Introduction

Candida glabrata accounts for approximately 15% of systemic infections related to the Candida genus. This species represents a serious clinical problem, as infection is associated with high rates of mortality [1, 2]. Since this species has few virulence factors, the large number of infections by C. glabrata is result from its inherently low susceptibility to azole antifungals, leading to development of resistance [3, 4].

Currently, echinocandins have been recommended as the first-line treatment by the Infectious Diseases Society of America (IDSA) for C. glabrata infections, especially in patients with previous exposure to azoles. In addition to a broad-spectrum activity against Candida species, echinocandins also have a favorable safety profile [5]. However, evidence of therapeutic failure in treatment with echinocandins has been reported. Reduced susceptibility of C. glabrata to caspofungin during prolonged therapy has also been reported [6]. Furthermore, a decrease in in vitro and in vivo susceptibility or resistance of C. glabrata to other echinocandins has been shown in other studies [7,8,9,10,11].

In order to study the profile of C. glabrata resistance to echinocandins, we evaluated the in vitro susceptibility of fluconazole/caspofungin-sensitive and -resistant C. glabrata clinical isolates against antifungal agents. Furthermore, we evaluated in vitro associations between antifungals that could demonstrate a synergistic action against resistant isolates.

Materials and Methods

Microorganisms

A total of 20 clinical isolates of C. glabrata obtained from oral lesions of patients with HIV infection were studied. The isolates were identified by the API®/ID32 yeast identification systems (bioMérieux, Marcy I’ Etoile, France) supplemented with conventional methods as needed. Four groups were formed from the first wild-type group: C. glabrata wild-type (Cg-Wt), C. glabrata fluconazole-resistant (Cg-FR), C. glabrata caspofungin-resistant (Cg-CR), and C. glabrata fluconazole/caspofungin-resistant (Cg-FCR). The groups Cg-FR, Cg-CR and Cg-FCR were obtained by exposure to sublethal concentrations of fluconazole and/or caspofungin by using the Fekéte-Forgács et al. [12] method. The isolates were stored in saline solutions until they were used. The reference strains C. albicans ATCC 14053, C. glabrata ATCC 90030, C. parapsilosis ATCC 22019 and C. krusei ATCC 6258 were obtained from the American Type Culture Collection (ATCC, Manassas, VA). They were used as quality controls for tests.

Antifungal Agents

Standard posaconazole, voriconazole, fluconazole, and amphotericin B powders were obtained from Sigma-Aldrich, São Paulo, SP, Brazil. Stock solutions of posaconazole, voriconazole, and amphotericin B were prepared in dimethyl sulfoxide (DMSO). Stock solution of fluconazole was prepared in sterile distilled water. Anidulafungin (Pharmacia & Upjohn Co. Kalamazoo, MI, USA), caspofungin (Laboratoires Merck Sharp & Dohme-Chibret, Clermont Ferrand, France), and micafungin (Astellas Pharma Tech Co., Takaoka, Toyama, Japan) were obtained from the manufacturer and prepared in sterile distilled water. The solutions were sealed and frozen at -70 °C until they were used. Final dilutions were made in RPMI 1640 medium (Sigma-Aldrich, São Paulo, SP, Brazil).

Antifungal Susceptibility Testing

All the isolates were tested for in vitro susceptibility to fluconazole, voriconazole, posaconazole, amphotericin B, caspofungin, micafungin, and anidulafungin according to the CLSI broth microdilution method (M27-A3) [13]. The inoculum density ranged from 0.5 × 103 to 2.5 × 103 cells mL−1. Minimum inhibitory concentrations (MICs) for the agents were read after incubation for 24 or 48 h. The MIC results for each agent were determined visually as specified in the CLSI documents M27-A3 [7] and M27-S4 [14]. The isolates were classified as sensitive (S), intermediate (I), or resistant (R) based on the CLSI interpretative breakpoints M27-A3 [13] and M27-S4 [14] and the definition by Pfaller et al. [15] and Pfaller et al. [16]. For anidulafungin and caspofungin MIC values ≥0.5, 0.25 and ≤0.12 μg mL−1 were considered to indicate R, I, and S, respectively. For micafungin MIC values ≥0.25, 0.12, ≤0.06 μg mL−1, were considered to indicate R, I, and S, respectively. For fluconazole MIC results of 32.00 and ≥64.00 μg mL−1 were categorized as sensitive dose-dependent and resistant, respectively. For voriconazole and posaconazole, MIC results ≥4.00, 2.00, and ≤1.00 μg mL−1 were categorized as R, I, and S, respectively.

Antifungal Interactions Assays

Interactions between antifungal agents were assessed by checkerboard assays. A modified version of the microdilution methodology from the CLSI document M27-A3 [13] was used. The FIC of each antifungal agent was read visually as the lowest drug concentration that resulted in ≥50% inhibition of the growth as compared to the positive control in azole-echinocandin association and 100% of inhibition in amphotericin B-echinocandin association. The reading was performed after 24 h of incubation at 37 °C for all antifungals. The fractional inhibitory concentration index (FICI) was obtained by summation of the combined effects of antifungal FICs; the obtained FICI values were interpreted as follows: FICI ≤ 0.5, synergistic; 0.5 < FICI < 4, indifferent; FICI ≥ 4, antagonistic [17].

Statistical Analysis

The in vitro susceptibility data were compared by analysis of variance (ANOVA) followed by Tukey’s post hoc test for multiple group comparisons and Student’s t test followed by the Mann–Whitney U test for independent samples. All tests were performed using the GraphPad Prism Program (version 5.0, GraphPad Software, San Diego, CA). A P value of <0.05 was assumed for the statistical significant differences.

Results

Table 1 summarizes the in vitro susceptibilities of all groups of isolates to azoles, amphotericin B and echinocandins. All antifungal agents exhibited good activity against the Cg-Wt group. Voriconazole and posaconazole MICs increased significantly (P < 0.0001) against the Cg-FR and Cg-CFR groups compared to those against Cg-Wt; 5 Cg-FR isolates were resistant to posaconazole, and 10 Cg-FR isolates were resistant to voriconazole. This observation demonstrated the expected cross-resistance among azole antifungals (Table 2). The Cg-CR group did not show resistance to azole antifungals. Amphotericin B MICs were higher in the Cg-FR and Cg-CFR groups compared to Cg-Wt (Fig. 1).

Table 1 Geometric mean MIC, MIC range, MIC50, and MIC90 (µg mL−1) of fluconazole, voriconazole, posaconazole, amphotericin B, caspofungin, anidulafungin, and micafungin for Candida glabrata isolates
Full size table
Table 2 Frequency of antifungal resistance among fluconazole-resistant, caspofungin-resistant, and fluconazole/caspofungin-resistant Candida glabrata isolates (n = 20)
Full size table
Fig. 1
figure 1

Minimal Inhibitory Concentrations (MIC) means of azole antifungals, echinocandins, and amphotericin B against Candida glabrata groups (Cg-FR, Candida glabrata fluconazole-resistant; Cg-CR, Candida glabrata caspofungin-resistant; Cg-FCR, Candida glabrata fluconazole/caspofungin-resistant

Full size image

The three echinocandins had good active against the Cg-Wt and Cg-FR groups. The MIC at which 90% (MIC90) of the both groups were inhibited was 0.125 µg mL−1 for caspofungin, 0.030 µg mL−1 for anidulafungin and 0.008 µg mL−1 for micafungin (Table 1).

Despite the good activity of echinocandins against the wild-group and fluconazole-resistant group, when they were tested against the Cg-CR there was a significant increase in anidulafungin and micafungin averages (P = 0.0389) (Fig. 1). Five strains showed intermediate MIC values and three strains were resistant to anidulafungin. Three strains were resistant to micafungin. In Cg-CFR strains, the anidulafungin MICs remained high while the micafungin MICs did not differ significantly between the wild and fluconazole-resistant groups; however, a resistance profile for any of the echinocandins (anidulafungin and micafungin) (Table 2; Fig. 1) was not observed.

The results of the in vitro checkerboard analysis of the combined antifungal effects against the Cg-CFR isolates are summarized in Table 3 and Fig. 2. The association of caspofungin and posaconazole led to synergism in 85% (FIC indexes mean = 0.271) of the isolates, indifference in 15%, and lack of antagonism. The association caspofungin plus voriconazole was synergistic to 35% (FIC indexes mean = 0.811) and indifferent to 65% of the isolates and caspofungin plus fluconazole showed 85% (FIC indexes mean = 1.905) of indifference and no synergism. The association micafungin plus posaconazole was synergistic to 30% and indifferent to 70% of isolates (FIC indexes mean = 1.165). When micafungin was associated with voriconazole, the observation was synergism in 15%, indifference in 60%, and antagonism in 25% of the isolates (FIC indexes mean = 3.463). Micafungin plus fluconazole was synergistic to 30% and indifferent to 55% of the isolates (FIC indexes mean = 2.288). In the associations of anidulafungin plus posaconazole and anidulafungin plus voriconazole, synergism was observed in 70% of the isolates with FIC indexes mean of 0.463, and 0.627, respectively; in these associations, no antagonism was observed. When anidulafungin was associated with fluconazole the same percentage (45%) of synergism and indifference was observed (FIC indexes mean = 1.633). In the associations of echinocandins with amphotericin B, synergism was observed in 20% (FIC indexes mean = 1.276), 15% (FIC indexes mean = 4.263), and 35% (FIC indexes mean = 1.200) of the isolates for caspofungin, micafungin and anidulafungin associations, respectively. Micafungin plus amphotericin B was antagonistic to 50% of the isolates.

Table 3 Interactions (%) of echinocandins with fluconazole, posaconazole, voriconazole and amphotericin B against Candida glabrata fluconazole/caspofungin resistant (Cg-FCR) (n = 20)
Full size table
Fig. 2
figure 2

Results obtained with antifungal combinations against Candida glabrata fluconazole/caspofungin resistant (Cg-FCR). AMB: amphotericin B, ANF: anidulafungin, CAS: caspofungin, MFG: micafungin, FLZ: fluconazole, PCZ: posaconazole, VRC: voriconazole

Full size image

Discussion

According to the current international guidelines, prophylactic therapy with azole antifungals is a risk factor for the development of reduced susceptibility or even resistance in Candida spp. isolates. On the other hand, pre-exposure to echinocandins has not been considered a risk so far [4]. However, this situation might be changing. A survey carried out in France reports that recent exposure to caspofungin influences the epidemiology of candidemia. In this study caspofungin MICs increased significantly for five major Candida species including C. glabrata [18].

In newer studies, it has been found that echinocandin resistance is most common in C. glabrata compared to that in other species, with rates exceeding 10% at selected institutions [19, 20]. This rate can be attributed to the high potential of C. glabrata for developing resistance mutations [21]. It has been reported that the reduced susceptibility to echinocandins is due to mutations in regions of the FKS1 and FKS2 genes; these genes encode subunits of the glucan synthase enzyme complex [22,23,24,25]. Echinocandin resistance in C. glabrata is related to pre-exposure to echinocandins and/or FKS mutant isolates [26].

Alexander et al. [11] analyzed 313 C. glabrata strains from bloodstream infections isolated at Duke Hospital, Iowa, USA. They found that between 2001 and 2010, echinocandin resistance increased from 4.9 to 12.3%, while fluconazole resistance increased from 18 to 30%. Among the isolates resistant to fluconazole, 14.1% were also resistant to one or more echinocandins. Further, 8% of patients infected with FKS mutant strains showed intermediate or resistant MICs and failure to treatment with any of the echinocandins.

Here we demonstrated cross-resistance between azole antifungals. On the other hand, all of them showed good activity against the caspofungin-resistant strains, in the same way as the echinocandins exhibited good activity against the Cg-FR strains. Multidrug-resistance between azoles and echinocandins is uncommon because they have different mechanisms of action; nevertheless Pfaller et al. [10] showed that 9.3, 9.3, and 8% of 162 fluconazole-resistant C. glabrata isolates were resistant to anidulafungin, caspofungin, and micafungin, respectively. This relationship highlights a possible co-resistance between these antifungal classes, which was not observed here.

We found an increase in anidulafungin and micafungin MICs against Cg-CR isolates. Thus constant exposure of wild-type C. glabrata strains to caspofungin, besides developing resistance to caspofungin, can also develop cross-resistance between other echinocandins. In Cg-CFR isolates, anidulafungin also showed higher MICs, while the average for micafungin significantly remained the same for the Cg-FR and Cg-Wt strains. Resistance profiles were not observed for any of the echinocandins. This difference indicates that micafungin can have a better effect than anidulafungin in fluconazole/caspofungin-resistant isolates. In contrast to our findings, a study performed by Farmakiotis et al. [27] showed that C. glabrata isolates from cancer patients that were resistant to fluconazole and caspofungin were also resistant to anidulafungin and micafungin.

Concerning the combinations of echinocandins with other antifungal agents against Cg-CFR, it has been shown that the highest percentages of synergistic outcomes were in the associations of posaconazole with caspofungin and anidulafungin as well as the association of voriconazole with anidulafungin.

The combination therapy of posaconazole with echinocandins is probably advantageous against fluconazole/caspofungin-resistant C. glabrata, since synergism was seen in up to 50% of the isolates. The synergism was less (30%) in the combination with micafungin, but antagonism was absent. Moreover, posaconazole appears to reduce the resistance of C. glabrata to caspofungin. Oliveira et al. [28] has already reported synergy of posaconazole and caspofungin against 18% of C. glabrata sensitive isolates (n = 199) and 4% of fluconazole resistant isolates, without evidence of antagonism.

Furthermore, the combination posaconazole plus caspofungin also shows in vivo synergy against C. albicans echinocandin-resistant isolates suggesting potential therapeutic usage for these combinations [29]. In addition, co-administration of posaconazole with echinocandins was seen to be well tolerated without any effect on either agent’s pharmacokinetics [30].

The associations of caspofungin with voriconazole, fluconazole and amphotericin B showed indifferent results. Our results were according to those of Barchiesi et al. [31], who reported indifferent results about the effects of caspofungin combined with amphotericin B against C. glabrata. The combination of caspofungin with fluconazole has no synergy; this outcome shows that the combined use of these antifungal agents is not recommended when resistance to them is established. Although the combination of caspofungin and voriconazole has not shown high percentages of synergism against Cg-CFR, this combination has been considered the preferred therapy for subsets of organ transplant recipients with invasive aspergillosis [32]. This combination deserves further studies to clarify its activity against C. glabrata.

Anidulafungin was seen to be a good option to combine with azole antifungals. In a study conducted by Karlowsky et al. [33], an additive effect was observed when anidulafungin was combined with fluconazole against C. glabrata. Combinations studies of anidulafungin with voriconazole have been most exploited against Aspergillus; they showed partial synergy and a lack of antagonism [34, 35]. According to our knowledge, this is the first study to evaluate the combination of anidulafungin and posaconazole against multi-resistant C. glabrata isolates. The high synergism found leads us to believe that this combination is a promising option for the treatment of infections caused by resistant C. glabrata. We also understand that further studies are needed to confirm our findings.

The association of micafungin with azoles and amphotericin B did not show significant synergism percentages; however, other authors have found positive results when micafungin was associated with voriconazole or amphotericin B. Baltch et al. [36] described that in human macrophages infected by C. glabrata, the combination of micafungin and voriconazole was more effective than single drugs. Olson et al. [37] showed that the combination of amphotericin B with micafungin markedly improved the therapeutic outcome in murine C. glabrata systemic infection. In addition, co-administration of micafungin with amphotericin B does not affect the pharmacokinetic action of micafungin; therefore, this combination is a safe option for treatment [38].

Considering the substantial benefit of combination therapy, combinations of echinocandins and azole antifungals are promising in the treatment of infections caused by multi-resistant C. glabrata. However, the interactions between amphotericin B and echinocandins did not show any advantage as compared to the use of amphotericin B alone. Since the results presented here are obtained in vitro, and checkerboard assay is a preliminary study of antimicrobial interactions, in vivo studies are needed to validate the safe use of these combinations.

As well as, more studies are needed to demonstrate the cross-resistance of C. glabrata to echinocandins. This is just an in vitro demonstration that pre-exposure to caspofungin may result in resistance to caspofungin, and cross-resistance may occur between other echinocandins; such a cross-resistance has already happened with the azole antifungals.