BioMed Research International

BioMed Research International / 2015 / Article

Research Article | Open Access

Volume 2015 |Article ID 938721 |

Hagen Frickmann, Ulrike Loderstaedt, Paul Racz, Klara Tenner-Racz, Petra Eggert, Alexandra Haeupler, Ralf Bialek, Ralf Matthias Hagen, "Detection of Tropical Fungi in Formalin-Fixed, Paraffin-Embedded Tissue: Still an Indication for Microscopy in Times of Sequence-Based Diagnosis?", BioMed Research International, vol. 2015, Article ID 938721, 11 pages, 2015.

Detection of Tropical Fungi in Formalin-Fixed, Paraffin-Embedded Tissue: Still an Indication for Microscopy in Times of Sequence-Based Diagnosis?

Academic Editor: Rosely Zancope-Oliveira
Received21 Oct 2014
Revised27 Feb 2015
Accepted09 Mar 2015
Published19 Apr 2015


Introduction. The aim of the study was the evaluation of panfungal PCR protocols with subsequent sequence analysis for the diagnostic identification of invasive mycoses in formalin-fixed, paraffin-embedded tissue samples with rare tropical mycoses. Materials and Methods. Five different previously described panfungal PCR/sequencing protocols targeting 18S and 28S ribosomal RNA gene fragments as well as internal transcribed spacer 1 and 2 fragments were evaluated with a collection of 17 formalin-fixed, paraffin-embedded tissue samples of patients with rare and/or tropical invasive mycoses, comprising chromoblastomycosis, coccidioidomycosis, cryptococcosis, histoplasmosis, mucormycosis, mycetoma/maduromycosis, and rhinosporidiosis, in a proof-of-principle analysis. Results. The primers of the panfungal PCRs readily and predominantly reacted with contaminating environmental fungi that had deposited on the paraffin blocks. Altogether three sequence results of histoplasmosis and mycetoma samples that matched the histological assessment were associated with sample age <10 years and virtually without PCR inhibition. Conclusions. The high risk of amplifying environmental contaminants severely reduces the usefulness of the assessed panfungal PCR/sequencing protocols for the identification of rare and/or tropical mycoses in stored formalin-fixed, paraffin-embedded tissues. Histological assessment remains valuable for such indications if cultural differentiation is impossible from inactivated sample material.

1. Introduction

Invasive mycoses are rare but associated with severe diseases with high mortality [1, 2]. Accordingly, reliable diagnostic approaches are necessary. Cultural growth with subsequent identification and antifungal susceptibility testing is desirable whenever possible as a diagnostic gold standard [3]. However, cultural growth of fungi is time-consuming and, as with, for example, Loboa loboi (causative agent of lobomycosis), Pneumocystis jirovecii, or Rhinosporidium seeberi (causative agent of rhinosporidiosis) [36], is not always possible. The time frame within which cultural results may be expected ranges from 1 day for very rapidly growing fungi such as Mucorales up to 6 weeks for slowly growing pathogens such as Histoplasma spp. or Paracoccidioides spp. [3]. Cultural growth of particularly harmful fungal pathogens such as Histoplasma capsulatum, which can cause invasive disease even in the immunocompetent human host and are readily transmitted under laboratory conditions [7], has to be performed under biosafety level 3 (BSL-3) conditions.

However, culture will fail if potential microbial causes of the pathological findings are not considered at the time of sample acquisition, so formalin-fixed, paraffin-embedded bioptic material is collected for subsequent histological assessment. After formalin-based sample inactivation, histological assessment is usually the method of choice in the diagnostic algorithm. Nevertheless, a recent 10-year single-center study impressively demonstrated an analytical correctness of no more than 79% of histological findings for assessed invasive mycoses [3]. Accordingly, preliminary histological results should be interpreted with care [8] and supported by cultural approaches whenever possible.

If culture is not possible for logistic, technical, or other reasons and if there is no alternative sample material available other than the formalin-fixed, paraffin-embedded material, PCR-based diagnostic approaches may be considered. Various PCR protocols for the specific detection of certain invasively growing fungi with human pathogenic potential, for example, Mucorales, in tissue have been introduced [915]. There are even published multicenter evaluations for PCR-based identification of invasive candidosis or aspergillosis [16]. Best results can be expected for short-fragment PCRs, because formalin-induced cross-linking between DNA strands or DNA–protein cross-links can inhibit the amplification of longer fragments. Such random events occur approximately every 1000 base-pairs, reducing the chances of long-fragment amplifications, particularly with low amounts of target DNA [1720]. However, the use of genus- or species-specific PCRs requires prior clinical suspicion, because primers and probes will bind only to their specific target organisms. Further, the sensitivity of PCR of formalin-fixed, paraffin-embedded tissues will decrease with low parasite density and greater sample age, as recently demonstrated for invasive amebiasis samples by our group [20].

If there is no explicit suspicion regarding the potential fungal pathogen, broad-range PCRs with consecutive sequence-based analysis have to be performed, usually amplifying longer DNA fragments of several hundred base pairs. Numerous protocols have been described for the sequence-based identification of fungi in formalin-fixed, paraffin-embedded tissue, usually targeting the 18S rRNA gene, the 28S rRNA gene, or ITS (internal transcribed spacer) fragments [10, 16, 2127]. These techniques are usually applied in industrialized, mostly Western, countries where diagnostic Sanger sequencing is readily available. Accordingly, they are predominantly evaluated with bioptic samples of patients with invasive mycoses that are typical for that part of the world, that is, invasive candidiasis, invasive aspergillosis, or, to a lesser degree, mucormycosis [21, 25, 27]. For other invasive fungi, in particular tropical fungi, evaluation data are scarce.

In this study, we evaluated five published PCR/sequencing protocols for the identification of fungal pathogens in formalin-fixed, paraffin-embedded bioptic tissue samples [2227] with samples of patients with rare and tropical mycoses from the Department of Infectious Disease Pathology at the German National Reference Centre for Tropical Diseases Bernhard Nocht Institute. The tissues were up to 30 years old and contained low concentrations of pathogens as assessed by microscopic analysis. This proof-of-principle assessment aimed to answer the questions whether the described diagnostic protocols provide reliable results with less usual fungal pathogens in complex sample matrices and whether there is still a role for microscopy in such instances.

2. Materials and Methods

2.1. Specimen Collection

In a 30-year period between 1984 and 2013, a specimen collection comprising a total of 17 paraffin-embedded, buffered formalin-fixed bioptic samples demonstrating histologically confirmed invasive mycosis due to tropical or other rare fungal pathogens was established at the Department of Infectious Disease Pathology of the Bernhard Nocht Institute for Tropical Medicine in Hamburg, Germany. Histological diagnoses comprised chromoblastomycosis (), coccidioidomycosis (), histoplasmosis (), histoplasmosis or cryptococcosis (), mucormycosis (), mycetoma/maduromycosis (), and rhinosporidiosis (). Culture-based diagnostic results were not available. The sample collection included lymphatic node tissue (), skin biopsy (), a biopsy of a nasal polyp (), vulva exudate (), lung tissue (), bone, muscular tissue, and connective tissue from the spine extension of the third thoracic vertebra (), tissue from the tricuspid valve (), bioptic material from the bottom lip (), material from the ethmoid (), and a biopsy from a wound of the foot (). The ages of the samples ranged from 1 to 30 years at the time of DNA extraction, with a median of 11 years in a left-shifted distribution (Table 1).

Sample I.D.Histological diagnosisSample materialSample age (years)

Case  1ChromoblastomycosisSkin biopsy30
Case  2MucormycosisMaterial from the ethmoid17
Case  3HistoplasmosisLymph node biopsy15
Case  4MucormycosisSkin biopsy14
Case  5Histoplasmosis or cryptococcosisLymph node biopsy13
Case  6ChromoblastomycosisSkin biopsy12
Case  7RhinosporidiosisBiopsy of a nasal polyp12
Case  8Mycetoma/maduromycosisVulva exudate11
Case  9RhinosporidiosisSkin biopsy11
Case  10Mycetoma/maduromycosisLung tissue11
Case  11HistoplasmosisBone, muscular tissue, and connective tissue from the spine extension of third thoracic vertebra11
Case  12HistoplasmosisTissue from the tricuspid valve11
Case  13ChromoblastomycosisSkin biopsy8
Case  14HistoplasmosisMaterial from the bottom lip8
Case  15CoccidioidomycosisLung tissue6
Case  16CoccidioidomycosisLung tissue5
Case  17Mycetoma/maduromycosisBiopsy from a wound of the foot1

To exclude microtome-associated contamination, three randomly chosen, recently obtained bioptic samples from patients with diseases other than invasive mycosis were included in the analyses as negative controls. As invasive mycosis is a rare diagnosis [1, 2], the risk of the negative control patients being infected with invasively growing fungi can be considered extremely low.

2.2. Sample Preparation

The small quantities of residual material from the tissue blocks in the specimen collection allowed for only one mode of DNA preparation, which has been reported to be optimal for the recovery of fungal DNA from formalin-fixed, paraffin-embedded tissues [26], with minor modifications. One 25 μm thick section was obtained from each paraffin block, including the blocks with the negative control tissues. Weighting of the samples was not performed. Instead of this, DNA amounts within the samples were measured as detailed below. The negative control samples were distributed among the tissue blocks and each section was cut at a different position of the disposable knife of the microtome. If all samples had been cut at the same position of the disposable knife, this might have allowed for DNA cross-contamination due to attached cells at that position of the knife from other, previously cut tissues. For deparaffinization, the sections were put into 1.5 mL tubes and exposed for 2 × 10 minutes to 1200 μL xylene and for 3 × 10 minutes to 1200 μL 100% ethanol under gentle constant agitation. Following each 10-minute step, the supernatant was discarded after centrifugation for 10 minutes at 13 000 ×g. Finally, the samples were air-dried. The further DNA extraction was performed according to the manufacturer’s instructions for the DNA FFPE tissue kit (Qiagen, Hilden, Germany) with the following modifications. Proteinase K digestion at 56°C was performed overnight until the suspension was clear. In line with the described protocol [26], fungal cell walls in the pellet were lysed with 400 units of Arthrobacter luteus lyticase L2524 (Sigma-Aldrich Corporation, St. Louis, MO, USA; a surrogate product for lyticase L4276 (Sigma-Aldrich Corporation) which was no longer available) per sample for 45 minutes at 37°C. The rest of the DNA extraction was performed exactly as described by the manufacturer of the DNA FFPE tissue kit. Subsequently, the material was used for PCR.

In addition to thick sections for PCR, thinner 4–6 μm sections were obtained to histologically confirm and, in case of Histoplasma capsulatum detection, semiquantify the presence of fungal pathogens in tissue by classical hematoxylin and eosin (HE), Giemsa, periodic acid-Schiff (PAS), and Grocott staining.

2.3. Molecular Diagnostic Approaches
2.3.1. Sample Quality Assessment and Inhibition Control PCR

The DNA amount in the samples was quantified using a Pico 100 Picodrop Microliter Spectrophotometer (Picodrop Ltd., Hinxton, UK) according to the manufacturer’s instructions. DNA quality was assessed in all samples by a Taqman PCR targeting a 155-base-pair fragment of the human 18S rRNA gene as previously described [12] with minor modifications (Table 2).

Sample I.D.Histological diagnosisCt-value of phocid herpesvirus PCRCt-value of human 18S rRNA gene PCRDNA concentration as measured by PicodropHistoplasma-specific nested PCR (and consecutive sequencing)Mucorales-specific seminested PCR (and consecutive sequencing)

Case  1Chromoblastomycosis222312.3 ng/µLN.a.N.a.
Case  2Mucormycosis181514.6 ng/µLNegativeNegative
Case  3Histoplasmosis1814246.3 ng/µLNegativeNegative
Case  4Mucormycosis2026.9 ng/µLNegativePositive (100% Lichtheimia/Absidia corymbifera)
Case  5Histoplasmosis or Cryptococcosis2019157.6 ng/µLNegativeNegative
Case  6Chromoblastomycosis2630414.5 ng/µLN.a.N.a.
Case  7Rhinosporidiosis262362.7 ng/µLN.a.N.a.
Case  8Mycetoma/maduromycosis2233618.9 ng/µLN.a.N.a.
Case  9Rhinosporidiosis182221.9 ng/µLN.a.N.a.
Case  10Mycetoma/maduromycosis251516.8 ng/µLN.a.N.a.
Case  11Histoplasmosis2130104.6 ng/µLNegativeNegative
Case  12Histoplasmosis241544.3 ng/µLNegativePositive (no interpretable sequence results due to poor sequence quality)
Case  13Chromoblastomycosis231433.8 ng/µLN.a.N.a.
Case  14Histoplasmosis2017175.5 ng/µLPositive (99% Histoplasma capsulatum)Negative
Case  15Coccidioidomycosis242116.8 ng/µLN.a.N.a.
Case  16Coccidioidomycosis2414107.5 ng/µLN.a.N.a.
Case  17Mycetoma/maduromycosis181326.3 ng/µLN.a.N.a.
Control 1Negative control2234609.7 ng/µLN.a.N.a.
Control 2Negative control2325655.3 ng/µLN.a.N.a.
Control 3Negative control1931676.6 ng/µLN.a.N.a.

Plasmids containing phocid herpesvirus 1 (PhHV-1) sequences were added to each sample as internal controls to exclude sample inhibition. Dilution was chosen to achieve cycle threshold- (Ct-) values of about 18. Primers and probes were used as previously described [28] for the amplification of an 89-base-pair fragment with slight modifications (Supplementary Material 1).

2.3.2. Specific PCR for Histoplasma spp. and Mucorales with Consecutive Sequencing

Samples with histological suspicion of histoplasmosis or Mucorales infection were subjected to specific PCRs with consecutive sequencing for confirmatory testing. Nested PCR for histoplasmosis targeting the gene encoding the unique fungal 100-kDa-like protein [29] and seminested PCR for Mucorales [30] were applied as described. Only cases with positive results were considered as confirmed infections. In case of negative PCR results, cases were defined as “histologically suspected” Histoplasma or Mucorales infections.

2.3.3. Panfungal PCRs

A total of five panfungal PCRs that have been evaluated for the diagnostic identification of invasive mycoses, including in some cases from formalin-fixed, paraffin-embedded tissue [2227], were applied; in the following these are referred to as PCR 1 to PCR 5. The PCR panel comprised three real-time PCRs and two traditional block cycler PCRs. The PCR targets comprised fragments of the 18S rRNA gene (PCR 1, according to [22, 23], applied on a LightCycler 2.0 (Roche, Basel, Switzerland)) and the 28S rRNA gene (PCR 3, according to [24, 27], applied on a Corbett RotorGene 6000 (Qiagen, Hilden, Germany)) as well as the internal transcribed spacer (ITS) regions ITS-1 (PCR 5, according to [25], applied on a TProfessional Basis cycler (BioMetra, An Analytik Jena Company, Jena, Germany)) and ITS-2 (PCR 2, according to [24, 27], applied on a Corbett RotorGene 6000, and PCR 4, according to [26], applied on a TProfessional Basis cycler) with species-dependent variable fragment lengths between 200 and 500 base pairs [2227]. All PCR protocols were adapted to the cyclers that were used prior to the analysis. The respective adaptations were cycler-specific and comprised reaction mixes, MgCl2 contents, optimal primer-probe-concentrations, and cycling protocols. Presentation in all details can be found in supplementary material 1 (see Supplementary Material available online at of this paper.

DNA of clinical Candida glabrata and Candida tropicalis isolates was used as positive controls. Water served as a negative control. The bands of the traditional block cycler PCRs were visualized using a FlashGel System (Lonza, Basel, Switzerland) according to the manufacturer’s instructions. Amplicons of positive PCRs were purified using the NAT Clean-up/Nucleospin Extrakt II kit (Macherey & Nagel, Düren, Germany) according to the manufacturer’s instructions and sent for Sanger sequencing to SeqLab GmbH (Göttingen, Germany). Sequences from the ab1-files obtained were aligned using BioNumerics 7.1 software (Applied Maths, Sint-Martens-Latem, Belgium). The alignment settings were open gap penalty 100%, unit gap penalty 0%, match score 100%, and fast algorithm (= minimum match sequence 2, maximum number of 98).

The sequences obtained were compared with deposited sequence information using the BLAST algorithm ( excluding human and model sequences and restricting the displayed results to fungal sequences. Best matches regarding both coverage and sequence identity were considered. To evaluate the quality of the results, the criteria suggested by the CLSI (Clinical and Laboratory Standards Institute) guideline MM18-A “Interpretive Criteria for Identification of Bacteria and Fungi by DNA Target Sequencing; Approved Guideline” [31] were employed. In detail, ≥99% identity was demanded for identification at species level, ≥97% identity for identification at genus level. Further, 0.8% separation between different species was demanded for a reliable discrimination. Deposition of obtained sequences to databases was not intended and was not carried out because no unambiguous diagnostic gold standard was available with which to confirm the fungal species identity.

2.4. Microscopic Assessment

Thin 4–6 μm neighboring sections that were cut in addition to the thick sections for PCR were analyzed by microscopy of HE-, Giemsa-, PAS-, and Grocott-stained slides by an experienced pathologist. Assessment of fungal elements was done using a Zeiss AxioImager M1 microscope equipped with an AxioCam MRc5 digital camera and AxioVisionRel 4.6 software (Zeiss, Jena, Germany). Briefly, nonoverlapping digital images (160 μm × 210 μm) were captured using a ×40 objective lens. Illustrations of representative mycosis samples are shown in Figure 1. In case of histological suspicion of Histoplasma capsulatum infection, semiquantification was performed. To do so, the numbers of pathogens per unit area (160 μm × 210 μm) were manually counted and averaged. In 4 out of 5 samples, 15 unit areas were assessed. The smaller size of the fifth sample allowed for the assessment of 13 unit areas only.

2.5. Ethics

Ethical clearance for the anonymous retrospective molecular assessment of residual materials from formalin-fixed, paraffin-embedded tissues for evaluation purposes was obtained from the Ethics Committee of the Medical Association of Hamburg, Germany (document number WF-028/13).

3. Results

3.1. Morphological Sample Characteristics

A total of 17 formalin-fixed, paraffin-embedded tissue specimens from patients with histologically confirmed invasive mycoses, deposited for 1 to 30 years (mean 11.5 years), were included in the analysis. Microscopic assessment allowed the detection of fungal elements on microscopic slides of all mycosis samples. The fungal density varied substantially from slide to slide owing to inhomogeneous distribution of the pathogens in tissue. Furthermore, in the case of filamentous fungi, elements of a multiply-cut filament were indistinguishable from single cuts of multiple filaments. For these reasons, semiquantification of fungi per unit area (160 μm × 210 μm) was abandoned as unreliable. In case of histological suspicion of nonfilamentous Histoplasma capsulatum, semiquantification was performed. An average of < 10 pathogens per sample unit was observed for all 5 samples, ranging from 2 to 9 fungal elements per sample unit.

3.2. Sample Quality Assessment

The mean DNA-amount in the fungal samples was 123.6 ng/μL, ranging from 12.3 to 618.9 ng/μL. In the fresh negative control material, higher DNA amounts were measured with a mean value of 647.2 ng/μL, ranging from 609.7 to 676.6 ng/μL (Table 2).

Positive results of the DNA-quality assessment 18S rRNA gene PCR were obtained for all but one sample. However, Ct-values were distributed over a wide range from 14 to 33. Ct-values did not seem to be influenced by sample age alone, as there was also a 17-year-old sample with a Ct-value as low as 15 (Table 2). The variety of sample materials (Table 1) made a direct comparison difficult, however.

Inhibition control PCR was positive for all analyzed samples. However, Ct-values ranged from 18 to 26, corresponding to low to moderate sample inhibition up to about 2.5 decadic logarithmic steps. Again, there was no obvious association between sample age and sample inhibition (Tables 1 and 2).

3.3. Results of Specific PCRs with Consecutive Sequencing

Five samples with histological suspicion of histoplasmosis and two samples with histological suspicion of Mucorales infections were subjected to specific PCR with consecutive sequencing. Only one out of five suspected histoplasmosis cases and one out of two suspected Mucorales infections could be confirmed by this approach (Table 2). Consequently, the remaining 5 unconfirmed cases were considered as histologically suspected only.

3.4. PCR/Sequencing Results in relation to Their Potential Predictors

The panfungal PCRs readily reacted with DNA from contaminating spores of environmental fungi, even in the negative control materials, as demonstrated by the sequencing results (Table 3). Details on PCR and sequencing results, comprising cycle threshold (Ct) values and peaks of the melting curve analyses, sequence fragment lengths as well as sequence identity, and coverage percentages, are presented in the supplementary materials 2–6 of this paper. In repeated instances (Table 3), sequence quality was noninterpretably poor, mostly due to overlapping sequences of different fungal species, when the used panfungal primers also reacted with fungal DNA of contaminants from the environment. Depending on the panfungal PCR used, sequence fragments of different fungal species were obtained from the same sample in several instances (Table 3). Primers of each PCR thus preferentially reacted with certain contaminants that differed from one protocol to another. Only in one case of histologically suspected histoplasmosis affecting cardiac valve material did all positive PCRs with interpretable sequence results suggest Candida parapsilosis, suggesting histological confusion of invasive candidiasis with histoplasmosis [8] in this instance.

Sample I.D.Histological diagnosisPCR 1PCR 2PCR 3PCR 4PCR 5

Case  1ChromoblastomycosisMultiple species, no discrimination possibleNoninterpretable, poor/overlapping sequencesUmbilicaria torrefacta/Placynthium nigrum Noninterpretable, poor/overlapping sequencesNoninterpretable, poor/overlapping sequences

Case  2MucormycosisAspergillus spp./Penicillium spp.Candida parapsilosis Noninterpretable, poor/overlapping sequencesNoninterpretable, poor/overlapping sequencesCandida parapsilosis

Case  3HistoplasmosisCryptococcus carnescens/Cryptococcus psychrotolerans/Cryptococcus peneaus/Taphrina maculans Noninterpretable, poor/overlapping sequencesVarious Umbilicaria spp.Various Penicillium spp.

Case  4MucormycosisDelitschia didyma Various Candida spp./Yarrowia spp./Candida galli Multiple species, no discrimination possible/Neofusicoccum spp.

Case  5Histoplasmosis or cryptococcosisLactarius indigo/Candida piceae/Glomus spp.Noninterpretable, poor/overlapping sequencesVarious Umbilicaria spp.Noninterpretable, poor/overlapping sequencesNoninterpretable, poor/overlapping sequences

Case  6ChromoblastomycosisXylomyces spp./Speiropsis spp./Jahnula spp./Brachiosphaera spp./Megalohypha spp./Lactarius indigo  Candida piceae/Glomus spp.Noninterpretable, poor/overlapping sequencesNoninterpretable, poor/overlapping sequences

Case  7RhinosporidiosisNoninterpretable, poor/overlapping sequencesNoninterpretable, poor/overlapping sequencesNoninterpretable, poor/overlapping sequencesCandida spp./Malassezia spp./Starmerella spp.

Case  8Mycetoma/maduromycosisLactarius indigo/Candida piceae/Glomus spp.Noninterpretable, poor/overlapping sequences

Case  9RhinosporidiosisMultiple species, no discrimination possibleNoninterpretable, poor/overlapping sequencesVarious Umbilicaria spp.Rhizoctonia solani Noninterpretable, poor/overlapping sequences

Case  10Mycetoma/maduromycosisMultiple fungal species including Penicillium spp./Aspergillus spp.Noninterpretable, poor/overlapping sequencesNoninterpretable, poor/overlapping sequencesVarious Penicillium spp./Talaromyces flavus Noninterpretable, poor/overlapping sequences

Case  11HistoplasmosisNoninterpretable, poor/overlapping sequencesCandida parapsilosis Noninterpretable, poor/overlapping sequencesTrametes spp./Tricholoma spp./Phellinus spp.

Case  12HistoplasmosisCandida parapsilosis Candida parapsilosis Candida parapsilosis Candida parapsilosis Candida parapsilosis/Tremellalesspp./Saccharomycetales spp.

Case  13ChromoblastomycosisMultiple fungal species including Penicillium spp./Aspergillus spp.Candida parapsilosis/various Candida spp.Penicillium spp.Aspergillus spp./Ajellomyces capsulatus/Histoplasma capsulatum Candida parapsilosis

Case  14HistoplasmosisAjellomyces capsulatus/Histoplasma capsulatum Noninterpretable, poor/overlapping sequencesPaecilomyces variotii Ajellomyces  capsulatus/ Histoplasmacapsulatum

Case  15CoccidioidomycosisMultiple species, no discrimination possibleCandida parapsilosis Noninterpretable, poor/overlapping sequencesCandida parapsilosis Candida parapsilosis

Case  16CoccidioidomycosisDelitschia didyma Candida parapsilosis Noninterpretable, poor/overlapping sequencesNoninterpretable, poor/overlapping sequencesNoninterpretable, poor/overlapping sequences

Case  17Mycetoma/maduromycosisDelitschia didyma Noninterpretable, poor/overlapping sequencesMyceliophthora thermophila/Thielavia terrestris/Neurospora crassa Madurella mycetomatis Noninterpretable, poor/overlapping sequences

Control 1Negative controlFunneliformis spp./Glomus spp./Candida salmanticensis Noninterpretable, poor/overlapping sequencesVarious Umbilicaria spp.Davidiella tassiana

Control 2Negative controlNoninterpretable, poor/overlapping sequencesNoninterpretable, poor/overlapping sequencesVarious Umbilicaria spp.Noninterpretable, poor/overlapping sequences

Control 3Negative controlNoninterpretable, poor/overlapping sequencesNoninterpretable, poor/overlapping sequencesUmbilicaria spp./Lasallia spp./Boreoplaca ultrafrigida/Rhizoplaca huashanensis Noninterpretable, poor/overlapping sequences

Sequences of the microscopically observed pathogens, however, were not detected with the exception of three instances. Panfungal PCRs 1 and 4 identified Histoplasma capsulatum sequence fragments in bioptic material from the patient’s bottom lip (Table 3). Of note, this was the only histoplasmosis case that was confirmed by specific PCR (Table 2). For PCR 1, however, sequence quality was unacceptably low with only 92% sequence identity and coverage of only 71% (supplementary material 2). Further, PCR 4 (Table 3) allowed the identification of Madurella mycetomatis from a biopsy of a wound on the foot of a mycetoma patient.

All sample materials with correct species identifications were younger than 10 years of age. Further, no significant PCR inhibition was detectable for the respective samples; the Ct-values were ≤20. Of note is that the missed coccidioidomycosis and chromoblastomycosis cases that were younger than 10 years all showed Ct-values ≥23 in the inhibition control PCR (Table 2). However, even in young samples without detectable relevant inhibition, not all PCR and sequencing approaches led to correct results as shown for the histoplasmosis sample from the bottom lip (Table 3). The number of correct sequence-based identifications was too low to allow a statistical assessment of favorable factors.

With the exceptions of panfungal PCRs 1 and 4, the other panfungal PCRs led to a varying number of failed reactions in the 17 microscopically positive samples assessed, that is, one failure for PCR 2, three failures for PCR 3, and as many as seven failures for PCR 5. In contrast, PCRs 1–4 showed positive results with negative control samples, while only PCR 5 was not associated with such false-positive reactions (Table 3). For PCR 1, Ct-value assessment against background fluorescence was impossible for unknown reasons, allowing the identification of positive test results only by melting curve analysis.

4. Discussion

In addition to the requirement for the presence of a critical number of pathogens in the analyzed materials, the histological identification of pathological fungi in tissue sections of patients with invasive mycosis demands specific skills and a high degree of experience, as suggested by the considerable number of mismatches in comparison with cultural diagnostic approaches [3, 8]. In line with these previous results, only one out of five histologically suspected histoplasmosis cases and one out of two suspected Mucorales infections could be confirmed by specific PCR in the approach described herein.

Molecular tests such as panfungal PCR with subsequent sequence analysis might help pathologists confirm the differential diagnosis, but nonspecies-specific diagnostic approaches are easily disturbed by environmental contamination, for example, located within the paraffin, unless protective measures are assured. Species-specific simplex or multiplex PCRs [916] might be an alternative. Other than panfungal PCRs with subsequent sequence analysis, however, such species-specific assays require a specific suspicion regarding the potential etiological agent for their selection.

Here we assessed the performance of five panfungal PCRs with subsequent sequence analysis that were designed for the diagnosis of invasive mycosis partly from rather difficult material such as formalin-fixed, paraffin-embedded tissues and evaluated in Western industrialized settings [2227] for the diagnosis of rare and/or tropical invasive mycoses, making use of a collection of 17 samples from the German National Reference Center for Tropical Infections. All samples were originally obtained from patients with confirmed invasive rare and/or tropical mycosis and were analyzed by experienced pathologists.

Some of the assessed panfungal PCR and sequencing protocols showed promising evaluation results in previous studies [2527] with invasive mycoses due to Candida spp. or Aspergillus spp. even from formalin-fixed, paraffin-embedded tissue samples after appropriate deparaffinization. The results presented here demonstrate, however, that panfungal PCRs from tissues in long-term stored paraffin blocks are generally prone to reactions with fungal spore contaminants, at least if no special preventive measures against contamination are employed. Contamination of the tissue-containing paraffin blocks with spores of environmental fungi was highly likely, as the blocks were stored for several years without particular precautions against deposition of spores. Contamination of paraffin with fungal spores might be an alternative explanation. Interestingly, the primers reacted more readily with DNA of environmental contaminants than with the rare and/or tropical fungi in our sample collection, thus considerably reducing the value of the protocols for use in cases of suspicion of tropical mycoses. Even more relevantly, positive PCR results occurred in negative control samples and failed PCR was infrequently observed in microscopically positive samples. Accordingly, PCR did not even reliably discriminate invasive mycosis from nonmycotic disease. Sample age and PCR inhibition seemed to be relevant factors here.

According to our analysis, the few correct sequencing results were observed in comparatively fresh samples that were younger than 10 years of age and that did not show considerable PCR inhibition. Even moderate inhibition prevented the amplification and subsequent sequencing of DNA fragments of the microscopically observed invasive fungal pathogens. The reliability of PCR and sequencing seems to be affected by sample quality and sample storage time. While 20 years of storage time marked the cut-off for diagnostic reliability of target-specific PCR as observed in a previous study [20], this time limit is obviously even shorter in the case of panpathogen PCRs with subsequent sequencing. Notably, the influence of storage time on the quality of DNA was not directly reflected in the semiquantitative results of a PCR targeting a 155-base-pair fragment of the human 18S rRNA gene owing to the variety of different sample materials and degrees of PCR inhibition. This variety limits the interpretability of the results of this proof-of-principle assessment with rare and/or tropical causes of invasive mycosis.

As in recent reports on fungal specimens [26], the mode of sample preparation was in principle effective in releasing fungal DNA and allowing its amplification even in samples that had been stored for years. However, PCR inhibition remained considerable in several samples. Alternative protocols for DNA preparation from formalin-fixed, paraffin-embedded tissue [26] might have led to an increased yield of fungal DNA from the microscopically observed pathogens but could not be assessed as the low quantities of residual material did not allow for comparative testing of different preparation schemes, an admitted limitation of the study.

Panfungal PCR with subsequent sequencing clearly performed less well than classical microscopy of stained sections for the identification of invasive rare and/or tropical fungal infections, at least when histology was performed by experienced pathologists. Matching sequence results were observed only in individual instances, underlining the importance of preserving histological skills in diagnostic routine. As suggested by the results of specific PCR analyses with the samples of histologically suspected histoplasmosis and mucormycosis as well as by previous studies [3, 8], however, specificity of microscopy shows considerable limitations as well.

5. Conclusion

We conclude that sequence results obtained after panfungal PCR can only be considered as confirmatory information in case of matching with preliminary histological results owing to the high risk of contamination of paraffin blocks with environmental fungal spores. Other histological techniques such as immunohistochemistry [8, 32] and in situ hybridization [8, 33] or, alternatively, specific PCRs for formalin-fixed, paraffin-embedded tissues [8, 29] can be used for the confirmation of diagnosis if mycological culture is not possible. Preliminary histological results should be interpreted with care and supported by cultural approaches whenever possible.

If mycological culture is possible, it is the reference method for the diagnosis of mycosis. In many mycoses, histology only allows a presumptive diagnosis that needs confirmation by mycological culture [8].

Conflict of Interests

The authors declare that there is no conflict of interests regarding the publication of this paper.

Authors’ Contribution

Hagen Frickmann and Ulrike Loderstaedt equally contributed to this work.


The authors are grateful to Steffen Lohr and Annett Michel for their excellent technical assistance. The PCR/sequencing analyses were funded by Grant 12K2-S-451315 “optimizing of microbiological diagnostic preanalytics for tropical conditions” of the German Ministry of Defense (MoD). The study was supported by the Körber Foundation to Paul Racz.

Supplementary Materials

Altogether, 6 supplementary materials are presented to provide further details regarding the applied methods and the observed results.

Supplementary material 1 describes the applied pan-fungal PCR protocols, the inhibition control PCR and the sample quality assessment PCR in detail.

The supplementary materials 2-6 are an addition to table 3 of the paper. They provide more detailed information regarding the observed ct-values, melting curve peaks (if applicable), detection of visible bands in electrophoresis (if applicable), readable sequence fragment lengths, coverage and sequence identities of pan-fungal PCRs 1-5. Thereby, they allow for a better interpretation of PCR and sequencing results.

  1. Supplementary Material


  1. K. Vandewoude, S. Blot, D. Benoit, P. Depuydt, D. Vogelaers, and F. Colardyn, “Invasive aspergillosis in critically ill patients: analysis of risk factors for acquisition and mortality,” Acta Clinica Belgica, vol. 59, no. 5, pp. 251–257, 2004. View at: Publisher Site | Google Scholar
  2. J. J. Castón-Osorio, A. Rivero, and J. Torre-Cisneros, “Epidemiology of invasive fungal infection,” International Journal of Antimicrobial Agents, vol. 32, supplement 2, pp. S103–S109, 2008. View at: Publisher Site | Google Scholar
  3. A. R. Sangoi, W. M. Rogers, T. A. Longacre, J. G. Montoya, E. J. Baron, and N. Banaei, “Challenges and pitfalls of morphological identification of fungal infections in histologic and cytologic specimens. A ten-year retrospective review at a single institution,” The American Journal of Clinical Pathology, vol. 131, no. 3, pp. 364–375, 2009. View at: Publisher Site | Google Scholar
  4. J. Schwarz, “The diagnosis of deep mycoses by morphologic methods,” Human Pathology, vol. 13, no. 6, pp. 519–533, 1982. View at: Publisher Site | Google Scholar
  5. J. C. Watts, “Surgical pathology and the diagnosis of infectious diseases,” American Journal of Clinical Pathology, vol. 102, no. 6, pp. 711–712, 1994. View at: Google Scholar
  6. V. Thankamani and L. Dev, “Rhinosporidium seeberi proven as a fungus for the first time after a century since its discovery,” Research in Biotechnology, vol. 3, no. 1, pp. 41–46, 2012. View at: Google Scholar
  7. D. L. Sewell, “Laboratory-associated infections and biosafety,” Clinical Microbiology Reviews, vol. 8, no. 3, pp. 389–405, 1995. View at: Google Scholar
  8. J. Guarner and M. E. Brandt, “Histopathologic diagnosis of fungal infections in the 21st century,” Clinical Microbiology Reviews, vol. 24, no. 2, pp. 247–280, 2011. View at: Publisher Site | Google Scholar
  9. H. Einsele, H. Hebart, G. Roller et al., “Detection and identification of fungal pathogens in blood by using molecular probes,” Journal of Clinical Microbiology, vol. 35, no. 6, pp. 1353–1360, 1997. View at: Google Scholar
  10. S. A. Balajee, L. Sigler, and M. E. Brandt, “DNA and the classical way: identification of medically important molds in the 21st century,” Medical Mycology, vol. 45, no. 6, pp. 475–490, 2007. View at: Publisher Site | Google Scholar
  11. D. J. Hata, S. P. Buckwalter, B. S. Pritt, G. D. Roberts, and N. L. Wengenack, “Real-time PCR method for detection of zygomycetes,” Journal of Clinical Microbiology, vol. 46, no. 7, pp. 2353–2358, 2008. View at: Publisher Site | Google Scholar
  12. P. D. Khot, D. L. Ko, R. C. Hackman, and D. N. Fredricks, “Development and optimization of quantitative PCR for the diagnosis of invasive aspergillosis with bronchoalveolar lavage fluid,” BMC Infectious Diseases, vol. 8, article 73, 2008. View at: Publisher Site | Google Scholar
  13. K. Hrncirova, M. Lengerova, I. Kocmanova et al., “Rapid detection and identification of mucormycetes from culture and tissue samples by use of high-resolution melt analysis,” Journal of Clinical Microbiology, vol. 48, no. 9, pp. 3392–3394, 2010. View at: Publisher Site | Google Scholar
  14. S. P. Hammond, R. Bialek, D. A. Milner, E. M. Petschnigg, L. R. Baden, and F. M. Marty, “Molecular methods to improve diagnosis and identification of mucormycosis,” Journal of Clinical Microbiology, vol. 49, no. 6, pp. 2151–2153, 2011. View at: Publisher Site | Google Scholar
  15. M. Fleischhacker, S. Schulz, K. Jöhrens et al., “Diagnosis of chronic disseminated candidosis from liver biopsies by a novel PCR in patients with haematological malignancies,” Clinical Microbiology and Infection, vol. 18, no. 10, pp. 1010–1016, 2012. View at: Publisher Site | Google Scholar
  16. P. L. White, R. Barton, M. Guiver et al., “A consensus on fungal polymerase chain reaction diagnosis?: a United Kingdom-Ireland evaluation of polymerase chain reaction methods for detection of systemic fungal infections,” Journal of Molecular Diagnostics, vol. 8, no. 3, pp. 376–384, 2006. View at: Publisher Site | Google Scholar
  17. R. M. Hagen, Y. P. Gauthier, L. D. Sprague et al., “Strategies for PCR based detection of Burkholderia pseudomallei DNA in paraffin wax embedded tissues,” Journal of Clinical Pathology—Molecular Pathology, vol. 55, no. 6, pp. 398–400, 2002. View at: Publisher Site | Google Scholar
  18. N. Quach, M. F. Goodman, and D. Shibata, “In vitro mutation artifacts after formalin fixation and error prone translesion synthesis during PCR,” BMC Clinical Pathology, vol. 4, article 1, pp. 1–5, 2004. View at: Publisher Site | Google Scholar
  19. K. Lu, W. Ye, L. Zhou et al., “Structural characterization of formaldehyde-induced cross-links between amino acids and deoxynucleosides and their oligomers,” Journal of the American Chemical Society, vol. 132, no. 10, pp. 3388–3399, 2010. View at: Publisher Site | Google Scholar
  20. H. Frickmann, K. Tenner-Racz, P. Eggert et al., “Influence of parasite density and sample storage time on the reliability of Entamoeba histolytica-specific PCR from formalin-fixed and paraffin-embedded tissues,” Diagnostic Molecular Pathology, vol. 22, no. 4, pp. 236–244, 2013. View at: Publisher Site | Google Scholar
  21. T. Henry, P. C. Iwen, and S. H. Hinrichs, “Identification of Aspergillus species using internal transcribed spacer regions 1 and 2,” Journal of Clinical Microbiology, vol. 38, no. 4, pp. 1510–1515, 2000. View at: Google Scholar
  22. J. Loeffler, N. Henke, H. Hebart et al., “Quantification of fungal DNA by using fluorescence resonance energy transfer and the light cycler system,” Journal of Clinical Microbiology, vol. 38, no. 2, pp. 586–590, 2000. View at: Google Scholar
  23. N. E. Jordanides, E. K. Allan, L. A. McLintock et al., “A prospective study of real-time panfungal PCR for the early diagnosis of invasive fungal infection in haemato-oncology patients,” Bone Marrow Transplantation, vol. 35, no. 4, pp. 389–395, 2005. View at: Publisher Site | Google Scholar
  24. P. D. Knot, D. L. Ko, and D. N. Fredricks, “Sequencing and analysis of fungal rRNA operons for development of broad-range fungal PCR assays,” Applied and Environmental Microbiology, vol. 75, no. 6, pp. 1559–1565, 2009. View at: Publisher Site | Google Scholar
  25. E. Dannaoui, P. Schwarz, M. Slany et al., “Molecular detection and identification of Zygomycetes species from paraffin-embedded tissues in a murine model of disseminated zygomycosis: a collaborative European Society of Clinical Microbiology and Infectious Diseases (ESCMID) Fungal Infection Study Group (EFISG) evaluation,” Journal of Clinical Microbiology, vol. 48, no. 6, pp. 2043–2046, 2010. View at: Publisher Site | Google Scholar
  26. C. Muñoz-Cadavid, S. Rudd, S. R. Zaki et al., “Improving molecular detection of fungal DNA in formalin-fixed paraffin-embedded tissues: comparison of five tissue DNA extraction methods using panfungal PCR,” Journal of Clinical Microbiology, vol. 48, no. 6, pp. 2147–2153, 2010. View at: Publisher Site | Google Scholar
  27. V. Rickerts, P. D. Khot, D. Myerson, D. L. Ko, E. Lambrecht, and D. N. Fredricks, “Comparison of quantitative real time PCR with sequencing and ribosomal RNA-FISH for the identification of fungi in formalin fixed, paraffin-embedded tissue specimens,” BMC Infectious Diseases, vol. 11, article 202, 2011. View at: Publisher Site | Google Scholar
  28. H. G. M. Niesters, “Quantitation of viral load using real-time amplification techniques,” Methods, vol. 25, no. 4, pp. 419–429, 2001. View at: Publisher Site | Google Scholar
  29. R. Bialek, A. Feucht, C. Aepinus et al., “Evaluation of two nested PCR assays for detection of Histoplasma capsulatum DNA in human tissue,” Journal of Clinical Microbiology, vol. 40, no. 5, pp. 1644–1647, 2002. View at: Publisher Site | Google Scholar
  30. R. Bialek, F. Konrad, J. Kern et al., “PCR based identification and discrimination of agents of mucormycosis and aspergillosis in paraffin wax embedded tissue,” Journal of Clinical Pathology, vol. 58, no. 11, pp. 1180–1184, 2005. View at: Publisher Site | Google Scholar
  31. Clinical and Laboratory Standards Institute, “Interpretive criteria for identification of bacteria and fungi by DNA target sequencing,” Approved standard MM18-A, 1st Edition, Clinical and Laboratory Standards Institute, Wayne, Pa, USA, 2009. View at: Google Scholar
  32. H. E. Jensen, B. Aalbaek, and H. Schønheyder, “Immunohistochemical identification of aetiological agents of systemic bovine zygomycosis,” Journal of Comparative Pathology, vol. 110, no. 1, pp. 65–77, 1994. View at: Publisher Site | Google Scholar
  33. R. M. da Silva, J. R. da Silva Neto, C. S. Santos et al., “Fluorescent in situ hybridization of pre-incubated blood culture material for the rapid diagnosis of histoplasmosis,” Medical Mycology, vol. 53, no. 2, pp. 160–164, 2015. View at: Publisher Site | Google Scholar

Copyright © 2015 Hagen Frickmann et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

More related articles

 PDF Download Citation Citation
 Download other formatsMore
 Order printed copiesOrder

Related articles

Article of the Year Award: Outstanding research contributions of 2020, as selected by our Chief Editors. Read the winning articles.