OUP user menu

Utility of DNA Sequencing for Direct Identification of Invasive Fungi From Fresh and Formalin-Fixed Specimens

Pablo A. Moncada MD, Indre Budvytiene MS, Dora Y. Ho MD, Stanley C. Deresinski MD, Jose G. Montoya MD, Niaz Banaei MD
DOI: http://dx.doi.org/10.1309/AJCPNSU2SDZD9WPW 203-208 First published online: 1 August 2013

Abstract

Objectives: To describe and discuss the utility and potential pitfalls of ribosomal RNA locus sequencing for direct identification of invasive fungi from fresh and formalin-fixed, paraffin-embedded specimens.

Methods: DNA was extracted from fresh and formalin-fixed, paraffin-embedded tissue and subjected to real-time polymerase chain reaction (PCR) targeting ITS2 and D2 regions of fungal ribosomal RNA locus. Cycle sequencing was performed on PCR products, and the identity of sequences was determined using a public database.

Results: Four clinical cases of invasive fungal infection are presented to illustrate the utility of DNA sequencing for determining etiology when microbiological culture is negative, for shortening the time to identification of slow-growing fungi, for guiding antifungal therapy, and for shedding light on the pathogenesis of disseminated fungal infection.

Conclusions: Fungal ribosomal RNA locus sequencing from fresh or formalin-fixed, paraffin-embedded specimens is a powerful tool for rapid and accurate diagnosis of patients with culture-negative or uncultured invasive mycosis.

Key Words:
  • Fungi
  • Invasive
  • Identification
  • Sequencing
  • Paraffin-embedded

Fungi are saprophytic eukaryotic organisms widely distributed in the environment.1 Except for dimorphic and dematiaceous molds, most fungi do not cause invasive disease in immunocompetent hosts. In fact, several species of yeast are considered normal human commensals.2 Conversely, many fungi behave as opportunistic pathogens in immunocompromised patients, causing rapidly progressing infections. Because invasive fungal infections (IFIs) can be life-threatening, especially in immunocompromised patients, it is critical that an accurate diagnosis be established early in the course of disease so that tailored antifungal therapy can be initiated.3 Conventional identification methods for fungi include microscopic visualization of the fungal element in fresh and/or formalin-fixed, paraffin-embedded (FFPE) specimens and isolation of the organism on fungal medium.4 Although microscopic examination may reveal clues about the identity of the fungus based on its size and shape, as well as the type of immunopathology present, morphologic identification is error prone.5 Definitive identification is accomplished by isolation of the organism on microbiological culture followed by phenotypic and genotypic testing of the isolate.4 In the absence of positive culture, ancillary tests such as serology and antigen assays may be helpful.68 Serology is widely used for the diagnosis of coccidiomycosis and histoplasmosis, but it lacks sensitivity, especially in patients with compromised humoral immune responses.7,8 Antigen tests also lack sensitivity in patients with localized infections as well as specificity due to cross-reactivity with other fungi.6,9,10

Clinicians may often be confronted with culture-negative or uncultured cases of IFI, thus necessitating empiric anti-fungal therapy. The wrong choice of antifungal agent, based on the incorrect interpretation of histopathologic findings, however, may lead to unnecessary morbidity and mortality.5 Therefore, culture-free methods capable of identifying invasive fungi directly from the specimen would be highly valuable. Pathogen-specific polymerase chain reaction (PCR) and in situ hybridization assays have been developed for the direct identification of invasive fungi,4,11,12 but the great phylogenetic diversity of fungi demands diagnostics with broader capabilities.4 Broad-range PCR coupled with amplicon sequencing has allowed direct detection and identification of bacteria from clinical specimens.13,14 In broad-range PCR, primers hybridize to highly conserved sequences and amplify the variable flanking sequences. Comparison of the target sequence with a database of known species allows identification to the species or genus level. The fungal ribosomal RNA operon encoding 5.8S, 18S, and 28S ribosomal subunit genes and the internal transcribed spacer regions (ITS1 and ITS2) is the most reliable and most frequently used locus for fungal identification by sequencing.15 Sequence-based identification of fungi from culture has proven to be more accurate than conventional methods.4,1619

In this case series, we discuss the advantages and limitations of ribosomal RNA locus sequencing for direct identification of fungi from fresh and FFPE specimens.

Materials and Methods

Patients

The patients received their care at the Stanford University Medical Center between 2009 and 2011. Case summaries of patients are presented in Table 1. Approval by the Stanford University Institutional Review Board was not necessary for a small case series.

Fungal Sequencing Assay

Sequence-based identification of invasive fungi from fresh and FFPE specimens was validated in the Clinical Microbiology Laboratory of the Stanford University Medical Center. Testing of clinical samples was requested by the infectious diseases physicians.

View this table:
Table 1

DNA Extraction

For FFPE, five 10-μm sections were cut using a new microtome blade. The QIAamp DNA Mini Kit (QIAGEN, Valencia, CA) and EZ1 DNA Tissue Kit (QIAGEN) were used to extract DNA from fresh and FFPE specimens, respectively, in accordance with the manufacturer’s protocols.

Amplification and Sequencing

Two separate real-time PCR reactions, one targeting 300 base pairs (bp) of ITS2 and another targeting 300 bp of the D2 region of the 28S gene, were performed on each sample. PCR was carried out in a 25-μL reaction mixture consisting of 0.5 μM of fungus-specific universal primers targeting ITS2 (ITS-3F, 5′-GCATCGATGAAGAACGCAGC-3′; ITS-4R, 5′-TCCTCCGCTTATTGATATGC-3′) or D2 (D2F, 5′-GWGACCGATAGCRMACAAGTA-3′; D1D2R, 5′-GCATATCAATAAGCGGAGGA-3′), 1× FastStart SYBR Green Master mix (Roche Applied Science, Indianapolis, IN), and RNase-free water. PCR was performed on a Rotorgene-6000 (QIAGEN). The PCR conditions included 95°C for 5 minutes, followed by 40 cycles of 95°C for 15 seconds, 58°C for 30 seconds, and 72°C for 35 seconds. The final step involved a 60°C to 95°C temperature ramp to generate a melting curve that was used to confirm the presence of amplicon with a melting temperature ranging from 80°C to 92°C. Cycle sequencing was performed with ITS-3F, ITS-4R, D2F, and D1D2R primers as previously described.20 DNA sequences were assembled with Lasergene software (DNASTAR, Madison, WI) and compared with those in the NCBI GenBank (http://www.ncbi.nlm.nih.gov/Genbank/) and CBS (http://www.cbs.knaw.nl/Medical/BioloMICSSequences.aspx). The following distance score criterion was used to identify fungal sequences: less than 1%, to the species level; 1% or more to 2% or less, to the genus level; and more than 2%, not identified. For Aspergillus and Trichophyton, a criterion of less than 0.5% was used to identify to the species level and 0.5% to 1% or less to the species complex level identification. Matching results for ITS2 and D2 regions were required.

Case Reports

Patient 1

A 44-year-old man with cystic fibrosis underwent bilateral lung transplantation in March 2010. His initial recovery was complicated by a sternal wound infection. A debridement specimen from the xiphoid tissue in June 2010 grew Aspergillus versicolor. The wound healed in response to combination therapy with caspofungin and voriconazole, followed by posaconazole. In December 2010, he presented with acute onset of bilateral lower extremity weakness and had extensive aortoiliac thrombosis and aneurysmal degeneration of the ascending aorta with mural thrombus. He underwent thrombectomy of the aorta and the external iliac arteries and stenting of the aortic pseudoaneurysm. The cultures from blood and the resected thrombus did not yield any growth, but histopathology showed numerous fungal elements. Fungal sequencing performed on the FFPE thrombus identified the organism as Aspergillus fumigatus. Given the discrepancy between A versicolor isolated from the sternal wound and A fumigatus identified from the thrombus, fungal sequencing was performed on the archived sternal tissue, and the fungus was identified as A fumigatus. Despite aggressive antifungal therapy, the patient’s condition deteriorated, and he was transitioned to comfort care until death. Autopsy revealed an organized thrombus within the aortic stent that contained abundant fungal hyphae. His lumbar spine also had evidence of discitis and osteomyelitis containing fungal elements.

Patient 2

An 87-year-old white man from the Central Valley, CA, with a medical history of an aortic porcine valve placement in 1998 due to calcified aortic stenosis, presented in April 2009 with a 2-month history of fever, chills, night sweats, and weight loss. Blood and bone marrow cultures for non-fastidious organisms as well as mycobacteria and fungi were negative. Transesophageal echocardiogram showed a normal-appearing aortic porcine heterograft, and full-body computed tomography (CT) was unremarkable. Abnormal laboratory results included elevated alkaline phosphatase and transaminases. Liver biopsy specimens demonstrated granulomas, but no organisms were seen by histopathology, although culture was not performed. Coccidioidomycosis serology testing was negative, but an aberrant immunodiffusion band, representing a Histoplasma “M” band, was detected at the Coccidioidomycosis Serology Laboratory (Davis, CA). Histoplasma urine antigen testing was initially negative, but when repeated at the Mira Vista Laboratory (Indianapolis, IN) it was positive.

Despite antifungal therapy with itraconazole with sequential changes to voriconazole and liposomal amphotericin B, symptoms and laboratory tests normalized only after coadministration of prednisone. Antifungal therapy was discontinued in May 2010, but fever, weight loss, and antigenuria recurred in August 2010. Antifungal therapy with itraconazole was reinitiated, but fever and weight loss again failed to resolve until a short course of prednisone was administered. The patient developed congestive heart failure with moderate to severe aortic regurgitation in the spring of 2011 while still receiving itraconazole. The patient underwent aortic valve replacement in August 2011. Histopathology revealed many yeast forms of variable sizes Image 1A. Given the unusual course of his infection and lack of confirmatory diagnosis by culture, fungal sequencing was performed on the valve and identified the organisms as Ajellomyces capsulatus, the sexual form of Histoplasma capsulatum. Fungal culture of the valve grew H capsulatum after 5 weeks. Liposomal amphotericin B was continued postoperatively, with a transition to itraconazole to be continued indefinitely. The patient has had no evidence of active infection 14 months after valve replacement.

Image 1

Microscopic images of fungal structures in specimens. A, Formalin-fixed, paraffin-embedded excised heart valve with Gomori-methenamine silver stain showing yeast forms (×100). B, Cerebrospinal fluid cytospin preparation with Gram stain showing budding yeast (×100).

Patient 3

A 77-year-old asymptomatic immunocompetent man was noted to have several lung nodules on a chest x-ray screening examination required for enrollment into a retirement community. Subsequent evaluation with CT confirmed the presence of multiple calcified nodules in the both lungs measuring up to 10 mm in diameter. A CT-guided needle biopsy specimen of the nodules revealed necrosis with abundant yeast forms, but fungal culture was negative. A thorough infectious diseases workup was remarkable for positivity of Histoplasma total antibody. Histoplasma urine antigen was negative. To definitively identify the yeast, fungal sequencing was performed on FFPE tissue and identified the organisms as H capsulatum. Because the lung nodules were largely calcified and the patient was asymptomatic, a decision was made not to initiate antifungal treatment but to follow closely with imaging. A chest CT was repeated 3 months later and showed unchanged nodules. No treatment has been administered, and the patient remains asymptomatic to date.

Patient 4

A 27-year-old man with a history of ruptured arterial venous malformations (AVM) underwent emergent placement of an extraventricular drain (EVD) for cerebral hypertension. On hospital day 10, partial resection of the AVM was performed. Two days later, he developed a fever and seizures. The cerebrospinal fluid (CSF) cell count and chemistries were normal. Blood, CSF, and urine cultures were negative. Ceftriaxone was empirically initiated. On hospital day 14, the EVD was replaced because of a high suspicion that it was infected. On hospital day 18, the patient developed a new episode of fever and seizures. Cerebrospinal fluid was xanthochromic, with a leukocyte count of 58 cells/mm3 with 25% neutrophils, 14% monocytes, and 61% lymphocytes; an erythrocyte count of 2,790 cells/mm3; protein of 90 mg/dL; and glucose of 73 mg/dL. Gram stain revealed a moderate number of budding yeast, some demonstrating collarettes Image 1B. The CSF bacterial and fungal cultures remained negative. The patient was started on amphotericin B. Fungal sequencing performed directly on CSF identified the yeast as Malassezia restricta. Antifungal therapy was changed to oral voriconazole. The remaining portion of the AVM was subsequently resected and the EVD was removed. Three weeks after completion of voriconazole, the patient had no evidence of central nervous system infection.

Discussion

Fungal sequencing is a powerful diagnostic tool for direct identification of fungi from fresh and FFPE specimens in patients with culture-negative invasive mycosis. Because the ribosomal RNA locus is in multiple copies (≥100 copies per genome) and contains diverse regions, targeting it maximizes the sensitivity of sequencing for fungal identication.4,21 Several studies have reported on the feasibility and performance of fungal sequencing for direct identification of fungi in specimens from patients with IFI.21,22 Sequencing was successfully completed on 97.3% of specimens when performed on fresh specimens infected with yeast compared with 63.2% to 70% when performed on FFPE specimens infected with various fungi.21,22 Lau et al21 reported successful identification of the fungi in 93.6% (fresh tissue, 96.8%; FFPE, 87.5%) and 64.3% (fresh tissue, 100%; FFPE, 54.5%) of culture-proven and solely histologically proven cases of IFI, respectively. These findings indicate that fresh tissue is superior to FFPE and suggest that not all fungal-like structures found in histopathology sections are real. Introduction of artifacts during staining can result in false-positive histology results. Although the specificity of fungal primer sequences prevents cross-priming of human DNA, contamination of the sample with commensal fungi or environmental spores could yield false-positive results and lead to mismanagement of patients.22

The cases presented here illustrate common clinical scenarios for which application of fungal sequencing can be helpful in illuminating the pathogenesis of IFI and in guiding antifungal therapy. In patient 1, sequencing results from an archived FFPE tissue section shed light on the pathogenesis of disseminated aspergillosis caused by A fumigatus that was incorrectly identified as A versicolor by phenotypic methods. Although the result did not have therapeutic implications in this case because the patient had renal insufficiency, important differences in antifungal susceptibilities have been reported across species of Aspergillus.23,24 In fact, while A fumigatus is generally susceptible to amphotericin B, A versicolor is most often resistant.25 Also, the ability to distinguish Aspergillus terreus or Aspergillus ustus from other Aspergillus species has important therapeutic implications, and thus correct identification of Aspergillus species could affect antifungal therapy and clinical outcome in such cases.26,27 In patient 2, evidence of systemic histoplasmosis had been primarily based on a positive Histoplasma urine antigen test, but the unusual course and extraordinarily slow response to antifungal therapy raised questions regarding possible cross-reactivity of the Histoplasma antigen test with another fungus.6 Thus, direct identification of H capsulatum from the excised prosthetic heart valve definitively established the diagnosis of histoplasmosis. This case also illustrates the advantage of sequencing to shorten the time to identify slow-growing fungi from weeks to days compared with conventional culture techniques.28 Because sequencing does not require the presence of viable organisms, it is also ideal for patients who have received antifungal treatment, as in patient 1, or in situations in which the entire specimen has been preserved in formalin for histopathology examination.13 In patient 3, sequencing from an FFPE biopsy section identified H capsulatum as the cause of pulmonary nodules in an asymptomatic elderly man. Because pulmonary nodules due to H capsulatum may not require treatment in an asymptomatic patient, the sequencing results determined that close follow-up was the most appropriate course in this case.29 In patient 4, the identification of the etiologic agent of EVD-associated meningitis due to M restricta was possible only by fungal sequencing directly from the CSF since this organism is not a known pathogen and does not grow on routine media.30 Sequencing may also be particularly useful for identifying invasive fungi growing in biofilms on foreign objects. Isolation of viable organisms from biofilms may be difficult due to the existence of organisms in nonreplicating states.31

Despite the many advantages of direct fungal identification with ribosomal RNA locus sequencing, infectious diseases practitioners must be aware of its limitations. First, the assay is not widely available because it has been developed and validated by only a limited number of clinical laboratories with expertise in DNA sequencing.32 Variables that are likely to affect the sensitivity and accuracy of the assay include amount of specimen obtained (open vs needle biopsy), fungal burden, fresh vs FFPE tissue, DNA extraction method, sequence targets, and primer selection.4,33 Second, sequence-based identification of fungi directly from specimens is subject to the same limitations that currently exist for identification of fungi from culture. Specific interpretive criteria for identifying fungi by DNA sequencing are currently lacking.4,34 Furthermore, the publicly available sequence database GenBank does not facilitate accurate identification of rare fungi due to incorrect nomenclature assigned to sequence entries, erroneous or truncated sequence entries, and incomplete taxon sampling of fungi-causing disease.19,35 Thus, when possible, sequence-based results should be correlated with phenotypic culture identification results and discrepancies resolved with additional genotypic and phenotypic testing.27 Third, not all fungi, including some clinically important fungi such as A fumigatus species complex and Fusarium solani species complex, can be identified using the ITS region.21 Accurate identification of these fungi requires sequence analysis of alternative genes.4 Fourth, because fungal identification by sequence analysis is highly susceptible to contamination with commensal fungi and environmental spores, it is imperative that testing is strictly limited to samples obtained from sterile sources with fungal elements present on the microscopic examination.21 Fifth, in the absence of positive culture, sequence results must be correlated with histopathology and ancillary test results (serology and antigen tests) to ensure the accuracy of sequence results.

In summary, when used cautiously, fungal ribosomal RNA locus sequencing from fresh or formalin-fixed specimens is a powerful tool for rapid diagnosis and management of patients with culture-negative or uncultured invasive mycosis.

CME/SAM

Upon completion of this activity you will be able to:

  • describe limitations of conventional microbiological methods for detection and identification of fungi in clinical specimens.

  • discuss limitations of existing nucleic acid tests and serological tests for diagnosis of invasive fungal infections.

  • discuss advantages and limitations of fungal identification directly from clinical specimens by DNA sequencing.

  • describe specimen types that are most suitable for fungal identification by DNA sequencing.

The ASCP is accredited by the Accreditation Council for Continuing Medical Education to provide continuing medical education for physicians. The ASCP designates this journal-based CME activity for a maximum of 1 AMA PRA Category 1 CreditTM per article. Physicians should claim only the credit commensurate with the extent of their participation in the activity. This activity qualifies as an American Board of Pathology Maintenance of Certification Part II Self-Assessment Module.

The authors of this article and the planning committee members and staff have no relevant financial relationships with commercial interests to disclose.

Questions appear on p 276. Exam is located at www.ascp.org/ajcpcme.

Acknowledgments

We thank the staff at the Clinical Microbiology Laboratory of Stanford University Medical Center.

Footnotes

  • Funding: This study was supported by the Department of Pathology at Stanford University School of Medicine.

References

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
View Abstract