Antifungal Susceptibility Testing—Getting the Answer you Want

Abstract & Commentary

Synopsis: Any of the process variables such as inoculum, medium, incubation conditions involved in antifungal susceptibility testing can affect the results dramatically even those thought to be too trivial to mention, such as sealing the plates.

Source: Rambali B, et al. Susceptibility testing of pathogenic fungi with itraconazole: A process analysis of test variables. J Antimicrob Chemother. 2001;48:163-177.

Several sources of variation, such as inoculum size and incubation conditions, are already known to influence the results of antimicrobial susceptibility tests including those involving antifungal drugs. However, virtually all investigations into the sources of variation have examined only 1 or 2 factors at a time. This approach does not allow the relative importance of each source of variation to be determined nor does it identify possible interactions between factors that may be greater than the individual factors. Like other activities, antifungal susceptibility testing is a process and, as such, is amenable to systematic analysis of each of the elements simultaneously. To this end, Rambali and colleagues set up a factorial design in order to identify which of 10 elements exerted the most influence on determining the susceptibility of 8 Candida yeast isolates and 8 molds of different genera including a strain of Aspergillus fumigatus to itraconazole.

Statistical analysis indicated that while the source of the medium, the cultivation of the inoculum, and the nature of the solvent used to dissolve the drug did not seem important, incubation time and atmosphere, the shape of the microtitration well, and whether the well was sealed all influenced yeasts and mold in both media. However, other factors, such as glucose concentration, initial inoculum, and incubation temperature, affected yeasts and molds differently (see Table 1).

Table 1: Factors that Affected Yeasts Differently from Molds
Growth of yeasts Growth of molds

Glucose concentration
(0.2 or 2.0 %)
Improved at 2 % No influence
Sticker seal
(yes or no)
No influence Decreased when
sealed
Initial cell concentration
(1,000 or 40,000 cfu/mL)
Improved when high No influence
Incubation time
(24 h or 48 h)
Optimal by 24 h Optimal by 48 h
Incubation temperature
(30°C or 35°C)
Improved at 35°C Improved at 30°C

The range of MICs was also impressive averaging 0.06 to 8.41 mg/L for RPMI and 0.02 to 2.69 mg/L for CYG—a 419- to 539-fold difference. This means that alterations in the process in one direction can result in a low MIC but in a very high MIC in the other direction. For instance, 9 of the 30 combinations in RPMI resulted in an MIC > 4 mg/L for C glabrata J940839 compared with only 2 of the combinations tested in CYG (see Table 2).

Table 2: Range of MICs obtained
Isolate RPMI
Lowest
MIC
Highest
MIC
Range CYG
Lowest
MIC
Highest
MIC
Range

Candida albicans
B2630
0.016 4 256 0.016 32 2048

Candida albicans
B59630
0.5 32 64 0.125 32 256

Candida glabrata
B63155
0.125 32 256 0.016 2 128

Candida glabrata
J931545
1 32 32 0.016 2 128

Candida glabrata
J940839
1 32 32 0.016 32 2048

Candida krusei
ATCC 6258
0.016 0.5 32 0.016 0.25 16

Candida parapsilosis  
ATCC 22019
0.016 0.25 16 0.016 0.016 1

Candida tropicalis
CDC 44
0.016 32 2048 0.016 32 2048

Aspergillus fumigatus
NCPF7099
0.016 32 2048 0.016 2 128

Aspergillus fumigatus
J980617
0.016 1 64 0.016 0.5 32

Fusarium oxysporum
J990081
0.125 32 256 0.5 32 64

Paecilomyces lilacinus J980407 0.125 32 256 0.016 8 512

Scedosporium apiospermum J961338 0.016 4 256 0.016 2 128

Trichophyton rubrum B68183 0.016 4 256 0.016 0.125 8

Average 0.06 8.41 419 0.02 2.69 539

Comment by J. Peter Donnelly, PhD

Finally, a group of scientists has risen to the challenge of systematically testing each of the processes involved in determining the MIC of yeasts and molds to an antifungal agent, and the results are a revelation. They clearly show that antifungal susceptibility tests are affected markedly by even trivial things such as the shape of the microtitration well and whether the opening is sealed. Moreover, different strains are affected differently and molds clearly require different conditions than yeasts. This is bad news for those striving to achieve a unified approach to antifungal susceptibility testing based on the conditions and materials. It also bodes ill for standardization since one method clearly doesn’t fit all and even minor variations in technique can lead to widely divergent results. The results also vindicate the notion that one cannot refer to the MIC but only to an MIC and also provide a plausible explanation for the marked inter-laboratory variation often seen in quality control programs in which participants purport to adopting the same technique but probably have only adhered to the main aspects while ignoring the apparently minor details. Last but not least, the consumers of such tests must now be aware that it is possible to obtain the result that best suits. Given the increasing number of available antifungal agents, it is important that appropriate tests are used to define susceptibility and detect resistance in vitro since these data are crucial for epidemiological purposes, patient management, and research. The results of this study are challenging to say the least as they report the influence of process variable for only one antifungal agent, itraconazole, but show clearly the need to explore much more thoroughly the influence of each element in the process of testing the susceptibility to the rest including fluconazole, amphotericin B, caspofungin, and those close to reaching the marketplace, such as voriconazole.

Dr. Donnelly, Clinical Microbiologist, University Hospital, Nijmegen, The Netherlands, is Associate Editor of Infectious Disease Alert.