Microdilution-broth based antifungal susceptibility testing (AFST)

Predicting therapeutic outcomes as well as guiding the antifungal drug discovery process based on AFST of pathogenic fungi remains challenging. In a recent review, Sanguinetti and Posteraro presented an overview of standard AFST methods, focusing on their advantages and disadvantages, as well as of new promising technologies and newer-generation methods (e.g. whole genome sequencing) that can predict resistance 13.

The Clinical and Laboratory Standards Institute (CLSI; formerly the National Committee for Clinical Laboratory Standards [NCCLS]) and The European Committee on Antimicrobial Susceptibility Testing (EUCAST) have developed reproducible methods for testing the activity of antifungal agents against yeasts (the CLSI M27, M44, M60 and the EUCAST E.Def 7.3 documents) and filamentous fungi (molds; the CLSI M38, M44, M51, M61 and EUCAST E.Def 9.3 documents) 17,18,19,20,21,22,23. These reference AFST methods, or their commercial counterparts such as Sensititre YeastOne (SYO, Thermo Fisher Scientific, MA, USA) rely on measuring growth of a defined fungal inoculum in a specific growth broth in the presence of different concentrations of the antifungal drug and allow the determination of the MIC (the minimum inhibitory concentration) of the drug resulting in complete or prominent growth inhibition. Clinical breakpoints have been determined by CLSI for anidulafungin, caspofungin, micafungin, fluconazole, and voriconazole against the prevalent Candida spp. whereas the EUCAST has set breakpoints for amphotericin B, anidulafungin, micafungin, fluconazole, itraconazole and voriconazole against the common Candida spp. and for amphotericin B, itraconazole, isavuconazole, posaconazole and voriconazole against the most common Aspergillus species 23,24. These AFST methods deliberately minimize measurement of tolerance or trailing growth (see further) because it is highly variable under different culture conditions. For drug-organism combinations for which clinical breakpoints are not available, epidemiological cutoff values are suggested based on normal ranges of susceptibility of wild-type populations 16. These should encompass the range of normal strain to strain variation within a species but exclude those organisms with known resistance mechanisms. Note that in some cases, published proposed epidemiologic cut-off values (see below) do not seem realistic. For example, for C. auris and fluconazole, they are so high that they fail their assignment as resistant. This has been recently addressed by the suggestion of more realistic, lower fluconazole breakpoints 25. The occurrence of resistance is often associated with a genetic difference between a susceptible and resistant isolate. Resistance may be also the result of transient and reversible adaptation 26,27.

Commonly used MIC end-point terminologies are: The minimum inhibitory concentration (MIC) is the lowest concentration of an antimicrobial agent that prevents or inhibits the visible growth of fungal cells, as established by a standardized endpoint. Standardized MIC endpoints include partial inhibition of growth relative to the growth observed in the control (>50% inhibition of growth as determined visually [CLSI] or spectrophotometrically [EUCAST] for azoles and echinocandins), complete inhibition (100%) of visual growth (CLSI) or >90% by spectrophotometer (EUCAST) as applied to amphotericin B 28. NOTE that MIC₅₀, MIC₉₀ and MIC₁₀₀ are NOT appropriate in this context. The MIC50 is the concentration of an antimicrobial agent at which 50 percent of the strains of an organism tested are inhibited. The MIC90 is the concentration of an antimicrobial agent at which 90 percent of the strains of an organism tested are inhibited. The term minimum effective concentration (MEC) is used to describe the effect of echinocandin agents on filamentous fungi and is defined as the lowest concentration that results in conspicuously aberrant growth as assessed microscopically. Aberrant growth of hyphae is defined as small, round, compact microcolonies, often with swollen ends to the hyphae, compared with the matt of hyphal growth in the control well that does not contain an antifungal agent.

Interpretation of the efficacy of a given antifungal drug, is determined by the use of clinical breakpoints (CBPs). The CLSI uses the term ‘breakpoint’ as ‘clinical breakpoint’ is redundant in light of the fact that breakpoints are only applicable under clinical conditions. Thus the CBPs for in vitro susceptibility testing are used to indicate those isolates that are likely to respond to treatment with a given antimicrobial agent administered using the approved dosing regimen for that agent 28. CLSI and EUCAST have established species-specific CBPs for some of the systemically active antifungal agents [CLSI 18, M60 Ed1; EUCAST E.Def 7.3 and E.Def 9.3]. The CBPs also provide information on the sensitivity of the CLSI/EUCAST methods to detect emerging resistance associated with acquired or mutational resistance mechanisms. The CBPs sort isolates into interpretive categories of susceptible (S), susceptible dose dependent (SDD; i.e. susceptibility is dose-dependent), intermediate (I), and resistant (R). The SDD category encompasses those organisms with MICs in a range that may respond to systemic therapy providing the drug levels in the blood are sufficiently high, especially relevant for fluconazole and voriconazole. For fluconazole and Candida glabrata, SDD is representing MICs < 32 µg/ml.. CBPs are established by taking into account microbiological (MICs and ECVs), clinical, molecular mechanisms of resistance, biochemical, pharmacokinetic and pharmacodynamic (PK/PD) data and provide the best cutoff value to predict clinical outcome for the treatment of a specific organism and antifungal agent 28.

Epidemiologic cutoff values (ECV/ECOFF) have been established to aid in the interpretation of MIC results when the lack of clinical data precludes the establishment of CBPs (CLSI, M57 and M59 29,30; EUCAST). As mentioned above, they are also the first for proposing BPs. It has been suggested that some CBPs may not detect known mutational resistance in different species. Because of that, an extensive effort has been undertaken to establish ECVs or the MICs/MECs that separate wild-type (WT) from non-WT strains; the latter are more likely to harbor acquired and mutational resistance mechanisms. ECVs as BP are species-and-method-specific 30 and are also useful for tracking the emergence of strains with decreased susceptibility to a given agent and therefore less likely to respond to therapy 30; ECVs can be used for the same purpose in surveillance studies 16. Epidemiologic cutoff values can also be used to identify isolates that are less likely to respond to therapy when clinical CBPs cannot be established because of the rarity of infection with unusual species of fungi. However, some PK/PD must be known about the bug/drug combination in order to determine the efficacy of a given ECV/ECOFF and account for intrinsic resistance.

When testing the antifungal activity of a novel antifungal agent, relative to standard antimycotics, we propose to use the abbreviation ‘IC₅₀ or IC₉₀’, defined as the minimum inhibitory concentration of the agent or standard antimycotic that inhibits growth of the fungus (or similar readout like metabolic activity, see further) by 50% or 90%, respectively. In this way, there will not be a mix up with MIC abbreviations, which can only be used in the context of standardized AFST assays and endpoints.

One of the major drawbacks with AFST methodologies described above is that they are time-consuming and/or have long turn-around-times. Another concern is the general subjectivity involved in reading MIC end-points and the inter-laboratory variability of MIC values, especially for methods involving visual endpoint reading and for specific antifungal drugs such as caspofungin. EUCAST testing is advantageous for yeast as an objective spectrophotometric endpoint reading is performed 23 and image analysis of disk diffusion assays 31 or measurement of optical density (OD) in broth microdilution assays can provide a normalized quantitative measure of the degree of growth inhibition in the presence of a drug relative to growth in the absence of drug. EUCAST recommends reading the MICs based on OD measurements using a wavelength of 530 nm, although other wavelengths can be used e.g. 405 nm or 450 nm.

Promising alternatives to the classical phenotypic AFST are phenotype-centered (or semi-molecular) approaches that combine a culture step with molecular analysis (i.e. by real-time PCR or matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF MS)) 13. However, in contrast to phenotypic methods, an important caveat of using PCR/sequencing is that it is suitable to detect resistance but not susceptibility as it can only detect resistance mechanisms that are already recognized. While the MALDI-TOF MS step of the analysis provides rapid analysis, the requirement for pre-assay cultured growth of the pathogen limits the ability to improve turn-around-times for more rapid diagnoses.