Epiglottic Histoplasmosis Presenting in a Nonendemic Region / In Reply

To the Editor.-O’Hara et al1 describe an epiglottic mass in a 78- year-old man. The mass was initially considered malignant and was not cultured. Small, intracellular, budding yeast were identified on histologie examination. The patient related a travel history to Africa and Asia, most recently returning from India. A differential diagnosis including Histoplasma capsulatum, microforms of Blastomyces aermatiditis, Candida, and Pneumocystis carinii was developed. A positive H capsulatum immunoglobuHn G (IgG) immunodiffusion test assisted in a final diagnosis of H capsulatum.

In the absence of culture, all other potential etiologic agents must be eliminated. Although I consider the authors’ ultimate diagnosis of H capsulatum as most likely accurate, other yeasts should have been considered. Candida glabrata is a saprophyte mainly encountered in urogenital infections, and other Candida species would probably demonstrate pseudohyphae. Microforms of B dermatiditis are often intermixed with larger forms. Growths of P carinii do not bud. Given this patient’s extensive travel history to Asia, Pnicillium marneffei should have been considered. Kok et al2 have recently described oropharyngeal ulceration caused by P marneffei, which can have an intracellular morphology similar to that of H capsulation. Distinction of P marneffei from H capsulatuni is based on the mechanism of cell division. P marneffei does not bud but divides by fission. The authors’ Gomori methenamine silver stain, demonstrating budding, does not support a diagnosis of P marneffei.

Small, yeastlike inclusions in a pulmonary macrophage in lavage fluid, arrow (Papanicolaou stain, original magnification 1000). Inset demonstrates cell block of lavage fluid demonstrating a similar cell containing yeast with carminophilic capsules, confirming Cryptococcus neoformans (mucicarmine stain, original magnification 1000).

Cryptococcus neoformans can closely simulate H capsulatum in size, and both produce narrow-necked buds. In the Figure, a bronchial lavage from a young patient with acquired immunodeficiency syndrome demonstrates small inclusions (arrow) in a Papanicolaou- stained macrophage. The inset demonstrates capsular staining by mucicarmine, establishing the diagnosis of C neoformans, which was subsequently confirmed by culture. Mucosal lesions by C neoformans can occur and demonstrate similar such cells containing these small, intracellular yeasts. In the case presented by O’Hara et al,1 C neoformans should have been eliminated by attempts to stain its carminophilic capsule or by specific staining of the fungus with the Fontana-Masson stain. The authors’ description of peripheral halos surrounding the yeast should have further prompted consideration of C neoformans.

The use of serology for identification of H capsulatum is of limited value in this case. A single positive immunodiffusion test for Histoplasma IgG cannot discriminate between recent or old infections. Histoplasma IgG would not appear until about 6 to 12 weeks after infection. It is unlikely that the man contracted the Histoplasma infection while visiting India, because the incidence of histoplasmosis in that country is low. However, the primary infection could have occurred during a trip to Africa.

The authors make an important point that fungi can be encountered in nonendemic areas because of the easy availability of travel. Specific identification of fungi is important, particularly because treatment may differ.

GORDON L. LOVE, MD

Department of Pathology, Laboratory 113

VA Outpatient Clinic

Martinez, CA 94553

1. O’Hara CD, Allegretto MW, Taylor GD, Isotalo PA. Epiglottic histoplasmosis presenting in a nonendemic region: a clinical mimic of laryngeal carcinoma. Arch Pathol Lab Med. 2004;128:574-577.

2. Kok I, Weenstra J, Rietra PJ, Dirks-Go S, BIaauwgeers IL, Weigel HM. Disseminated Pnicillium marneffei infection as an imported disease in HIV-I infected patients: description of two cases and review of the literature. Neth J Med. 1994;44:18-22.

In Reply.-We appreciate the interest Dr Love has shown in our recent report describing a 78-year-old retired soil science professor who presented with an eroded epiglottic mass that clinically mimicked a primary carcinoma and that was subsequently determined by tissue microscopy to be epiglottic histoplasmosis.1 Although the clinical presentation of epiglottic histoplasmosis is rare, this case was particularly interesting, because the patient presented in Alberta, Canada, a region where histoplasmosis is not endemic. This case demonstrates that with the mobility of populations, infectious diseases may present in nonendemic regions. Both clinical and laboratory vigilance are required to identify these atypical presentations of infectious pathogens.

This case also illustrates that the specific identification of fungal pathogens may be difficult to establish because of morphologic similarities among some fungal species in tissue sections. Tissue cultures may not be obtained, especially when an infectious etiology is clinically unexpected, as occurred in our case. Our patient’s epiglottic mass consisted of a dense histiocytic infiltrate that contained small, uninucleated yeasts that demonstrated no hyphae or pseudohyphae. In addition to staining with periodic acid-Schiff and Gomori methenamine silver stains, these yeasts were mucicarmine negative, excluding the possibility of Cryptococcus neoformans infection. The differential diagnosis of small yeast forms in tissue sections is extensive, and we completely agree with Dr Love’s differential diagnosis of fungal pathogens for our patient. As described, Blastomyces sp and Candida infections were excluded based on morphologic features.1 The patient’s positive immunodiffusion test for Histoplasma capsiilntum was considered clinically significant, considering his prolonged clinical presentation and that he lived in a nonendemic region where local environmental exposure to H capsulatum was unlikely.

When morphologically similar infectious pathogens are encountered, additional complementary identification techniques, such as direct immunofluorescence, polymerase chain reaction amplification, and in situ hybridization may be useful in establishing a specific diagnosis. Since the report of this case, in situ hybridization for H capsulatum was performed on the patient’s epiglottic biopsy at the Mayo Clinic in Rochester, Minn, using DNA probes designed from variable regions of the 18S and 28S ribosomal RNA genes of H capsulatum.2 The highly specific in situ hybridization technique was strongly positive and confirmed the patient’s original diagnosis of epiglottic histoplasmosis.

CAROLYN D. O’HARA, MD, FRCPC

Dynacare Kasper Medical Laboratories, and Department of Laboratory Medicine and Pathology

University of Alberta

Edmonton, Alberta, Canada

PHILLIP A. ISOTALO, MD, FRCPC

Department of Pathology and Molecular Medicine

Queen’s University

Kingston, Ontario, Canada

1. O’Hara CD, Allegretto MW, Taylor GD, lsotalo PA. Epiglottic histoplasmosis presenting in a nonendemic region: a clinical mimic of laryngeal carcinoma. Arch Pathol Lab Med. 2004;128:574-577.

2. Hayden RT, Qian X, Roberts GD, Lloyd RV. In situ hybridization for the identification of yeastlike organisms in tissue section. Diagn Mo1 Pathol. 2001; 10:15-23.

Copyright College of American Pathologists Jan 2005