Fungal culture

The most direct and usually conclusive means of establishing the diagnosis of a fungal infection is to grow the fungus from a patient sample. Numerous different samples can yield a fungus, including blood, cerebrospinal fluid, pus, urine, tissue, respiratory samples (sputum, bronchoscopy lavage), pleural, pericardial or peritoneal fluid, skin scraping, hair, nail clippings, oral or vaginal samples. Processing of these samples may involve centrifugation or softening/liquidisation to allow spreading of the sample onto an agar medium. A number of laboratory manuals and specific describe the methods for doing this although remarkably few comparative studies comparing one method with another have been done.

The importance of media selection

The yield of most fungi is improved by direct culture of samples on so-called ‘fungal media’. For some fungi, cultures are always or almost always negative on bacterial media, examples being Histoplasma, Mucorales and Coccidioides spp. The culture of Aspergillus spp. on bacterial media is ~30% less effective than on fungal media [1].

General purpose media that are commonly used for fungal culture are Sabouraud dextrose, malt extract and less commonly brain heart infusion medium. To prevent contamination of the medium by bacteria, chloramphenicol is used, but prevents the growth of Actinomyces, which others grows well on Sabouraud dextrose agar. For reducing the frequency of environmental fungal growth, cycloheximide is added, but this reduces the yield of many opportunistic fungi including Aspergillus spp., Cryptococcus neoformans and Mucorales isolates. Therefore if cycloheximide is used, one agar plate not containing it should also be used in parallel.

Media for selective culture and presumptive identification

There are some additional specialised media that can be used directly on specimens to select for growth of Aspergillus species. e.g. Oxoid Sigma Pitt Cotty

Blood culture for fungi

Numerous blood culture systems have been used, including ‘standard’ bacterial cultures and more specialised systems such as the lysis centrifugation system, the Septi-Check systems, and special bottles within an automated system (sometimes of Mycobacteria and fungi). A small number of comparisons have been done of these systems, and the choice matters a great deal more when the range of potential infecting fungi includes Cryptococcus neoformans, Histoplasma capsulatum and Penicillium marneffei. In current automated blood culture systems, Candida glabrata grows more slowly and much more often in the anaerobic bottle, so both aerobic and anaerobic bottles should be filled with blood. The yield from blood culture is higher if patients are not taking antifungal agents [2] and if at least 20mL of blood is cultured.

Incubation conditions

The standard temperature for incubation of fungi is 30ºC and cultures should be incubated in a humidified environment for 21 days. They should be inspected daily for at least a week, and at least 3 times weekly thereafter. Some fungi are very slow to grow, notably Histoplasma spp., and may need longer incubation times. Culture of respiratory cultures at 42ºC prevents the growth of Candida spp. allowing Aspergillus spp. to flourish without competition.

Handling growth and procedures for identification

Once colonies are visible, they should be inspected carefully for their morphology. Yeasts can be further identified by their growth pattern on specialised media such as cornmeal agar and with biochemistry testing. Some presumptive identification of yeasts can be made based on colour and colony morphology on chromogenic media. Some rapid testing systems are available which are more or less comprehensive and reliable. Published comparisons of these methodologies are available.

Filamentous fungi are harder to identify. Without spores or other sterile structures, the only real distinction that can be made is whether the hyphae are septate or not (i.e. are likely be Mucorales, Basidiobolus ranarum or Conidiobolus spp). Media that can assist with sporulation include potato dextrose agar. If the colony will not spore, only molecular identification is possible. Once the colony has sporulated identification by microscopy using phenol cotton blue or another simple means of highlighting the distinctive structures of the fungus allows an experienced mycologist to identify to at least genus level in the vast majority of cases. Species identification of filamentous fungi can be very difficult or impossible without additional data such as colour and morphology of additional media, differential temperature growth rates and molecular information. Many cryptic (ie identical appearance) species have been described in many genera.

MALDI-TOF

Many fungi can now be speciated with matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF) [Chalupova, J et al  2013]. Some of the colony from a positive culture is mixed with a specialized reagent and directly analysed in the mass spectrometer. The time to result is approximately 1 minute. Calibrators and controls are used regularly to ensure high quality performance of the MALDI-TOF mass spectrometer. The ionisation spectrum is used by an integrated database to provide identification. However, few filamentous fungi are included in current commercial databases, whereas yeasts, especially Candida spp. are well represented. One of the challenges is that filamentous fungi often are quite variable in their growth pattern and therefore protein spectra may vary with media, time of growth and temperature, for example.

There are 2 commercial systems available for clinical microbiology labs: Bruker (Autoflex)  and BioMérieux (VITEK MS). They differ in databases, algorithms used to identify organisms and instrumentation. Each database is proprietary and not open access.

Not only can yeasts be identified from culture plates, but also directly from blood cultures. MALD-TOF performs less well when blood contains mixed species of Candida or mixed bacterial/fungal cultures.

For Aspergillus and other filamentous moulds including genera Mucor, Rhizopus and Lichtheimia 82-97% of isolates were correctly identified to the species (complex) level. Identification is also excellent for dermatophytes. Rare moulds are rare, so large studies are difficult to do.

Compared to standard methods, MALDI-TOF MS improves turn-around time for bacterial and yeast identification by an average of 1.5 days, and do not require subculture. 

 Ref : Chalupova j et al. Biotechnol Adv 2013, Nov 7, pii: S0734-9750(13)00194-8. doi: 10.1016.

Transmissible fungi from the laboratory

Plates should not be opened outside a secure environment if there is any possibility of potentially airborne, primary pathogenic fungi such as Aspergillus spp., Coccidioides spp., Histoplasma spp., Blastomyces dermatitidis, Penicillium marneffei. Laboratory transmission of infection has occurred with severe consequences. Measures to deal with inadvertent opening of plates in the open lab have recently been published [3].

Full details of fungal identification can be found in numerous books and identification keys (e.g. Fungal Infection: Diagnosis and Management 2012).

Culture interpretation

The culture of most fungi from a human sample makes the diagnosis as external contamination is rare. A positive culture of Trichophyton spp., Microsporum spp. Cryptococcus neoformans, Coccidioides spp., Histoplasma spp., Blastomyces dermatitidis, Penicillium marneffei represent disease. The 2 common exceptions are Aspergillus spp. and Penicillium spp., and occasionally other pathogenic fungi such as the bread mould Rhizopus, are plate contaminants or found in specimens but not causing disease. The interpretation of these cultures is shown in the table below.

Interpretation of positive cultures of Aspergillus, Candida and Penicillium spp.

Specimen	Aspergillus	Candida	Penicillium
Blood	Rare, may be significant or laboratory contaminant	Pathogenic and requires treatment (rare exceptions related to skin colonisation)	Not significant, a laboratory contaminant
Venous catheter	Rare, may be significant or laboratory contaminant	Probably significant – bloodstream infection likely, possibly silent.	Not significant, a laboratory contaminant
Tissue	Significant, if confirmed by histology	Significant, if confirmed by histology	Significant, if confirmed by histology
Pus or sterile site aspiration	Significant	Significant	Rare, probably significant, could be laboratory contamination
CSF	Significant	Significant	Probably laboratory contamination
Urine	Rare, possibly significant	Usually not significant. In immunocompromised  or intensive care may signal disseminated candidiasis
Sputum, other respiratory samples	Often significant – allergic, chronic or invasive disease, especially if A. fumigatus	Rarely significant.	Not significant. Possibly significant in allergy.
Peritoneal discharge after surgery	Rare, probably laboratory contaminant	Often significant –Candida peritonitis	Rare, probably laboratory contaminant
CAPD fluid	Rare, possibly significant	Significant and serious	Rare, probably laboratory contaminant
Nails	Probably significant, if toenails.	Probably significant, if nail fold or clinical appearance is that of superficial white onychomycosis	Laboratory or skin contaminant
Ear swab	Probably significant, if A. niger. If A. fumigatusconsider invasive otitis.	Probably significant	Laboratory or skin contaminant
Cornea scraping	Probably significant	Probably significant	Rare, probably laboratory contaminant