ranged from 0.002 to 2.0 μg/ml for the susceptible strains and from 0.03 to 16 μg/ml for the resistant strains. Plates were incubated at 35°C for 24 h (48 h for amphotericin B), and readings were performed both visually and spectrophotometrically. Anidulafungin, caspofungin, and micafungin MICs were considered to be the first concentrations of the antifungal agent at which the turbidity in the well was 50% less than that in the control well. Amphotericin B MICs were considered to be the first concentration of the antifungal agent at which the turbidity in the well was 100% less than that in the control well. Each test was run in quintuplicate and repeated on 2 different days. MIC values are the median MIC of 5 replicate determinations on 2 different days.

After MIC readings, the whole suspension from each well above the MIC were withdrawn and plated in duplicate onto 150-mm Sabouraud dextrose agar (SDA) plates. The inoculated plates were incubated at 35°C, and minimum fungicidal concentrations (MFCs) were recorded after 48 to 72 h. The MFC was defined as the lowest concentration of drug yielding no growth (27). Each isolate was tested in triplicate.

Antifungal susceptibility testing was performed as described above with and without 50% human serum in the medium (21, 33). The human serum was prepared from the blood of healthy volunteers.

C. tropicalis ATCC 750 was utilized as an additional strain, and C. parapsilosis ATCC 22019 was included on each day of testing as a CLSI-recommended quality control strain.

(iv) Killing curves.

Killing experiments were performed against two strains (C. glabrata 11, and C. glabrata 30). Briefly, three to five colonies of each isolate from a 48-h growth plate were suspended in 10 ml of sterile distilled water, and the turbidity was adjusted using spectrophotometric methods to 0.5 McFarland standard (approximately 1 × 10⁶ to 5 × 10⁶ CFU/ml). The solution was further diluted at 1:50 and 1:10 with RPMI 1640 medium buffered with morpholinepropanesulfonic acid (MOPS), and 250 μl of the adjusted fungal suspension was added to 250 μl of either RPMI 1640 medium buffered with MOPS free of drug or the growth medium plus an appropriate amount of drug. Drugs were used at final concentrations of 0.25×, 1×, 4×, 32×, 64×, 128×, and 256× the MIC. Test solutions were placed on a shaker and incubated at 35°C. At 0, 4, and 24 h following the introduction of the test isolate into the system, 50-μl aliquots were removed from each test solution. After 10-fold serial dilutions, 50-μl aliquots of undiluted and diluted samples were streaked onto SDA plates for colony count determination. Following incubation at 35°C for 48 to 72 h, the number of CFU on each plate was determined. Fungicidal activity was considered to be achieved when the number of CFU per milliliter was <99.9% of the initial inoculum size (27). The limit of detection was 19 CFU. Experiments were performed in triplicate.

FKS gene sequence analysis.

Sequencing of partial portions of the FKS1 and FKS2 genes was performed using previously described primers (30). DNA from yeast cultures grown at 37°C overnight in YPD (2% yeast extract, 4% Bacto Peptone, 4% dextrose) broth medium was extracted using the QIAamp DNA Mini Kit (Qiagen, Milan, Italy) according to the manufacturer's instructions. PCR products were sequenced, the sequences of both strands were analyzed using a 3130 Genetic Analyzer (Applied Biosystems, Foster City, CA), and multiple-amino-acid alignments were derived by using Vector NTI Advance (version 11.5; Invitrogen Inc., Milan, Italy).

In vivo studies.

CD1 female mice (Charles River, Calco, Italy) weighing 25 g were rendered neutropenic by intraperitoneal (i.p.) administration of cyclophosphamide (200 mg/kg of body weight/day) on day −4 and days 1 and 4 postinfection.

The mice were infected intravenously with 1.35 × 10⁸ CFU/mouse of C. glabrata 11, with 1.6 × 10⁸ CFU/mouse of C. glabrata 30, and with 1.2 × 10⁸ CFU/mouse of C. glabrata 51 given in a 0.2-ml volume. All drugs were administered i.p. for 6 consecutive days in a 0.2-ml volume, starting 24 h postchallenge. Each echinocandin was given at 0.25, 1, 5, and 10 mg/kg/day, while amphotericin B (the control drug) was given at 1 mg/kg/day. Drug efficacy was assessed by determining the number of CFU per kidney pair. Briefly, the mice were sacrificed, the kidneys were homogenized, and diluted or undiluted aliquots, including the entire organ, were grown on SDA plates for colony count determination. Tissue burden experiments were performed on days 3, 5, and 7 postinfection, which corresponded to a total of 2, 4, and 6 days of therapy, respectively. The susceptibilities of C. glabrata 30 to the echinocandins were determined after 7 days of animal treatments. Briefly, colonies of the tissue burden counts obtained either from control or treatment groups were analyzed, and the MIC values did not show any difference from the initial strain. There were from 6 to 9 animals in each group. Animal experiments were conducted with the approval of the University of Ancona Ethics Committee.

Statistical analysis.

In vitro and in vivo studies were compared among the different groups using the Kruskal-Wallis test and Dunn's test for post hoc comparisons between each treatment group and the control group (Prism 5; GraphPad Software). Two-sided P values of <0.05 were considered statistically significant.

RESULTS

The in vitro susceptibilities of 28 clinical isolates of C. glabrata against amphotericin B, anidulafungin, caspofungin, and micafungin, either with or without 50% human serum, are reported in Table 1 and Fig. 1. The geometric mean MICs of anidulafungin, caspofungin, and micafungin were 0.10, 0.04, and 0.02 μg/ml, respectively. In the presence of human serum, the geometric mean MICs of anidulafungin, caspofungin, and micafungin increased to 1.08, 0.32, and 0.62 μg/ml, respectively. Multiple-comparison analysis of MIC values showed that the overall susceptibilities to caspofungin and to micafungin were not statistically different in the absence of human serum, whereas the isolates were less susceptible to micafungin than to caspofungin in its presence (P < 0.05). In general, anidulafungin was the less active echinocandin against the C. glabrata isolates. The MFCs obtained without serum showed that caspofungin was more active than anidulafungin, while micafungin was superior to either caspofungin or anidulafungin. However, the addition of serum rendered caspofungin and micafungin equally effective. Again, both drugs were more effective than anidulafungin (Fig. 1).

Table 1.

In vitro susceptibilities of 28 clinical isolates of C. glabrata to amphotericin B, anidulafungin, caspofungin, and micafungin^a

Isolate	Median MIC (μg/ml)b								Median MFC (μg/ml)c
	RPMI				RPMI plus 50% serum				RPMI				RPMI plus 50% serum
	AMB	ANID	CAS	MICA	AMB	ANID	CAS	MICA	AMB	ANID	CAS	MICA	AMB	ANID	CAS	MICA
C. tropicalis ATCC 750	0.5	0.06	0.06	0.06	0.5	0.25	0.125	1.0	1.0	0.06	0.06	0.125	0.5	0.5	0.125	1.0
C. parapsilosis ATCC 22019	0.5	0.25	0.25	0.5	0.5	>2.0	1.0	>2.0	1.0	0.5	0.5	0.5	0.5	>2.0	1.0	>2.0
C. glabrata
1	1.0	0.25	0.03	0.016	0.5	2.0	0.25	1.0	2.0	2.0	1.0	0.03	2.0	>2.0	0.5	2.0
2	0.5	0.06	0.008	0.008	0.5	1.0	1.0	0.5	1.0	1.0	>2.0	0.06	1.0	>2.0	2.0	1.0
3	0.5	0.06	0.03	0.016	0.5	1.0	0.5	0.5	1.0	1.0	0.5	0.125	1.0	>2.0	1.0	1.0
4	2.0	0.50	0.03	0.03	0.5	2.0	0.25	1.0	1.0	2.0	1.0	0.03	2.0	>2.0	0.5	2.0
5	1.0	0.06	0.03	0.016	0.5	2.0	0.25	1.0	2.0	2.0	0.5	0.03	2.0	>2.0	0.5	1.0
6	1.0	0.06	0.06	0.016	0.5	2.0	0.25	1.0	2.0	2.0	1.0	0.06	1.0	>2.0	1.0	2.0
7	1.0	0.06	0.06	0.016	0.5	1.0	0.25	0.5	2.0	1.0	1.0	0.125	1.0	>2.0	1.0	2.0
8	1.0	0.03	0.06	0.03	0.5	2.0	0.25	1.0	2.0	>2.0	0.5	0.125	2.0	>2.0	0.5	2.0
9	2.0	0.06	0.03	0.016	0.5	1.0	0.25	0.5	2.0	2.0	1.0	0.03	1.0	>2.0	2.0	1.0
10	1.0	0.06	0.06	0.016	0.5	1.0	0.25	0.5	2.0	1.0	1.0	0.06	1.0	>2.0	2.0	1.0
11	2.0	0.06	0.03	0.03	0.5	0.5	0.25	0.5	2.0	1.0	0.5	0.06	1.0	>2.0	2.0	0.5
12	1.0	0.125	0.03	0.016	0.25	1.0	0.25	0.5	2.0	1.0	0.5	0.03	0.5	>2.0	0.5	2.0
13	1.0	0.125	0.016	0.03	0.5	1.0	0.25	0.5	2.0	2.0	1.0	0.06	1.0	>2.0	0.5	1.0
14	1.0	0.125	0.03	0.03	0.5	1.0	0.25	0.5	2.0	>2.0	1.0	0.06	2.0	>2.0	0.5	1.0
15	1.0	0.125	0.03	0.03	0.5	1.0	0.5	0.5	2.0	1.0	0.5	0.06	2.0	>2.0	0.5	1.0
16	1.0	0.125	0.06	0.03	0.5	1.0	0.5	0.5	2.0	2.0	1.0	0.125	1.0	>2.0	1.0	1.0
17	1.0	0.125	0.03	0.03	1.0	1.0	0.5	0.5	2.0	>2.0	0.125	0.06	1.0	>2.0	0.5	1.0
18	2.0	0.125	0.03	0.03	2.0	1.0	0.25	1.0	2.0	1.0	0.5	0.06	2.0	>2.0	0.5	1.0
19	2.0	0.125	0.03	0.03	2.0	1.0	0.25	1.0	2.0	1.0	0.5	0.06	2.0	>2.0	0.5	1.0
20	2.0	0.06	0.03	0.03	2.0	1.0	0.25	1.0	2.0	2.0	0.5	0.06	2.0	>2.0	0.5	1.0
21	1.0	0.06	0.06	0.03	1.0	1.0	0.25	0.5	2.0	1.0	0.5	0.06	1.0	>2.0	0.5	1.0
22	1.0	0.125	0.06	0.03	1.0	1.0	0.25	0.5	2.0	1.0	0.5	0.06	1.0	>2.0	0.5	1.0
23	1.0	0.06	0.06	0.016	1.0	1.0	0.5	1.0	2.0	>2.0	0.5	0.06	1.0	>2.0	0.5	0.5
24	1.0	1.0	0.03	0.008	0.5	1.0	0.25	0.5	2.0	1.0	0.06	0.016	1.0	>2.0	0.5	1.0
26	1.0	0.125	0.06	0.008	1.0	0.5	0.25	0.25	2.0	0.5	1.0	0.016	1.0	>2.0	1.0	0.5
27	1.0	0.125	0.125	0.03	0.5	1.0	0.5	1.0	2.0	2.0	2.0	0.125	1.0	>2.0	1.0	1.0
28	0.5	0.125	0.06	0.03	0.5	1.0	0.25	0.5	2.0	1.0	1.0	0.125	1.0	>2.0	1.0	0.5
29	0.5	0.25	0.125	0.016	0.5	1.0	0.5	0.5	1.0	1.0	2.0	0.06	0.25	>2.0	1.0	0.5

^aAMB, amphotericin B; ANID, anidulafungin; CAS, caspofungin; MICA, micafungin.

^bThe AMB MIC was defined as the lowest drug concentration at which there was complete inhibition of growth after 48 h of incubation; the ANID, CAS, and MICA MICs were defined as the lowest concentration at which there was a visually prominent reduction in growth (approximately 50%) relative to the drug-free growth control after 24 h of incubation.

^cThe MFC was defined as the lowest concentration of antifungal compound yielding no growth.

Fig 1.

Geometric means of the MICs (top) and MFCs (bottom) of C. glabrata clinical isolates (n = 28) with or without serum (S). The error bars represent the 95% confidence Intervals. Φ, P < 0.05 for caspofungin (CAS) versus anidulafungin (ANID); ¶, P < 0.05 for micafungin (MICA) versus CAS; §, P < 0.05 for MICA versus ANID.

Then, C. glabrata 11 was selected to compare the fungicidal activities of all the drugs (Fig. 2). In these experiments, drugs were utilized at concentrations of 0.25×, 1×, 4×, and 32× the MIC with or without human serum. After 24 h of incubation, amphotericin B yielded killing activity at a concentration of 32× the MIC, regardless of the absence or presence of serum. In the absence of serum, micafungin exerted fungicidal activity starting from 1× the MIC after 24 h of incubation. At the same time interval, caspofungin was fungicidal at doses of 4× and 32× the MIC, while anidulafungin yielded a fungicidal effect at 32× the MIC. In the presence of serum, micafungin, and caspofungin were fungicidal at 4× and 32× MIC, while anidulafungin did not reach fungicidal activity.

Fig 2.

Time-kill plots for amphotericin B, anidulafungin, caspofungin, and micafungin against C. glabrata isolate 11. Each data point represents the mean result ± SEM (standard error of the mean). The dotted lines represent a 99.9% growth reduction compared with that of the initial inoculum (fungicidal effect). ●, control; □, 0.25× MIC; △, 1× MIC; ▽, 4× MIC; ♢, 32× MIC.

C. glabrata 11 was also utilized for in vivo studies. Mice were infected with 1.35 × 10⁸ CFU/mouse, and the drug activities were studied on days 3, 5, and 7 postinfection. The results for kidney tissue burden obtained with this strain are reported in Fig. 3. After 2 days of treatment, caspofungin was active starting at a dose of 1 mg/kg/day, while micafungin was active at doses of 5 and 10 mg/kg/day. On day 5 postinfection, all doses of caspofungin were effective at reducing the burden. Micafungin was active starting at a dose of 1 mg/kg/day. Although, anidulafungin showed a trend in reduction, the agent did not significantly decrease the fungal burden with respect to the control after 2 and 4 days of treatment at any tested doses. On day 7 postinfection, all three echinocandins at 5 and 10 mg/kg/day were effective at reducing the counts against the controls. At 1 mg/kg/day, either caspofungin or micafungin, but not anidulafungin, was active. At the lowest dose tested (0.25 mg/kg/day), only caspofungin was active.

Fig 3.

Kidney tissue burdens of neutropenic CD1 mice infected intravenously with C. glabrata 11. The mice were treated daily with amphotericin B at 1 mg/kg/day (■) and anidulafungin (△), caspofungin (□), or micafungin (○) at doses of 0.25, 1, 5, and 10 mg/kg/day. C, control. Tissue burden experiments were performed on days 3, 5, and 7 postinfection. There were from 6 to 9 animals in each group. The horizontal lines represent the median fungal burdens. *, P < 0.05 versus the results for the control.

In order to investigate the effects of anidulafungin, caspofungin, and micafungin against echinocandin-resistant isolates of C. glabrata, two additional strains were selected for further in vitro and in vivo studies. C. glabrata 30 is a laboratory strain harboring a mutation in FKS2 (C1999A; Fks2p-P667T), while C. glabrata 51 is a clinical isolate harboring another FKS2 mutation (A1997G; Fks2p-D666G). In vitro susceptibility tests of these two strains are shown in Table 2. Time-kill studies were performed against C. glabrata 30, and the results are shown in Fig. 4. In these experiments, drugs were utilized at concentrations of 0.25×, 1×, 4×, 32×, 64×, 128×, and 256× the MIC with or without human serum. Amphotericin B yielded killing activity against this isolate after 24 h of incubation at a concentration of 32× the MIC either with or without serum. Anidulafungin, caspofungin, and micafungin exerted fungicidal activity at 128× the MIC in the absence of serum after 24 h of incubation; in the presence of serum, all three echinocandins were not active against this isolate of C. glabrata.

Table 2.

In vitro susceptibility to amphotericin B, anidulafungin, caspofungin, and micafungin of two echinocandin-resistant strains of C. glabrata

C. glabrata isolatea	Median MIC (μg/ml)b
	RPMI				RPMI plus 50% serum
	AMB	ANID	CAS	MICA	AMB	ANID	CAS	MICA
30	1.0	0.5	2.0	0.5	1.0	1.0	0.5	0.5
51	1.0	1.0	2.0	0.25	>16	2.0	2.0	0.5

^aC. glabrata 30 is a laboratory strain selected in vitro by plating the isolate C. glabrata 11 on 20 μg/ml caspofungin-containing YPD agar plates (Fks2p-P667T); C. glabrata 51 is a clinical isolate bearing a mutation in the FKS2 gene (Fks2p-D666G).

^bAMB, amphotericin B; ANID, anidulafungin; CAS, caspofungin; MICA, micafungin. The AMB MIC was defined as the lowest drug concentration at which there was complete inhibition of growth after 48 h of incubation; the ANID, CAS, and MICA MICs were defined as the lowest concentration at which there was a visually prominent reduction in growth (approximately 50%) relative to the drug-free growth control after 24 h of incubation.

Fig 4.

Time-kill plots for amphotericin B, anidulafungin, caspofungin, and micafungin against C. glabrata isolate 30. Each data point represents the mean result ± SEM. The dotted lines represent a 99.9% growth reduction compared with that of the initial inoculum (fungicidal effect). ●, control; □, 0.25× MIC; △, 1× MIC; ▽, 4× MIC; ♢, 32× MIC; ×, 64× MIC; *, 128× MIC; 196, 256× MIC.

Both resistant strains of C. glabrata were utilized to compare the efficacies of the echinocandins in vivo, and the results are shown in Fig. 5. Against C. glabrata 30, micafungin was effective at doses of 1, 5, and 10 mg/kg/day and caspofungin was active at 5 and 10 mg/kg/day. Similarly, against C. glabrata 51, micafungin was effective at doses of 1, 5, and 10 mg/kg/day, while caspofungin was active only at a dose of 10 mg/kg/day. Although for both strains anidulafungin at 5 and 10 mg/kg/day showed a trend toward reduction of CFU with respect to the controls, a statistically significant difference was never reached by using the multiple-comparison analyses.

Fig 5.

Kidney tissue burdens of neutropenic CD1 mice infected intravenously with C. glabrata 30 and C. glabrata 51. The mice were treated daily with anidulafungin (△), caspofungin (□), or micafungin (○) at doses of 0.25, 1, 5, and 10 mg/kg/day. C, control. Tissue burden experiments were performed on day 7 postinfection. There were 8 or 9 animals in each group. The horizontal lines represent the median fungal burdens. *, P < 0.05 versus the results for the control.

DISCUSSION

Our findings showed that all 28 clinical isolates recovered from patients hospitalized in our department presented anidulafungin, caspofungin, and micafungin MICs within the previously reported ranges for wild-type strains of C. glabrata (3, 5, 24, 25). Also, our results showed that the MICs were within the susceptibility ranges for all three echinocandins, with the exception of two isolates showing an intermediate MIC value for anidulafungin (0.25 μg/ml). In fact, the recently proposed CLSI clinical interpretive MIC breakpoints (CBPs) for anidulafungin and caspofungin against C. glabrata are ≤0.12 μg/ml for susceptible isolates (S), 0.25 μg/ml for intermediate isolates (I), and ≥0.5 μg/ml for resistant isolates (R), while the CPBs for micafungin are ≤0.06 μg/ml for S isolates, 0.12 μg/ml for I isolates, and ≥0.25 μg/ml for R isolates (24).

It has been reported that echinocandins exert fungicidal activity against yeasts (10, 11). Therefore, we investigated this characteristic by determining either the MFCs or the killing curves. In general, our MFCs were within the reported ranges for all three echinocandins (11), with a rank order of activity of micafungin > caspofungin > anidulafungin.

Interestingly, a similar rank order was maintained when the “cidal” activity was investigated by killing experiments. In general, all three echinocandins exerted fungicidal activities against the susceptible isolate of C. glabrata. Our results are in agreement with those previously reported for caspofungin by Nagappan et al. (19). These authors assessed the in vitro activity of caspofungin against fluconazole-susceptible and -resistant isolates and observed that the drug was fungicidal at concentrations of 1 μg/ml and 4 μg/ml. A previous study reported that micafungin was fungicidal at concentrations ranging from 4 to 16 times the MIC₈₀ against C. glabrata isolates with MIC₈₀s ranging from 0.0039 to 0.25 μg/ml (9). Similar to this study, our data showed that micafungin exerted fungicidal activity starting from 1 to 32 times the MIC (0.06/2.0 μg/ml). In our hands, anidulafungin was fungicidal at concentrations of 32 times the MIC against the susceptible isolate, whereas previous data reported “cidal” activity starting from 4 times the MIC (16, 20). It can be hypothesized that this difference might be due to a slight modification of the experimental procedure (i.e., drug preparation, drug lot, subcultured volumes, “cidal” definition, etc).

It is known that echinocandins bind serum proteins at very high levels (i.e., >99% to human plasma proteins for anidulafungin and approximately 97% to albumin for caspofungin) (15, 21). Odabasi et al. (17) evaluated the effects of protein binding on the activities of caspofungin, anidulafungin, and micafungin against Candida and Aspergillus species. They observed that adding human serum sharply increased the MICs of micafungin and anidulafungin and modestly affected the MIC of caspofungin. However, they also found that the increase in MICs does not appear to be consistent with the rates of protein binding for the three compounds. Therefore, we performed in vitro studies by adding 50% human serum to RPMI 1640. Similar to what was observed by others (15, 17, 21), the addition of serum to the medium increased the MICs of all three drugs. We also performed the experiments by adding 50% fetal bovine serum to the medium, and we obtained similar results (data not shown).

In our hands, the ratios of geometric mean MICs (MIC values with/without serum) were 10.8, 8.0, and 31 for anidulafungin, caspofungin, and micafungin, respectively, while the ratios of geometric mean MFCs (MFCs values with/without serum) were 0.9 and 17 for caspofungin and micafungin, respectively (the ratio for anidulafungin was not determined because the tested concentrations were too low with respect to the fungicidal range).

In general, our results are in agreement with previous in vitro studies showing an increased echinocandin MIC when 50% serum or bovine serum albumin was added to RPMI 1640 (3, 12, 17).

Killing experiments conducted in the presence of serum showed that both caspofungin and micafungin started to be fungicidal at 4 times the MIC and that the addition of serum did not modify the fungicidal activity of caspofungin while it decreased that of micafungin. These results are in line with the higher serum binding levels of micafungin compared to caspofungin (21). Time-kill plots of anidulafungin in the presence of serum never reached the fungicidal effect, and additional studies should be performed by using various drug lots and by increasing the antifungal agent concentration.

Since in vitro/in vivo correlation is not yet understood, we compared the in vivo activities of all three echinocandins in a neutropenic murine model of candidiasis. In our hands, caspofungin proved to be the most active drug (in terms of either time or dose effectiveness) against this susceptible isolate. In fact, caspofungin started to be effective after only 2 days of treatment, while after 6 days, the lowest effective doses were 0.25, 1, and 5 mg/kg/day for caspofungin, micafungin, and anidulafungin, respectively.

Recently, Andes et al. (1, 2) investigated the in vivo activities of all three echinocandins against Candida spp., including C. glabrata, in a neutropenic murine model of disseminated candidiasis. To compare the potencies of antifungal agents, they calculated the 24-h static dose of each echinocandin and the doses required to achieve a 1-log-unit reduction in colony counts (1). They observed that caspofungin required less drug on a mg/kg basis for efficacy against all organisms than did the other two drugs. Actually, the mean static doses were 21.1, 2.47, and 0.33 mg/kg/24 h for anidulafungin, micafungin, and caspofungin, respectively, while mean doses to achieve 1-log-unit reduction were 39, 5.88, and 1.16 mg/kg/24 h for anidulafungin, micafungin, and caspofungin, respectively. In agreement with our results against C. glabrata, the echinocandins showed the following rank order of activity: caspofungin > micafungin > anidulafungin (1).

Our in vivo data on anidulafungin are similar to those observed by Gumbo et al. (14). These authors studied the activity of the drug in a neutropenic murine model of disseminated candidiasis due to a fluconazole-susceptible C. glabrata isolate. They found that doses of 8 and 10 mg/kg resulted in progressive declines in kidney fungal density, while data for mice that received 2 and 3 mg/kg did not differ significantly from the controls.

When therapeutic options are limited (i.e., azole resistance, renal insufficiency, drug intolerance, etc.), an important clinical question is whether an infection due to a yeast isolate with reduced susceptibility to a given echinocandin might be treated by an increased dose of the same drug or by selecting a new drug belonging to the same family. Recently, Brzankalski et al. (6) showed that caspofungin dose escalation may overcome the in vitro resistance of C. glabrata and be effective in vivo against resistant isolates. Additionally, the same authors suggested that aminocandin, an investigational echinocandin, has some potential in the treatment of C. glabrata infections due to caspofungin-susceptible isolates and that higher doses may be required against isolates with reduced susceptibility to caspofungin (6). Also, Garcia-Effron et al. (13) performed a genetic analysis of FKS1 and FKS2 genes from 13 echinocandin-resistant C. glabrata isolates. They demonstrated that FKS mutations influenced the β-1,3-d-glucan synthase kinetics and the FKS gene expression and that the mutations were linked to an echinocandin reduced-susceptibility phenotype. In the current study, we investigated the in vivo effects of the available echinocandins against two echinocandin-resistant C. glabrata isolates, one harboring the mutation Fks2p-P667T and the other the mutation Fks2p-D666G.

In our in vivo experiments, a fungicidal effect (i.e., organ sterilization) was never observed regardless of the drug or strain tested, the dosages, or the duration of therapy. In general, all three echinocandins at the active doses showed lower killing rates against the resistant strains than the susceptible strain.

Interestingly, we observed that micafungin retained its efficacy against both fks2 mutant strains, being effective at doses as low as 1 mg/kg/day. In C. albicans, Slater et al. (28) investigated the effects of three doses of micafungin (5 h, 29 h, and 53 h postinfection) in a murine model of disseminated candidiasis due to C. albicans fks1 heterozygous and homozygous mutants at Ser645. They observed that fungicidal activity in animals infected with an FKS1/fks1 heterozygote was reached only with doses as high as 20 mg/kg, while animals infected with the homozygous fks1 mutant failed to respond to any dosage.

In our study, we also demonstrated that caspofungin dose escalation may overcome in vitro resistance. In fact, caspofungin was still active at 5 or 10 mg/kg against the two resistant strains. Anidulafungin showed a trend toward reduction of CFU with respect to the controls at 5 and 10 mg/kg/day, but statistically significant differences were never reached. Our results are partially comparable to those reported by Wiederhold et al. (33). They compared caspofungin and anidulafungin in vitro and in vivo against two clinical isolates of C. glabrata with caspofungin MICs of ≥1 μg/ml and found that, despite enhanced in vitro potency of anidulafungin, treatment with the echinocandin did not result in reductions in tissue burdens greater than those achieved by treatment with caspofungin.

Recently, Wiederhold et al. (32) demonstrated that higher doses of caspofungin (5 and 10 mg/kg) did improve survival against an fks1p-S645P C. albicans-resistant isolate, but not against another isolate bearing the mutation fks1p-F641S. The authors concluded that the caspofungin effect against resistant C. albicans isolates may be associated with the virulence of the strain. Overall, these results suggest that there might be a linkage between the increased echinocandin MICs, the specific FKS mutations, and the potential for a successful clinical outcome.

In conclusion, we compared in vitro and in vivo effects of anidulafungin, caspofungin, and micafungin against the difficult-to-treat fungal opportunistic pathogen C. glabrata. While all three drugs were often fungicidal in vitro, they were not able to completely eradicate the infection in this murine neutropenic model of candidiasis. Caspofungin, followed by micafungin, was the most active drug at reducing the kidney burden of mice infected with an echinocandin-susceptible strain. Interestingly, micafungin showed the best in vivo antifungal activity against two resistant mutants of C. glabrata bearing specific mutations in the FKS2 hot spot region. A limitation of this in vivo study is that the mutant isolates showed a low level of resistance to echinocandins, and extrapolations to other mutants with more dominant mutations cannot be made. Further studies with a larger number of strains showing various levels of echinocandin resistance, as well as several schemes for therapies, should be done to better predict treatment success in clinical practice.