Narrow-spectrum histochemical staining of fungi

Archives of Pathology & Laboratory Medicine,  Nov 2003  by Youngberg, George A,  Wallen, Ellen D B,  Giorgadze, Tamar A

• E-mail
• Print
• Link

Although culture is the gold standard for fungal pathogen identification, histologic techniques also play an important role. Histologic methods provide quick results, and material for culture may not be available.

Fungal organisms in tissue sections can sometimes be distinguished simply by morphologic features, such as size, narrow versus broad-based budding, endosporulation, associated hyphae or pseudohyphae, and double-contour cell walls. However, useful morphologic features may not be present, and on occasion the morphology can be misleading. Histoplasma and Paracoccidioides, like Blastomyces, can exhibit prominent cytoplasmic retraction artifact in tissue sections. Size can differ in fungal variants, such as microform (small-strain) blastomycosis or large-form (African) histoplasmosis. Free endospores can be confused with small yeast forms. Several of the dimorphic fungi occasionally show the unexpected presence of hyphae in tissue, while 1 species of Candida does not generally form pseudohyphae. Fungal morphology can also be altered by therapy and possibly by the immunologic status of the host.1 In contrast, immunohistochemical stains and DNA methods are generally highly effective for specific fungal identification, but these methods are not widely available outside a reference laboratory setting. Therefore, other techniques still have a role in the differential diagnosis of fungi.

Broad-spectrum fungal histochemical stains, like Gomori methenamine silver, are used to detect the presence of fungal organisms, regardless of type. However, some commonly available histochemical stains react selectively, staining certain types of fungal organisms but not others. These stains, which may be termed narrow-spectrum fungal stains, are useful for differential diagnosis. The use of narrow-spectrum staining for Cryptococcus is well known and has been described in numerous publications. Mucicarmine stains the capsule of Cryptococcus. Mucicarmine may also stain the cell wall of Blastomyces, but weakly, and mucicarmine does not stain Coccidioides, Histoplasma, or Candida. The Fontana-Masson (FM) stain is potentially even more useful, since mucicarmine may miss capsule-deficient forms of Cryptococcus. The FM stain reacts with a melanin-like compound in the wall of the organism. The FM stain may also stain spherules of Coccidioides, but is reported to not stain Blastomyces, Histoplnsnia, or Candida.2 (In our laboratory, however, a standard FM method has stained a broader group of organisms than what was previously reported in the literature, in spite of careful adherence to a published method and avoidance of overstaining; the specificity of the stain could be increased somewhat by using a modified method with a reduced staining time, albeit with some consequent loss of sensitivity.)

One of the least known but potentially very useful narrow-spectrum stains is Congo red. Congo red reacts with the yeast forms of Blastomyces (as noted by Sheehan and Hrapchak3), which demonstrate the typical salmon coloration, as well as the characteristic greenish birefringence on polarized light examination (Figure 1; eg, organism identified by the arrow). Staining intensity is somewhat variable, but in many cases it may be easier to detect and identify Blastomyces organisms with Congo red than with any other stain. Congo red also has been reported to stain spherules of Coccidioides.2

The periodic acid-Schiff (PAS) stain is used widely as a broad-spectrum stain for fungal detection. It performs almost as well as methenamine silver in this regard, and it demonstrates fungal morphology better than methenamine silver. Furthermore, artifacts that can mimic fungi on PAS are generally different from those that can mimic fungi on methenamine silver, so the use of both stains reduces the incidence of false positives. In some cases, PAS can strongly stain degenerated organisms that are virtually invisible on a hematoxylin-eosin stain. However, it is not widely appreciated that PAS can also function as a narrow-spectrum fungal stain. Histoplasmn cell walls tend to stain poorly with PAS.2 In contrast, other organisms in the differential diagnosis, such as Candida, microform Blastomyces, or the yeast form of Malassezia (Pityrosporum), show reliable and strong cell wall staining with PAS. In the presence of small budding yeast-form organisms, strong cell-wall staining on methenamine silver (Figure 2, A; eg, organism identified by the arrow) coupled with poor staining on PAS (Figure 2, B; eg, organism identified by the arrow) supports a diagnosis of histoplasmosis. Another implication of this is that PAS should never be used as the sole broad-spectrum stain if histoplasmosis is a consideration.

The Ziehl-Neelsen stain is a narrow-spectrum fungal stain. The potential utility of this stain was not fully realized until A. R. Laurain, MD, from our institution, brought its properties to the attention of the Armed Forces Institute of Pathology, where a detailed study was performed.4 Ziehl-Neelsen staining demonstrated some organisms (rare to one third) with positive acid-fast cytoplasmic staining in 60% of blastomycosis cases and 47% of histoplasmosis cases. No acid-fast staining was seen in cases of cryptococcosis or candidiasis. Very rare acid-fast endospores were seen in coccidioidomycosis. Acid-fast cytoplasmic staining of Blastomyces is illustrated. The acid-fast cytoplasmic staining of some of the organisms stands out in sharp contrast to the blue-counterstained, thick, double-contour cell walls (Figure 3; eg, organism identified by the arrow; insets, similar acid-fast organisms from a different case).