Antifungal Susceptibility Testing

It is now possible for the clinical microbiology laboratory to perform reliable in vitro antifungal susceptibility tests on a wide range of yeasts and moulds. The aim of this article is to a provide review of what is currently available to the clinical laboratory along with some practical comments on antifungal susceptibility testing. Major advances with the standardisation and clinical interpretation of in vitro antifungal susceptibility testing have been made in recent years. These include the introduction of standard reference methods for both yeasts and moulds by theCLSI (1,2,3) and publication of interpretive breakpoints, especially for Fluconazole and Itraconazole against Candida infections (4).  In the antifungal susceptibility testing world, the CLSI have set the benchmark methodology by providing laboratory tested, reproducible, consensus peer review standards that are updated on a regular basis. To do antifungal susceptibility testing you will need to purchase the relevant CLSI standards and quality control stains.

CLSI M27-A2 standard for yeasts:

See CLSI. Reference Method for Broth Dilution Antifungal Susceptibility Testing of Yeasts; Approved Standard-Second Edition.  CLSI document M27-A2. CLSI, Pennsylvania, USA 2002.

The M27-A2 standard is intended for testing yeasts including Candida species and Cryptococcus species (C. neoformans and C. gattii) by using either a macro or micro broth dilution test system. It recommends the use of RPMI-1640 medium (with glutamine and phenol red, without bicarbonate)(Sigma #R-7755 St. Louis, USA) supplemented with 0.2% glucose and buffered to a pH of 7.0 with 0.165 mol/L MOPS (3-[N-morpholino] propanesulfonic acid)(Sigma #M-6270), inoculum standardized to 0.5 McFarland using a densitometer and incubation at 35oC. It is essential to use the exact reagents as listed from Sigma. Plates are read at 24 hours for non-fastidious yeasts like Candida or at 48 to 72 hours for slower growing yeasts like Cryptococcus. 

The microdilution wells should be visualised with the aid of a reading mirror and the growth in each well should be compared with that of the growth control. A numerical score from 0 to 4 is given to each well using the following scale: 0 = optically clear, 1 = slightly hazy, 2 = prominent reduction in turbidity compared with that of the drug-free growth control, 3 = slight reduction in turbidity compared with that of the drug-free growth control, 4 = no reduction in turbidity compared with that of the drug-free growth control. The MIC for amphotericin B is the lowest concentration with a score of 0 (optically clear). The MICs for the azoles and 5FC are the lowest concentrations with a score of 2 (prominent decrease in turbidity). 

Interpretation of results:  For the most part Amphotericin B MICs for Candida species cluster between 0.25 and 1.0 ug/ml.  However it must be stressed that the M27 method does not consistently permit detection of resistant strains and isolates with MICs of > 1.0 ug/ml should be considered likely to be resistant. Antibiotic Medium 3 supplemented with 2 % glucose may permit more reliable detection of resistance but this medium is not standardized and substantial lot-to-lot variability is possible. On a brighter note, interpretative breakpoints for Candida against Fluconazole, Itraconazole and 5-Fluorocytosine have been established (table 1).  These breakpoints have also been used for isolates of Cryptococcus where there is also some correlation between elevated MIC and treatment failure1,5.

Trailing end points.  Some azoles, particularly fluconazole, exhibit a phenomenon known as trailing. Trailing occurs when the turbidity continually decreases as the drug concentration increases but the suspension fails to become optically clear (partial inhibition of growth over an extended range of antifungal concentrations). For most isolates, the difference between reading at 24 hours versus 48 hours is minimal and will not alter the interpretative category (i.e. does not change whether the isolate would be read as “susceptible” or “resistant”). However some isolates show a dramatic rise in MIC over time (e.g. for fluconazole from 0.5 ug/ml at 24 hours to 256 ug/ml at 48 hours). This trailing phenomenon has been reported as occurring in about 5% of isolates,6 however some studies have reported that up to 20% of C. albicans isolates read at 48 hours may show trailing to fluconazole that would alter the interpretation from “susceptible” to “resistant” (Fig. 1) (7,8). To help resolve this issue the M27-A2 methodology for Candida has provided both 24 hour and 48 hour microdilution MIC ranges for the two QC strains and eight systemic antifungal agents (table 2). Ideally plates should be read at 24 hours whenever there is sufficient growth.

Etest showing a trailing end point.

The CLSI M27-A2 (1) method is now the benchmark to validate all other methods against. As a result several commercial systems are now available to the clinical laboratory for antifungal susceptibility testing. 

CLSI M44-A standard for yeasts by disk diffusion:

See CLSI. Method for Antifungal Disk Diffusion Susceptibility Testing of Yeasts; Proposed Guideline.  CLSI document M44-A. CLSI, Pennsylvania, USA 2003.

The M44-A standard is a newly established methodology for disk diffusion testing of Candida species. Other yeast genera and moulds have yet to be validated using this method. The standard includes zone interpretive criteria for fluconazole and recommended quality control ranges for fluconazole and voriconazole. It recommends the use of Mueller-Hinton agar supplemented with 2% glucose and 0.5 ug/ml methylene blue dye medium. Mueller-Hinton agar is readily available and shows acceptable batch-to-batch reproducibility, the glucose provides a suitable growth for most yeasts and the addition of methylene blue enhances the zone edge definition. The pH of the medium needs to be between 7.2 and 7.4 at room temperature after gelling. The inoculum is standardized  to 0.5 McFarland using a densitometer and plates should be incubated at 35C for 24 hours. Some strains where insufficient growth has occurred after 24 hours may need  to be read after 48 hours incubation. Commercially prepared paper disks for fluconazole (25 ug) and voriconazole (1 ug) are available from Oxoid and Becton Dickinson.

Interpretative zone sizes and equivalent MICs have been set for Candida against Fluconazole (table 3). Disk tests are inexpensive and easy to set up and provide an ideal screening test . However it is recommended that all strains that appear resistant  should be confirmed against the M27-A2 microbroth dilution standard. Disk testing may be adapted for use with other fungi including sporulating moulds but once again the results need to be validated by using the appropriate CLSI reference method.

CLSI M38-A standard for moulds:

See CLSI. Reference Method for Broth Dilution Antifungal Susceptibility Testing of Filamentous Fungi; Approved Standard.  CLSI document M38-A. CLSI, Pennsylvania, USA 2002.

The M38-A standard describes a method for testing antifungal susceptibility of filamentous fungi (moulds) that cause invasive infections, including Aspergillus spp., Fusarium spp., Pseudallescheria (Scedosporium) spp., zygomycetes and other pathogenic moulds. It recommends the use RPMI-1640 medium (with glutamine, without bicarbonate, and with phenol red as a pH indicator)(Sigma #R-7755 St. Louis, USA) supplemented with 0.2% glucose and buffered to a pH of 7.0 with 0.165 mol/L MOPS (3-[N-morpholino] propanesulfonic acid)(Sigma #M-6270) as used in the M27-A2 standard for yeasts. 

However for the moulds the inoculum preparation of conidial or sporangiospores suspensions must be adjusted using a spectrophotometer with a test inoculum in the range 0.4x104 to 5x104 CFU/ml providing the most reproducible MIC data. The optical density (OD) at 530nm required is dependant on the conidial or sporangiospores size of the mould being tested; i.e. for Aspergillus and Sporothrix species the OD = 0.09 - 0.11; for Fusarium, Pseudallescheria (Scedosporium), and Rhizopus species the OD = 0.15 – 0.17; for Bipolaris and Histoplasma species the OD = 0.2. Note the addition of a very small drop of Tween 20 as a wetting agent will help to facilitate the preparation of Aspergillus inocula. 

Microdilution trays are incubated at 35oC; and may be read at 24 hours for Rhizopus species; 48 hours for Aspergillus, Fusarium and Sporothrix species; and 72 hours for slower growing moulds like Pseudallescheria (Scedosporium) species. Most moulds may be read at 48 hours.  Once again turbidity in the microdilution wells should be scored with the aid of a reading mirror and compared with that of the growth control. A numerical score from 0 to 4 is given to each well using the following scale: 0 = optically clear or absence of growth, 1 = slight growth (25% of growth control), 2 = prominent reduction in growth (50% of growth control), 3 = slight reduction in growth (75% of growth control), 4 = no reduction in growth.  The MIC for amphotericin B, Itraconzole, Voriconazole and Posaconazole is the lowest concentration with a score of 0 (optically clear).  The MICs for 5-fluorocytosine, Fluconazole and Ketoconazole are the lowest concentrations with a score of 2 or lower (50% growth reduction).

Interpretation of results:  For Amphotericin B end points are typically well defined with most moulds clustering between 0.5 and 2.0 ug/ml.  However some species such as Aspergillus terreus, Acremonium strictum, Pseudallescheria boydii and Scedosporium prolificans show higher MICs in the range of 2 to 16 ug/ml (although very little data are available, MICs above 2 ug/ml have been associated with treatment failures). RPMI medium may also be unreliable in detecting resistance to Amphotericin B. For 5-Fluorocytosine most mould MICs are greater than 64 ug/ml, the only exception are some isolates of Aspergillus and the dematiaceous fungi. Similarly for Fluconazole most mould MICs are greater than 64 ug/ml, the only exception are some isolates of the dimorphic fungi and dermatophytes. For Itraconazole, Voriconazole and Posaconazole the end points are typically well defined with MICs ranging from 0.0313 to 16 ug/ml.  No reliable breakpoints have been published for mould MICs and there is very limited in vivo data available. 

No commercial systems are yet available, although some yeast plates like the Sensititre YeastOne test may be utilised by adjusting the inoculum to the relevant OD using a spectrophotometer as described above. Inoculum density and growth controls are also essential. The main problems with testing moulds are inoculum standardisation, slow growth rates, non-sporing moulds and the interpretation of end points. Testing for antifungal susceptibility of moulds is technically more demanding.

Commercially available systems:

Sensititre® YeastOne™ Test Panel (manufactured by TREK but supplied in Australia by Dutec Diagnostics). This is a microtitre broth dilution method based on the CLSI M27-A2 standard described above. Each test consists of a disposable microtitre plate, which contains dried serial dilutions of six antifungal agents, Amphotericin B (range 0.008-16 mg/ml), Fluconazole (range 0.125-256 mg/ml), Itraconazole (range 0.008-16 mg/ml), Ketoconazole (range 0.008-16 mg/ml) and 5-Fluorocytosine (range 0.03-64 mg/ml), Voriconazole (range 0.008-16 mg/ml) in individual wells (Fig.2). The wells also contain Alamar Blue as a colorimetric indicator, which greatly improves the end point readability by a colour change from blue to pink. Results are expressed as an MIC and comparative studies against the CLSI method have shown favorable results (9,10).

Overall, the Sensititre YeastOne is a robust and reproducible test, easy to set up, the end points are clearly visible and results are within the expected range (Table 4). Each plate gives MIC results for 6 common antifungal agents. Excellent shelf life and the test also works with moulds, especially those that sporulate freely like Aspergillus.

Sensititre® YeastOne™ Test Panel

Fungitest® (manufactured in France and supplied in Australia by Bio-Rad Laboratories Pty Ltd. ex Sanofi Diagnostics Pasteur).  This is a modified microtitre broth breakpoint test also based on the CLSI M27-A2 standard described above. Each plate has six antifungal agents at two different concentrations, Amphotericin B (2 and 8 mg/ml), Fluconazole (8 and 64 mg/ml), Itraconazole (0.5 and 4 mg/ml), Ketoconazole (0.5 and 4 mg/ml), Miconazole (0.5 and 8 mg/ml) and 5-Fluorocytosine (2 and 32 mg/ml) in individual wells, plus control growth and no growth wells.  The wells also contain a redox indicator to improve the end point readability by a colour change from blue to pink (Fig. 3). Results are expressed as Susceptible, Intermediate or Susceptible Dose Dependent or Resistant with MIC’s given relative to the break points e.g. Susceptible MIC <0.5 mg/ml 11,12.

The Fungitest is easy to set up and the end points are clearly visible. Most results are within the expected range but it may be difficult to detect resistant strains, especially to Fluconazole12. Excellent shelf life, but we have not yet evaluated this for test for moulds.

Fungitest®

Etest®(manufactured by AB Biodisk in Sweden and supplied in Australia by Australian Laboratory Services Pty Ltd). This is an agar diffusion method using a strip with a predefined concentration gradient of the antimicrobial agent being tested. This gradient allows for an MIC determination to be made. Etests have been extensively used for susceptibility testing of bacteria and they are also available for antifungal agents, including Amphotericin B (range 0.002-32 mg/ml), Ketoconazole (range 0.002-32 mg/ml), Itraconazole (range 0.002-32 mg/ml), Fluconazole (range 0.016-256 mg/ml), Voriconazole (range 0.002-32 ug/ml) and 5-Fluorocytosine (range 0.002-32 mg/ml) (fig. 3). The two difficulties with Etests against fungi have always been which medium to use and with the interpretation of the end point.  The later problem remains but it is now recommended that modified RPMI-1640 agar supplemented with 0.2% glucose be used in line with the CLSI standard 13,14,15.

Etests are simple to perform and the methodology is similar to that used in bacteriology. Individual antifungals may be tested and there is reasonable correlation with the CLSI standard. Etest end points are often difficult to read due to the trailing effect seen with some strains [about 20% of yeasts (Fig. 4)]. As a result MIC's may tend to be higher than expected so it is important to repeat any “resistant” results in parallel with a control stain or to have them confirmed by another method.

Etest and Neo-Sensitab test.

Neo-Sensitabs®(manufactured by Rosco in Denmark and supplied in Australia by Dutec Diagnostics). This is a simple agar diffusion method using tablets to determine the susceptibility of fungi to antifungal agents (fig.3). Once again there have been problems with which media to use and with the interpretation of the end points. Most users have now changed over to modified RPMI-1640 agar supplemented with 0.2% glucose and have adopted the basic CLSI guidelines to create a more standardised test11. However, recent studies have used Mueller-Hinton agar supplemented with 2% glucose and 0.5 mcg/ml methylene blue as the medium (see CLSI M44-P method above) and a Biomic plate reader to electronically read and interpret zones sizes16. A large number of antifungals is available in tablets; Amphotericin B (10 ug), Ketoconazole (15 ug), Itraconazole (8 ug), Fluconazole (25 ug), Voriconazole (1.0 ug), 5-Fluorocytosine (1 and 10 ug), Clotrimazole (10 ug), Miconazole (10 ug), Econazole (10 ug), Natamycin (10 ug) and Nystatin (50 ug).

Neo-Sensitabs are cheap, agar diffusion tests are easy to set up and show potential for the screening of large numbers of isolates for resistance. However, individual disk zone sizes are often not able to differentiate between Susceptible and Susceptible Dose Dependent isolates and the correlation between zone size and MIC is more variable10. Once again, resistant isolates need to be confirmed by using one of the appropriate CLSI methodology.

Tips from the bench.

Inoculum preparation:

• Yeasts (Candida and Cryptococcus)
• Sub-culture onto Sabouraud’s Dextrose Agar and incubate for 24-48 hours at 35OC to obtain a freshly grown pure culture.
• Use a sterile cotton tip swab to harvest colonies and vortex thoroughly to ensure a homogenous suspension (no clumped cells).
• Moulds
• Sub-culture to Potato Dextrose Agar slopes and incubate for 7 days at 35oC for most moulds, except Fusarium spp. which should be incubated for 2 days at 35oC then 5 days at 25oC) to ensure maximum sporulation (shorter incubation times may be used if results are required urgently bearing in mind the need for sufficient numbers of conidia to harvest).
• Use a sterile paster pipette to gently wash the surface of the colony with sterile distilled water or saline (for Aspergillus use 1 drop of Tween 20 as a wetting agent to help disperse the conidia). 
• For filamentous fungi, which do not sporulate profusely, add one drop of the wetting agent Tween 20 to the sterile distilled water and then break up the mould by vortexing it with small glass beads to make a suitable suspension.  Allow the suspension to stand for a few minutes so that larger hyphal segments have time to settle, then use top homogenous layer to obtain the desired inoculum density.

Inoculating trays

• Always mix, mix and mix to ensure yeast cells or conidia are evenly distributed through the suspension.
• Use a multi-channel pipette.  This is important to achieve an even distribution of the inoculum, avoiding skipped wells and for reducing air bubbles.
• Incubate trays in moist chamber if not sealed.

Reading MICs

• Always compare to the growth control well, if unsure what to score, look for end-point well.
• Skipped wells may occur due to (a) inconsistent inoculum dispensed into wells, (b) poor growth habit of organism, and (c) a manufacturing error when the drug as dispensed into wells.  Repeat the test if 2 or more skipped wells are present.
• For trailing isolates check the inoculum density, repeat the test and read at 24 hours.
• If in doubt, or the result is unexpected then repeat the test.

Conclusion:

The main practical difficulties with antifungal susceptibility testing are usually with the inoculum preparation and end point determination. An inoculum sterility and density check should be done for each test [i.e. spread 100 ml of the inoculum onto a clean SDA plate and count the CFU’s at 24 hours]. This also acts as a growth control for slower growing fungi [i.e. you may have to read some plates at 48 or 72 hours]. If the inoculum is too light or heavy the test should be repeated. End point reading may also be enhanced by the use of colorimetric indicators, plate readers and automated zone readers, which allow for a more quantitative result. No one test is fool proof, if results are unexpected then the test should be repeated for confirmation.

In conclusion, we are at an exciting stage in the development of in vitro antifungal susceptibility testing and systems are now available that allow the clinical laboratory to perform these tests with some confidence. However many issues remain. It goes without saying that you need a culture, which is not always possible. The identity of the culture is often predictive of its in vitro susceptibility, so which isolates should be tested? Only those clinically warranted or should the laboratory screen all isolates?  “MC&S” (microscopy, culture and sensitivity) is not a viable option in medical mycology.  In many cases the clinical relevance of in vitro antifungal susceptibility results remains difficult to interpret and expert advice from a consulting microbiologist or infectious disease specialist may be required.

References:

• CLSI. Reference Method for Broth Dilution Antifungal Susceptibility Testing of Yeasts; Approved Standard-Second Edition.  CLSI document M27-A2 [ISBN 1-56238-469-4]. CLSI, Pennsylvania, USA 2002.

• CLSI. Method for Antifungal Disk Diffusion Susceptibility Testing of Yeasts; Proposed Guideline.  CLSI document M44-P [ISBN 1-56238-488-0]. CLSI, Pennsylvania, USA 2003.

• CLSI. Reference Method for Broth Dilution Antifungal Susceptibility Testing of Filamentous Fungi; Approved Standard.  CLSI document M38-A [ISBN 1-56238-470-8]. CLSI, Pennsylvania, USA 2002.

• Rex JH, Pfaller MA, Galgiani JN et al. (1997).  Development of interpretative breakpoints for antifungal susceptibility testing; conceptual framework and analysis of in vitro – in vivo correlation data for fluconazole, itraconazole and Candida infections.  Clin Infect Dis. 24:235-47.

• Witt MD, Lewis RJ, Larsen RA, et al. (1996).  Identification of patients with acute AIDS-associated cryptococcal meningitis who can be effectively treated with fluconazole: The role of antifungal susceptibility testing.  Clin Infect Dis.  22:322-328.

• Arthington-Skaggs BA, Warnock DW, Morrison CJ.  (2000.  Quantitation of Candida albicans ergosterol content improves the correlation between in vitro antifungal susceptibility test results and in vivo outcome after fluconazole treatment in a murine model of invasive candidiasis.  Antimicrob Agents Chemother.  44:2081-2085.

• Pfaller M.  2002.  Focus on Fungal Infections 12, Phoenix Arizona March 2002.

• Australian Candidemia Study Update June 2004.

• Espinel-Ingroff A, Pfaller M, Messer SA. et al.  (1999).  Multicenter comparison of the Sensititre YeastOne colorimetric antifungal panel with the National Committee for Clinical Laboratory Standards M27-A reference method for testing clinical isolates of common and emerging Candida spp., Cryptococcus spp., and other yeasts and yeast-like organisms.  J Clin Micro. 37:591-595.

• Ellis DH. Et. Al.  1999.  Evaluation of the Sensititre YeastOne microtitre panel, Etest and Neo-sensitab disk methods for antifungal susceptibility tests against 4 reference yeasts. Abstract MP3.24 9th IUMS International Congress of Mycology, Sydney, August 1999.

• Swinne D, Raes-Wuytack C, Van Looveren K, Desmet P.  (1999).  Comparative evaluation of Fungitest, Neo-Sensitabs and M27T-CLSI broth microdilution methods for antifungal drug susceptibility testing of Candida species and Cryptococcus neoformans.  Mycoses 42:231-237.

• Witthuhn F, Toubas D, Beguinot I. et al  (1999).  Evaluation of the fungitest kit by using strains from human immunodeficiency virus-infected patients: study of azole drug susceptibility. J Clin Micro. 37:864-6.

• Etest Technical Guide 4b AB Biodisk, Sweden.

• Chen CA, O’Donnell ML, Gordon S, Gilbert GL. (1996).  Antifungal susceptibility testing using Etest: comparison with broth macrodilution technique.  J Antimicrob Chemo. 37:265-73.

• Arikan S, Gur D, Akova M. (1997). Comparison of Etest, microdilution and colorimetric dilution with reference broth macrodilution method for antifungal susceptibility testing of clinically significant Candida species isolated from immunocompromised patients.  Mycoses 40:291-296.

• Meis, FGM, M. Petrou, J. Bille, D. Ellis, D. Gibbs and the Global Antifungal Surveillance Group.  2000. A Global Evaluation of the Susceptibility of Candida Species to Fluconazole by Disk Diffusion.  Diagnostic Microbiology and Infectious Disease 36: 215-223.

• Barry, A.L. et al. 2000. Quality control limits for broth microdilution susceptibility tests of ten antifungal agents.  J. Clin Micro. 38:3457-3459.