Current perspectives on echinocandin class drugs

Abstract

It has been nearly a decade since caspofungin was approved for clinical use as the first echinocandin class antifungal agent, followed by micafungin and anidulafungin. The echinocandin drugs target the fungal cell wall by inhibiting the synthesis of α-1,3-d-glucan, a critical cell wall component of many pathogenic fungi. They are fungicidal for Candida spp. and fungistatic for moulds, such as Aspergillus fumigatus, where they induce abnormal morphology and growth properties. The echinocandins have a limited antifungal spectrum but are highly active against most Candida spp., including azole-resistant strains and biofilms. As they target glucan synthase, an enzyme absent in mammalian cells, the echinocandins have a favorable safety profile. They show potent MIC and epidemiological cutoff values against susceptible Candida and Aspergillus isolates, and the frequency of resistance is low. When clinical breakthrough occurs, it is associated with high MIC values and mutations in Fks subunits of glucan synthase, which can reduce the sensitivity of the enzyme to the drug by several thousand-fold. Such strains were not adequately captured by an early clinical breakpoint for susceptibility prompting a revised lower value, which addresses the FKS resistance mechanism and new pharmacokinetic/pharmacodynamic studies. Elevated MIC values unlinked to therapeutic failure can occur and result from adaptive cell behavior, which is FKS-independent and involves the molecular chaperone Hsp90 and the calcineurin pathway. Mutations in FKS1 and/or FKS2 alter the kinetic properties of glucan synthase, which reduces the relative fitness of mutant strains causing them to be less pathogenic. The echinocandin drugs also modify the cell wall architecture exposing buried glucans, which in turn induce a variety of important host immune responses. Finally, the future for glucan synthase inhibitors looks bright with the development of new orally active compounds.

Echinocandin antifungal drugs

The echinocandin drugs anidulafungin (ANF), caspofungin (CSF) and micafungin (MCF) were the first members of the lipopeptide class of antifungal drugs. They are cyclic hexapeptides N-linked to a fatty acyl side chain (Figure 1) and are potent inhibitors of β-1,3-d-glucan synthase, which is responsible for biosynthesis of β-1,3-d-glucan, the major cell wall biopolymer [1]. These fungal-specific drugs show concentration-dependent antifungal activity against susceptible Candida spp. and Aspergillus spp. without cross-resistance to existing antifungal agents, which enables them to be effective against azole-resistant yeasts and moulds [2–4]. The echinocandins demonstrate fungicidal activity against most species of Candida [5,6] but they exhibit complex growth inhibitory behavior with moulds, such as Aspergillus where they induce a sea urchin-like multi-budded structure with slowed growth and lysis of rapidly growing bud tips (Figure 2) [7–9]. All three echinocandin drugs have been approved by the US FDA for the treatment of esophageal candidiasis and invasive candidiasis including candidemia. The echinocandins are now the preferred systemically active antifungal agents for the treatment of invasive candidiasis [10–12]. CSF is also approved for empirical therapy for presumed fungal infections in febrile neutropenic patients, while MCF is approved for the treatment of esophageal candidiasis and for prophylaxis of Candida infections in patients undergoing hematopoietic stem-cell transplantation during the period of neutropenia [13–19]. Echinocandin drugs are highly effective against azole-resistant strains of Candida [20–23] and Candida-forming biofilms [24–26], but they are less active against Zygomycetes, Cryptococcus neoformans or Fusarium spp. [3]. There is now a broad clinical experience using the echinocandins to treat both mucosal and invasive forms of candidiasis [13,14,27–37]. Overall, echinocandin drugs demonstrate a high therapeutic index with strong efficacy and excellent safety and tolerability profiles with few drug interactions and related adverse events [13,29,35,38–41]. Although reports of resistance to echinocandin treatment in Candida spp. are increasing, clinical failure remains relatively low, even in spite of the expanded use of these drugs [42–44].

Figure 1. The three approved echinocandin drugs.

Cyclic hexapetides are best differentiated by their aliphatic tails (circled).

Figure 2. Morphological change in Aspergillus fumigatus hyphae following exposure to caspofungin.

Cells were grown for 18 h in Roswell Park Memorial Institute (RPMI) medium in the (A) absence or (B) presence of caspofungin at 0.5 mg/ml and visualized by light microscopy (100×).

Glucan synthase complex

The β-1,3-d-glucan synthase is a multisubunit enzyme complex that catalyzes the transfer of sugar moieties from activated donor molecules to specific acceptor molecules forming glycosidic bonds in the reaction UDP-glucose + ([1,3]-β-d-glucosyl) (N) → UDP + ([1,3]- β-d-glucosyl) (N + 1) [45,46]. Most of our understanding of the genetics of glucan synthase has come from biochemical studies in yeast [47–53] and more recently with kinetic studies of FKS-resistant mutants from Candida spp. [42,43,54]. The enzyme complex has at least two subunits, Fks and Rho [55,56]. Fks is the presumed catalytic subunit [57] and is encoded by three related genes, FKS1, FKS2 and FKS3; it is the target of the echinocandin drugs. FKS1 is essential in Candida albicans [58,59] but in Saccharomyces cerevisiae and Candida glabrata, FKS1 genetic disruptants remain viable due to paralog FKS2. These genes are calcineurin dependent [60] and are downregulated by the immunosuppressive drug FK506. Rho, a GTP-binding protein in the Rho/Rac subfamily of Ras-like GTPases, regulates the activity of glucan synthase [61]. A variety of biochemical and genetic studies have pointed to the direct or indirect involvement of other proteins contributing to glucan synthase complex, although their firm association has not been established [62,63]. It is still not understood how the echinocandin drugs interact to inhibit glucan synthase. Enzymatic removal of the aliphatic tail of the drugs render them inactive [64]. Kinetic studies of drug transport suggest that both saturable high affinity and low affinity transport systems exist in yeast cells to move the echinocandin drugs into the cell in an energy-independent manner [65]. However, it is not clear whether transport into the cell is required for echinocandin action, which may explain why the drugs are not subject to the action of multidrug resistant transporters prominent in azole-resistant strains and biofilms [24,25,66]. It has been speculated that the tail of the drugs may intercalate into the bilayer resulting in inhibition. Such drug-target interactions would not require the drugs to enter the cell and they may act on the enzyme from the extracellular face of the cell membrane. Such behavior would be comparable to that of cardiac glycosides on the membrane-bound sodium pump [67]. Topology models for the location of resistance hot spots support such a model (Figure 3) [5,52]. The mechanistic nature of the interaction between echinocandins and glucan synthase remains ambiguous and will not be fully resolved until a 3D structure with bound drug is obtained.

Figure 3. Membrane topology model for glucan synthase showing putative resistance mutations and action of echinocandin drugs.

Transmembrane segments deduced from analysis of amino acid relative hydrophobicity.

FKS resistance mechanism

Echinocandin resistance resulting in clinical failure from susceptible species such as C. albicans, C. glabrata and Candida tropicalis is uncommon [44,68,69]. Yet, as the clinical use of echinocandins expands, Candida spp. isolates with reduced susceptibility to these drugs are increasingly encountered [42,43,68,70–80]. Recently, a collection of 490 C. glabrata bloodstream isolates from a population-based surveillance study identified 16 isolates (3.3%) with an elevated MIC value to one or more of CSF, MCF or ANF [68]. In another study of 133 Candida strains covering six species, 2.9% of isolates were considered resistant based on mechanistic criterion [69]. Fungi possess adaptive mechanisms that help protect against cellular stresses, such as those encountered following inhibition of glucan synthase by echinocandin drugs. These stress adaptation responses may result in elevated echinocandin MIC values [44,81], as determined in vitro, but they are not typically associated with clinical failures [82–84]. Rather, echinocandin resistance resulting in clinical failure is associated with mutations in two highly conserved ‘hot spot’ regions of FKS1 and/or FKS2 [42–44,68,85]. These limited-spectrum amino acid substitutions in the Fks subunits of glucan synthase confer cross-resistance among the class of drugs [44]. In C. albicans and most other Candida spp., mutations occur in hot spot regions of FKS1, which is the most prominently expressed FKS gene [69]. In C. glabrata, comparable mutations are found in either FKS1 or FKS2, which are both expressed and comprise the active subunit [42,71,74,79]. Typically, FKS3 is weakly expressed or not at all [42,71,74,79]. Mutations in FKS result in elevated MIC values (0.5–2 logs depending on the drug) and reduce the sensitivity of glucan synthase (inhibition constant producing 50% reduction in activity [IC₅₀] or K_i) to the drug by 30- to more than 3000-fold (Figure 4) [42,43,85]. The majority of echinocandin clinical breakthrough infections have been observed with C. albicans and C. glabrata, as expected by their prevalence in causing invasive disease. Amino acid changes at Ser645 (S645P, S645F and S645Y) are the most abundant and cause the most pronounced resistance phenotype (Figures 4 & 5) [42,43]. Other mutations (C. albicans: F641, L642, T643, L644, A645, L646, R647, D648) account for smaller increases [43], which is apparent from their affects on glucan synthase drug sensitivity (Figure 4). While the most prominent FKS mutations are associated with clinical resistance, not all FKS mutations confer the same phenotype and thus, they are less likely to result in breakthrough disease. On the basis of phenotypic behavior in MIC testing and animal models, as well as kinetic inhibition of glucan synthase, the allele order of resistance associated with the FKS genotype is: S645, F641>>L642, T643, L644, L646, R647, D648>P649. This limited spectrum of FKS mutations in Candida spp. cause reduced susceptibility across the entire class of echinocandin agents, which makes this mechanism of resistance ideal for molecular-based testing [86].

Figure 4. Effect of FKS mutations on sensitivity of glucan synthase to echinocandin drugs.

Glucan synthases with characteristic hot spot one amino acid substitutions were isolated and inhibition of activity (IC₅₀) by each of the echinocandin drugs was determined, as described by Park et al. [85].

Figure 5. Distribution and frequency of FKS mutants in Candida albicans and Candida glabrata.

Specific mutations in referred clinical isolates (n = 88) are shown for hot spot one mutations in FKS1 for C. albicans and both FKS1 and FKS2 for C. glabrata.

Clinically important nonsusceptible fungi

The recently described species Aspergillus lentulus demonstrates reduced pleiotropic susceptibility to most systemic antifungal drugs [87]. A. lentulus isolates are overall less susceptible to CSF although they retain susceptibility to ANF and MCF [88]. In this way, they resemble resistant mutants selected following cell wall digestion and exposure to CSF [89]. Analysis of the A. lentulus FKS1 sequence did not reveal a polymorphism at any of the known hotspot regions of the gene [90] indicating that this rare species [91] has an unknown resistance mechanism that is stable and confers clinical resistance [87]. One possibility is that this species produces excess glucans, which limit the effective concentration of antifungals. Increased resistance to ANF and other antifungal agents was observed for Candida biofilms overexpressing FKS1 [92]. It was surmized that excess glucans may act to sequester the hydrophobic drugs. In Cryptococcus neoformans, the echinocandins are active in vitro on glucan synthase yet fail to show highly active antifungal properties [93]. The mechanism of action is unclear, although in the case of C. neoformans, melanin may play a protective role [94]. In non-Aspergillus moulds, the echinocandins are weakly active as antifungal agents, although they are potent inhibitors of glucan synthase [95]. Inhibition of β-1,3 glucan formation may result in compensatory changes to the cell wall, such as other glucans or chitin being synthesized in these organism, but this has not been reported.

Paradoxical drug action

Growth of susceptible fungi at echinocandin concentrations much greater than the MIC is termed ‘the paradoxical effect’. The effect was first noted by Stevens and colleagues in a subgroup of C. albicans isolates grown in the presence of high CSF concentrations [96]. The prevalence of the paradoxical effect for CSF has been reported at 90, 60, 40 and 10% in Candida parapsilosis, C. albicans, C. tropicalis and Candida krusei isolates, respectively [97]. It is rarely seen with C. glabrata. These strains demonstrated a normal susceptibility pattern with a typical low MIC, but then paradoxically showed break-through growth at high drug concentrations. This drug-insensitive growth was conditional, as strains undergoing paradoxical growth showed normal susceptibility properties when cultured and re-challenged with drug. The mechanism responsible for paradoxical growth remains elusive. It is not related to modification of glucan synthase (FKS mutations) complex or to its upregulation in the presence of drug [98]. The growth behavior is consistent with adaptive stress responses observed in S. cerevisiae and C. albicans, which may lead to reduced susceptibility [99,100]. Overproduction of chitin as a compensatory cell wall component may account for some cases of paradoxical behavior [101,102]. Seemingly paradoxical behavior was reported for CSF in a murine model of pulmonary aspergillosis [103], although it was not reproduced in a systemic candidiasis mouse model with MCF utilizing strains displaying paradoxical behavior [104]. A paradoxical increase in circulating Aspergillus antigen was observed during treatment with CSF in a patient with pulmonary aspergillosis [105] and with MCF and CSF in animal models [106]. The clinical significance of the paradoxical effect remains unclear.

Hsp90 & echinocandin action

Fungal cells possess an array of mechanisms that enable them to adapt to stresses imposed by their environment or from the action of antifungal agents [44]. The molecular chaperone Hsp90 helps protect fungal cells following exposure to antifungal drugs by enabling the function of its client protein calcineurin [107–109]. This Ca²⁺/calmodulin-dependent serine threonine phosphatase helps regulate responses to membrane stress in yeasts and moulds [108,110]. Hsp90 is an essential chaperone responsible for folding and maturation of client proteins including signal transduction proteins with unstable conformations [111]. When stress conditions overwhelm the function of Hsp90, otherwise silent polymorphisms are expressed and can be selected if they provide a survival advantage [81]. Inhibition of glucan synthase following exposure to echinocandin drugs activates calcineurin-dependent stress responses and the downstream effector Crz1 [112,113]. The key cellular responses generated as a result of cell wall stress caused by echinocandins are influenced by Hsp90 to the extent that compromising Hsp90 function reduces echinocandin tolerance in wild-type (WT) strains and resistance in clinical isolates [81,109]. Chemical or genetic impairment of Hsp90 reduces the tolerance of Candida and Aspergillus to echinocandin drugs, as well as to other cell wall inhibitors such as nikkomycin Z [114].

MIC & breakpoints

The Clinical Laboratory Standards Institute (CLSI) has defined laboratory standards to determine MICs for WT strains and the MIC distributions from large-scale surveillance studies are used to define epidemiological cut-off values (ECVs). The European Committee on Antimicrobial Susceptibility Testing (EUCAST) work to the principle that breakpoints for susceptibility testing should not divide WT distributions for clinically important microorganisms, thereby ensuring that all individuals of a species that lack drug-specific mechanisms of resistance can be categorized as either susceptible, intermediate or resistant [115]. A WT organism is defined as a strain that does not possess acquired resistance to a given antimicrobial agent [116–118]; its MIC distribution covers three–four twofold dilution steps around the modal MIC [115]. Since the introduction of CSF in 2001, extensive surveillance has been conducted on the susceptibility of clinical strains to the echinocandin drugs [119–121]. Recently, a total of 8271 clinical isolates, obtained from blood or other normally sterile sites from more 100 medical centers worldwide during the period 2003–2007 were evaluated, including strains of C. albicans (4283), C. glabrata (1236), C. parapsilosis (1238), C. tropicalis (996), C. krusei (270), Candida lusitaniae (99), Candida guilliermondii (88) and Candida kefyr (61) [122,123]. WT MIC distributions demonstrate very low MICs typical of susceptible C. albicans, C. glabrata, C. tropicalis, C. krusei and C. kefyr and higher values typical of C. parapsilosis, C. guilliermondii and C. lusitaniae for all three echinocandins (Table 1).

Table 1.

Epidemiological cut-off values for anidulafungin, caspofungin and micafungin and eight species of Candida.

Species	Antifungal agent	No. of isolates tested	MIC (µg/ml)			Isolates with MIC of ≤2 µg/ml (%)
			Range	Mode	Epidemiological cut-off values (%)†
Candida albicans	Anidulafungin	4283	0.007–1	0.03	0.12 (99.7)	100.0
	Caspofungin	4283	0.007–0.5	0.03	0.12 (99.8)	100.0
	Micafungin	4283	0.007–0.5	0.015	0.03 (97.7)	100.0
Candida glabrata	Anidulafungin	1236	0.015–4	0.06	0.25 (99.4)	99.9
	Caspofungin	1236	0.015–8	0.03	0.12 (98.5)	99.8
	Micafungin	1236	0.007–2	0.015	0.03 (98.2)	100.0
Candida tropicalis	Anidulafungin	996	0.007–2	0.03	0.12 (98.9)	100.0
	Caspofungin	996	0.007–>8	0.03	0.12 (99.4)	99.9
	Micafungin	996	0.007–1	0.015	0.12 (99.1)	100.0
Candida kefyr	Anidulafungin	61	0.015–0.12	0.06	0.25 (100.0)	100.0
	Caspofungin	61	0.007–0.03	0.015	0.03 (100.0)	100.0
	Micafungin	61	0.015–0.06	0.06	0.12 (100.0)	100.0
Candida krusei	Anidulafungin	270	0.015–0.5	0.03	0.12 (99.3)	100.0
	Caspofungin	270	0.015–1	0.06	0.25 (96.3)	100.0
	Micafungin	270	0.015–0.25	0.06	0.12 (97.8)	100.0
Candida lusitaniae	Anidulafungin	99	0.06–1	0.5	2 (100)	100.0
	Caspofungin	99	0.03–1	0.25	0.5 (98.0)	100.0
	Micafungin	99	0.007–1	0.12	0.5 (99.0)	100.0
Candida parapsilosis	Anidulafungin	1238	0.015–4	2	4 (100.0)	93.1
	Caspofungin	1238	0.015–4	0.25	1 (98.6)	99.9
	Micafungin	1238	0.015–2	1	4 (100)	100.0
Candida guilliermondii	Anidulafungin	88	0.06–4	2	16 (100.0)	92.0
	Caspofungin	88	0.03–>8	0.5	4 (95.5)	95.5
	Micafungin	88	0.015–>8	0.5	4 (98.9)	98.9

^†Percentage of isolates for which MIC is less than or equal to the epidemiological cut-off values.

Adapted from [123].

Clinical breakpoints (CBPs) assist in determining whether an anti-infective is useful in the treatment of an infection and are typically set prior to a drug being used clinically. They require an integrated knowledge of WT distributions of MICs, an assessment of the pharmacokinetics/pharmacodynamics (PK/PD) and an evaluation of clinical outcome. Once established, breakpoints are reviewed when anti-infective agents have been in clinical use for some time, especially when new mechanisms of resistance are elucidated [117]. The CLSI Subcommittee for Antifungal Testing proposed a clinical interpretive MIC breakpoint (CBP) for echinocandin drugs against Candida spp. [124]. Conventionally, clinical breakpoints are used to indicate those isolates likely to respond to treatment with a given antimicrobial agent at the approved dosing regimen. It is different from the ECV, which provides the most sensitive measure of the emergence of strains with reduced susceptibility to a given agent. An initial CBP for susceptible isolates was established with MIC ≤2 mg/ml for all three echinocandin drugs and all species of Candida, which followed from an analysis of existing clinical, PK/PD and microbiological data [84]. A resistant breakpoint was not established because resistant isolates with known FKS genotypes were not included in the original database. The Committee recommended that isolates with a MIC more than 2 µg/ml be called non-susceptible. An initial CBP value of ≤2 µg/ml encompassed 99.9–100% of the five most susceptible Candida spp. (C. albicans, C. glabrata, C. tropicalis, C. krusei and C. kefyr) isolates, while the ECVs were eight- and 64-fold lower. The ECVs for the three less susceptible species, C. lusitaniae, C. parapsilosis and C. guilliermondii, were similar to the CBPs for all three of the echinocandins [123].

FKS-mediated resistance helps redefine breakpoints

It is now well established that clinically resistant Candida infections are associated with strains carrying mutations in FKS1 and/or FKS2 [42,43]. As many of these strains were obtained from patients failing therapy, the MIC values were not captured by the proposed susceptibility breakpoint, especially with MCF and ANF [42,43,80,115,123,125,126]. The CLSI breakpoint was inclusive of all Candida species, but it became apparent that a single breakpoint (≤2.0 µg/ml) for all species and echinocandin drugs did not adequately reflect the clinical experience with strains harboring FKS genotypes, as many of the strains were not captured by the susceptibility breakpoint, especially for MCF and ANF (Figure 6). Kinetic studies of the glucan synthase enzyme complex from FKS mutants also raised questions about the proposed CBP and suggested that a lower MIC cutoff of 0.25–0.5 mg/ml would be more inclusive of strains failing therapy (Figure 6) [42,43,127]. These findings coupled with better PK/PD indices for the echinocandins [128–130] provided a new basis to develop species-specific CBPs for both susceptibility and resistance, which would encompass data on strains with FKS genotypes [131]. Revised echinocandin breakpoints of ≤0.25 µg/ml/ml (susceptible), 0.5 µg/ml (intermediate) and ≥1 (resistant) for C. albicans, C. glabrata, C. tropicalis and C. krusei were recently proposed [131]. The revised CBP is applicable to all echinocandin drugs; although, a caveat was noted for MCF and C. glabrata in which a lower susceptible breakpoint of ≤0.06 µg/ml was proposed to provide adequate clinical utility. The CBP was not appropriate for either C. parapsilosis or C. guilliermondii, which are inherently less susceptible due to polymorphisms in the hot spot one region of FKS1 [54]. Unlike FKS mutants with strong resistance phenotypes (high MIC and IC₅₀), these strains typically respond well to standard therapy [132], presumably because the polymorphisms only weakly affect the sensitivity of glucan synthase (IC₅₀) for drug [54]. The moderate enzyme sensitivity maintains the observed clinical efficacy of echinocandin drugs with these strains [123]. However, they are shifted somewhat(1 log vs 3 logs) in susceptibility, which may account for sporadic reports of clinical failure [12,79]. Despite the good clinical efficacy, the current Infectious Diseases Society of America guidelines do not recommend the echinocandins as first-line therapy for C. parapsilosis fungemia [10]. To date, there are no clinical strains of C. parapsilosis or C. guilliermondii reported that contain characteristic hot spot mutations in FKS1 (apart from naturally occurring polymorphisms) that would increase an already high MIC and confer clinical failure due to reduced sensitivity of the enzyme for drug. A CBP for resistant C. parapsilosis and C. guilliermondii has not been determined, although isolates of these two species with MICs ≥8 µg/ml are likely to be resistant to the echinocandin drugs [123].

Figure 6. IC₅₀ as a function of MIC for echinocandin drugs.

A composite of Candida spp. in reference library (n = 55) containing WT and FKS mutations were plotted as a function of IC₅₀ and MIC for each of the echinocandin drugs, as shown. The original CLSI breakpoint of ≤2 µg/ml (solid line) fails to capture a significant number of FKS mutants resulting in treatment failure, especially for micafungin and anidulafungin. The new proposed breakpoint (broken line) of ≤0.25–0.5 µg/ml is more inclusive of all FKS mutants. The WT ECR represents Candida spp. (e.g., Candida parapsilosis) with inherent elevated MIC levels.

CLSI: Clinical Laboratory Standards Institute; ECR: Echinocandin resistance; WT: Wild-type.

Partially adapted with permission from [42,43].

Serum effect on drug potency & MIC testing

A wide range of animal and human studies indicate that echinocandin drugs are tightly bound to serum proteins CSF 80–96% [133,134], ANF more than 84% and MCF more than 99% protein bound [135]. The binding of serum to echinocandin drugs reduces their relative antifungal properties, up to 128-fold in some Candida spp. [135–137]. Comparable effects are also observed when evaluating the relative potency of the drugs directly on the glucan synthase target [135]. Relative potency differences for susceptible Candida and Aspergillus strains (20-fold or more) observed in standard MIC testing (e.g., CLSI M27-A3) are minimized in the presence of serum [135]. The effect of serum in neutralizing MIC differences was also consistent with a murine candidiasis model in which CSF and ANF were comparable in efficacy while MCF was only somewhat less effective over the dose range (0.05–1 mg/kg/day) in reducing kidney fungal burdens (ED99) [135]. Serum protein binding influences the antifungal properties of echinocandin drugs and standardized MIC values obtained in the absence of serum may only partially reflect in vivo efficacy. Recently, it was noted that the addition of 50% human serum to the MIC assay improves the identification of echinocandin-resistant Candida strains containing FKS hot spot mutations by broadening the separation between WT and FKS mutant-bearing resistant populations [138]. Despite this promising observation, which could aid in detection of bona fide resistance, this modification cannot be routinely applied to testing due to safety and standardization difficulties involving human serum. To circumvent this problem, studies have been initiated to evaluate commercial bovine serum albumin (BSA) as a standardized alternative to human serum. In one study, 79 Candida spp. strains from seven species and including 40 FKS hot spot mutants were evaluated. The addition of BSA was found to mimic that of human serum and helped distinguish between susceptible and FKS-resistant Candida strains [138]. A larger-scale study is in progress using CLSI and EUCAST protocols to evaluate the merits of adding BSA to obtain a clearer indication of resistant isolates from standardized susceptibility testing [139].

FKS mutations & virulence

The virulence of fungal cells is in part linked to their ability to rapidly divide and grow through cell extension. Glucan synthase is critical to growth and development as it must be able to produce glucan polymers necessary for cell wall deposition and remodeling with high efficiency. Like most critical cellular biosynthetic enzymes, it operates at a basal level that is displaced from maximum velocity (V_max). This property affords glucan synthase the ability to rapidly increase its catalytic capacity based on cellular demand. Detailed kinetic studies of glucan synthase isolated from drug-resistant cells with associated FKS hot spot mutations revealed that resistance has a cost to the enzyme. Many of the FKS mutations result in reduced V_max for the mutant glucan synthase enzymes [42,43], which limits the biosynthetic capacity of cells. Depending on the mutation and its affect on glucan synthase, the cells regulate the relative expression of FKS1 and FKS2 [42] but the overall effect is reduced biosynthesis. Some cells show thicker cell walls attributable in part to compensatory increases in cell-wall chitin content [140]. The FKS1 mutants with the highest chitin content had reduced growth rates in liquid medium and impaired capacity for yeast-to-hyphal transformation. They also show attenuated virulence in a Drosophila minihost model, as well as in a model of disseminated candidiasis in a competitive mixed (isogenic WT and FKS mutants) challenge infection [140]. Similar attenuated virulence associated with decreased V_max has been observed with C. glabrata [Perlin Ds, Unpublished Data]. Collectively, these results strongly suggest that FKS1 mutations conferring echinocandin resistance have a likely fitness cost resulting in reduced virulence because the cells cannot keep up with normal cell wall expansion associated with growth demand. Such a fitness cost would be expected to limit the epidemiological and clinical impact of these strains, which may help account for their low frequency in the clinic.

PK/PD considerations

Echinocandin drugs are only suitable for intravenous administration, as they have poor oral bioavailability and have half-lives ranging from 10–24 h. The tissue distributions are greatest in target organs for fungal disease namely liver, kidneys, large intestine, spleen and lungs. Low or trace levels of drug are found in the urine and cerebrospinal and vitreous fluids [141]. CSF is metabolized by hydrolysis and N-acetylation with loss in the feces and urine as degradation products [142]. ANF undergoes a slow chemical degradation at physiologic temperature and pH to an inactive compound, with approximately 30% excreted in the feces and <1% in urine [143]. MCF is metabolized to M-1 (catechol form) by arylsulfatase, with further metabolism to the M-2 (methoxy form) by catechol-o-methyltransferase [142]. Pharmacodynamic studies using in vivo candidiasis models have demonstrated that the 24 h area under the concentration-time curve (AUC)/MIC is a good indicator of the echinocandin exposure– response relationship. The 24 h AUC/MIC target for a stasis end point identified free-drug 24 h AUC/MIC against C. albicans for all echinocandin drugs [128]. ANF, CSF and MCF show linear PK values following intravenous administration with peak serum concentrations of 10 µg/ml and AUC values of 110 mgh/l [142]. The echinocandin drugs show a concentration-dependent killing and a prolonged (12–24 h) post-antifungal effect [129,130,144] and it suggests that escalating doses may have additional antifungal benefits. In an in vivo model, ANF redistributed from serum to tissue where it persisted for many days. Four days after a single drug dose of more than 4 mg/kg, the measured drug concentrations in the tissue were in excess of the MIC₉₀ for clinical isolates of Candida, which is similar to the tissue concentration of CSF. Murine studies demonstrate that fungicidal activity of the drugs is maximized against C. albicans when the ratio of total serum drug concentration to organism MIC (C_max/MIC) approaches 10 or if the 24 h AUC:MIC ratio exceeds 10–20 [128–130]. Caspofungin also shows concentration-dependent PD in the treatment of invasive pulmonary aspergillosis with the C_max:minimum effective concentration (MEC) ratio most closely associated with reduction of pulmonary fungal burden [103].

Drug–bug–immune responses

The observation that the addition of low levels of CSF to monocyte cultures increased anti-fungal activity was noted early in echinocandin research [145]. Dectin-1 is a pattern recognition receptor for β-1,3/β-1,6-linked glucans; it is a type II transmembrane protein belonging to the NK-like C-type lectin-like receptor family [146,147] and its activation results in phagocytic, proinf lammatory and antimicrobial responses [148–150]. At sub-MIC concentrations, CSF was found to increase β-glucan exposure on C. albicans through cell wall remodeling which triggered enhanced macrophage inflammatory responses via interaction with dectin-1 [151]. In A. fumigatus, the conidial surface is composed of a hydrophobic layer of RodA protein covalently bound to the conidial cell wall. This layer insulates the spores making them immunologically inert. The removal or disruption of this surface layer results in conidial morphotypes inducing immune activation [152]. As the live conidia swell, they display β-glucans on their cell surface and with hyphal formation, they trigger dectin-1 recruitment to phagosomal membranes of alveolar macrophages [153,154]. Exposure of macrophages to germinating conidia reveals a direct correlation between surface β-glucan display and inflammatory cytokine/chemokine induction. The innate immune system ignores the less threatening dormant conidia, but responds vigorously to germinating condia and hyphae [155]. Echinocandins do not completely inhibit in vitro growth of A. fumigatus, rather they induce morphological changes in fungal hyphae (Figure 2) [7]. Echinocandin-treated hyphae enhance antifungal properties in a macrophage system [145]. Echinocandin-induced morphological changes in A. fumigatus hyphae and other fungi are accompanied by increased β-glucan exposure resulting in dectin-1-mediated inflammatory responses caused by macrophages [156,157]. Echinocandins stimulate release of TNF and CXCL2, and TLR 9 is also upregulated [156]. Overall, the echinocandins compensate for their relatively weak growth inhibitory properties in Aspergillus and other molds by unmasking glucans, which can then prime and stimulate components of the innate immune system, even when compromised. Finally, it has been observed that biofilms are significantly more susceptible to the presence of human phagocytes in the presence of an echinocandin than in its absence suggesting a combined action of the drug and host defense [158].

Oral glucan synthase inhibitors

The echinocandins are available only in intravenous formulation, which limits their use in less severe infections or as oral step-down agents. The enfumafungins are natural products that are structurally different from the echinocandins, but like the echinocandins, they inhibit fungal β-1,3-d-glucan synthesis. Enfumafungin is a potent antifungal compound isolated from Hormonema carpetanum [159,160]. It is a natural triterpene glycoside that inhibits glucan synthase and shows highly potent antifungal activity in vitro against Candida and Aspergillus, although its chemical structure is distinct from the echinocandins. Recently, Merck described a semi-synthetic derivative of enfumafungin, MK-3118, which is being developed as an oral therapy for fungal infections [161,162]. MK-3118 is a potent inhibitor of glucan synthases from C. albicans (IC₅₀ 0.6 ng/ml) and A. fumigatus (IC₅₀ 1.7 ng/ml). Against 160 strains of seven Candida species, it had a MIC₉₀ of no more than 1 µg/ml and against 40 Aspergillus spp., it had a MEC₉₀ of no more than 0.015 µg/ml [163]. MK-3118 also shows promising in vivo efficacy in murine models of candidiasis [164] and aspergillosis [165]. MK-3118 was also active in vitro against echinocandin-resistant Candida spp. isolates containing FKS mutations. This latter observation suggests that this molecule either binds at a site separate from the echinocandins on Fks1 or it is active on another component of glucan synthase. More detailed biochemical and genetic studies will be required to assess this molecular property. The spontaneous mutation frequency in C. albicans for conferring resistance to MK-3118 was less than 4.6 × 10⁻⁹ mutations per cell per generation, which is comparable to that of CSF at 2.6 × 10⁻⁹ [163]. This level of mutant generation portends a low frequency of resistance, except at high burdens. But this remains to be determined experimentally.

Future perspective

The clinical experience with echinocandin drugs has been high successful, as the class exhibits potent efficacy, especially with Candida spp., with minimal side effects and a low frequency of resistance. A revised break-point has been proposed to better guide treatment when resistant isolates are encountered. As clinical resistance is associated with a narrow spectrum of point mutations in the FKS genes, real-time molecular diagnostics may be of value in diagnosing resistant strains. However, not all FKS mutations have the same potential for therapeutic failure and a rigorous PK/PD approach is needed to better understand weak FKS mutants. It remains to be seen whether commonly encountered strains with reduced susceptibility but which respond to therapy, such as C. parapsilosis will begin to show higher-order breakthrough infections. The effect of echinocandin drugs in stimulating the immune system by exposing critical antigens is an important consideration in the use of these drugs with moulds, and this drug–bug–immune response needs to be better understood. Finally, there is much excitement as a new generation of oral glucan synthase inhibitors is being developed with great promise to benefit patients requiring potent step-down therapy. Hopefully, there will also be an emphasis on broader spectrum inhibitors to meet the challenges of other fungal pathogens.

Executive summary.

Echinocandin antifungal drugs

• ▪The echinocandin drugs anidulafungin, CSF and MCF are the first members of the lipopeptide class of antifungal drugs that target the fungal cell wall by inhibition of β-1,3-d-glucan synthase.
• ▪These drugs show concentration-dependent antifungal activity against susceptible Candida spp. and Aspergillus spp. and are effective against azole-resistant yeasts and moulds; they are also active on most biofilms.

Glucan synthase complex

• ▪The enzyme complex in Candida spp. consists of a catalytic subunit that is the target of echinocandin drugs and is encoded by three related genes, FKS1, FKS2 and FKS3; it also contains a regulatory component Rho1.

FKS resistance mechanism

• ▪Echinocandin resistance resulting in clinical failure from susceptible Candida species is largely uncommon, although population-based surveillance studies suggest the level of in vitro resistance in some species is approximately 3%.
• ▪Echinocandin resistance resulting in clinical failure is associated with mutations in two highly conserved ‘hot spot’ regions of FKS1 and/or FKS2. These mutations confer cross-resistance among the class of drugs and result in several log order decreased sensitivity at the level of glucan synthase.
• ▪In Candida albicans and most other Candida spp., mutations occur in hot spot regions of FKS1, which is the most prominently expressed FKS gene. In C. glabrata, FKS1 and FKS2 are major sources of resistance.

Clinically important nonsusceptible fungi

• ▪These fungi typically have glucan synthase enzymes that are sensitive to drug but fail to show significant antifungal susceptibility.
• ▪The resistance mechanism is largely unknown but may involve compensatory changes in cell-wall composition or presence of excess glucan that acts as a drug sink.

MIC & breakpoints

• ▪An initial clinical break point for susceptible isolates was established for all three echinocandin drugs and all species of Candida, which followed from an analysis of existing clinical, pharmacokinetics/pharmacodynamics (PK/PD), epidemiological and microbiologic data.

FKS-mediated resistance helps redefine breakpoints

• ▪Clinical isolates containing characteristic FKS hot spot mutations resulting in clinical failures were not captured by the initial susceptibility cut-off value. A new echinocandin breakpoint was proposed based on studies of PK/PD, enzyme inhibition kinetics and epidemiologic MIC data to better account for these isolates.

FKS mutations & virulence

• ▪FKS mutations conferring elevated MICs are associated with reduced virulence due to alterations in the catalytic potential of mutant glucan synthase and the net result on cellular physiology and morphology.

PK/PD considerations

• ▪Echinocandin drugs show concentration-dependent killing and a prolonged post-antifungal effect.
• ▪Pharmacodynamic studies demonstrate that the 24-h area under the concentration-time curve (AUC)/MIC best describes the echinocandin exposure-response relationship.

Drug–bug–immune responses

• ▪Exposure of Candida and Aspergillus to echinocandins alters the cell wall composition resulting in surface exposure of glucans that stimulate host immune responses.

Oral glucan synthase inhibitors

• ▪Enfumafungins are potent inhibitors of glucan synthase that are chemically distinct from the echinocandins and are being developed as a next generation oral antifungal class.