Current Concepts in the Epidemiology, Diagnosis, and Management of Histoplasmosis Syndromes

Marwan M. Azar, MD; James L. Loyd, MD; Ryan F. Relich, PhD; L. Joseph Wheat, MD; Chadi A. Hage, MD


Semin Respir Crit Care Med. 2020;41(1):13-30. 

In This Article


Culture, Histopathology, and Cytopathology

The gold standard methods for the diagnosis of histoplasmosis include identification of yeast on histopathology or cytopathology of clinical specimens, and growth of mold in culture incubated at room temperature (Table 1). Histopathologic examination demonstrates a granulomatous response that is either necrotizing or nonnecrotizing (Figure 1A, B) as well as small narrow-budding yeast cells morphologically consistent with H. capsulatum. The differential for histoplasmosis includes other granulomatous disorders such as sarcoidosis, lymphoma, other fungal infections, and mycobacterial infections. Microscopic examination reveals ovoid, narrow-based budding yeast within tissues, often found phagocytosed within macrophages (Figure 2). Free yeast cells may also be observable, especially in many body fluids and bronchoalveolar lavage (BAL) fluid. Yeast cells are generally small (2–4 μm in greatest dimension) while those of H. capsulatum var. duboisii are larger (~6–12 μm in greatest dimension). Unlike Coccidioides immitis/posadasii, identification of H. capsulatum cannot be definitively based upon microscopic observations alone, as other organisms, including Candida glabrata, Leishmania spp. amastigotes, and several others can be indistinguishable (Figure 3). Table 2 lists some look-alike organisms and their attributes, which can be used to help distinguish them from H. capsulatum yeast cells. Primary microbiologic detection is highly specific. In disseminated disease, fungus can be grown from extrapulmonary tissues and blood. The mold form produces a white, fluffy mycelium that can take up to 6 weeks to grow in culture (Figure 4A) but usually grows in 2 to 4 weeks. Microscopic examination of growth reveals septate hyphae that, in young cultures, bear pear-shaped microconidia; however, thick-walled, spiny macroconidia (Figure 4B) are readily observable in older cultures. Although culture remains the gold standard for the diagnosis of histoplasmosis, it requires several weeks, which can lead to delays in patient treatment. Cytology and histopathology, by contrast, can lead to faster diagnosis but both sensitivity and specificity are lower. Flexible bronchoscopy and transbronchial lung biopsy may be required to collect tissue specimens for cultures and pathology though thrombocytopenia or critical illness may present barriers to invasive procedures. Soft clues that may point to the diagnosis of severe histoplasmosis include elevated serum alkaline phosphatase, lactate dehydrogenase, or ferritin levels and the presence of hepatosplenomegaly, and/or mediastinal or hilar adenopathy on imaging.[40] In a recent study, an aspartate aminotransferase/alanine transaminase (AST/ALT) ratio was significantly higher for disseminated histoplasmosis (2.69) than disseminated coccidioidomycosis or blastomycosis (≤1.14) and bacterial sepsis (0.89), serving as a possible early indicator prior to resulting of confirmatory testing.[41]

Figure 1.

Caseating (A) and noncaseating (B) granulomas, which are typically observed in tissues from patients with histoplasmosis. Original magnifications: ×100 (A) and ×200 (B).

Figure 2.

Tinctorial and morphological attributes of Histoplasma capsulatum in stained clinical specimens. H. capsulatum yeast cells are small (generally, 2–4 μm in greatest dimension), ovoid, and budding occurs on a narrow base. (A) Diff-Quik; BAL fluid smear showing extracellular yeast forms. (B) Giemsa stain; touch preparation of an aortic mass demonstrating numerous yeast forms within a histiocyte and in the surrounding space. (C) GMS stain; abundant black-colored yeast cells in a touch preparation of an aortic mass. Both extracellular and intra-histiocytic forms are seen. (D) Gram stain; red-colored yeast cells in a blood culture smear. (E) H&E stain; liver core biopsy containing numerous intracellular and extracellular yeast forms. Note the presence of colorless halos surrounding the yeast, which represents an artifact of specimen processing and not true encapsulation. (F) Mucicarmine stain; yeast forms are barely visible without the aid of increased contrast. (G) Periodic-acid Schiff stain; magenta-colored yeast forms are evident scattered throughout the epidermis of a skin biopsy specimen. (H) Wright-Giemsa; yeast forms evident within a monocyte in a peripheral blood smear. Scale bar, 10 μm; original magnification, ×1,000. BAL, bronchoalveolar lavage.

Figure 3.

Microorganisms commonly confused with H. capsulatum yeast forms in clinical specimens. (A) Blastomyces dermatitidis; H&E-stained tissue section showing broad-based budding yeast cells with thick, refractile walls. (B) Candida glabrata; Gram-stained BAL fluid smear containing purple-colored yeast cells among numerous red-staining polymorphonuclear neutrophils. Recall that H. capsulatum stains red or fuchsia in properly stained Gram stains. (C) Cryptococcus neoformans; mucicarmine-stained BAL fluid smear demonstrating several red-colored encapsulated yeasts. The polysaccharide-laden capsule of encapsulated strains of both C. neoformans and Cryptococcus gattii stains magenta with mucin stains, but the cell wall and the cytoplasmic compartment are visible within the stained capsule. (D) Leishmania spp.; Giemsa-stained touch preparation from an infected tissue permits visualization of numerous amastigotes both within and outside of host cells. A distinguishing feature of Leishmania spp. amastigotes is the presence of a kinetoplast (arrow heads) in close proximity to the parasite nuclear material. (E) Pneumocystis jirovecii; GMS-stained BAL fluid containing several black-colored cyst forms. The cyst forms of this fungus are often described as looking like crushed ping-pong balls and have prominent, central cyst wall thickenings (darker stained objects within cysts). (F) Toxoplasma gondii; H&E-stained tissue section showing a tissue cyst containing numerous bradyzoites. T. gondii bradyzoites often lack the defined cellular margin seen in H. capsulatum and T. gondii does not stain positive by GMS or PAS. (G) Trypanosoma cruzi; amastigotes of T. cruzi in an H&E-stained heart tissue section. Like Leishmania spp. and T. gondii, T. cruzi will not stain positive in GMS or PAS-stained preparations. Scale bar, 10 μm; original magnification, ×1,000. BAL, bronchoalveolar lavage.

Figure 4.

Colonial and microscopic morphologies of H. capsulatum mold form. (A) Off-white, cottony mycelia growing on Mycosel Agar after ~2 weeks of incubation at 28°C in an ambient atmosphere. (B) Lactophenol cotton blue-stained cellulose-tape preparation demonstrating large, rounded, tuberculate macroconidia. Original magnification, ×400.

Antigen Detection

Although isolation of mold in culture and demonstration of yeast on histopathology remain the gold standards for the diagnosis of histoplasmosis, variable culture yield and contraindications to biopsy can limit their clinical utility in real-world settings. Antigen detection represents a valuable diagnostic tool that provides rapid and noninvasive evidence of acute, chronic, and disseminated histoplasmosis. First developed in 1986, Histoplasma antigen testing underwent several modifications leading to its most recent iteration as a third-generation test to allow for quantification.[42] The assay performance is highest when performed on urine specimens and is contingent on the presenting clinical syndrome of histoplasmosis (Table 1). Sensitivity is high for chronic and acute histoplasmosis (88 and 83%), but lower for the subacute form (30%) and mediastinal histoplasmosis where antigen testing is most often negative.[43] Sensitivity is also increased in immunocompromised patients including HIV/AIDS,[44] treatment with tumor necrosis factor (TNF)-α inhibitors,[45] or immunosuppression following solid organ transplantation,[46] and those with severe disease because of a higher burden of infection.[44–48] In a multicenter retrospective review of 158 patients with disseminated histoplasmosis, the sensitivity of Histoplasma urine antigen was 91.8%.[48]

Notably, detection of antigenemia is slightly less sensitive than detection of antigenuria but combining both urine and serum antigen testing results in improved diagnostic performance, with a sensitivity exceeding 80% in acute pulmonary histoplasmosis[49] and 90% in disseminated histoplasmosis.[48]

Antigen detection may be useful on BAL fluid where the sensitivity has ranged from 70 to 93%.[50,51] In one study, the sensitivity of BAL Histoplasma antigen was superior to those of both serum and urine antigen and was the only evidence of antigen positivity in 10% of the sample.[51] In another study, the sensitivity of antigen testing increased to 96.8% when combined with BAL cytopathology.[51] If antigen testing on urine or serum is negative but the pretest probability for histoplasmosis is high, a bronchoscopy with BAL and transbronchial biopsy should be considered.

Detection of antigen in cerebrospinal fluid (CSF) is useful for diagnosis of Histoplasma meningitis. In a recent study, CSF antigen was sensitive for Histoplasma meningitis in immunocompromised patients and those with severe disease (93 and 81%, respectively) but less so in normal hosts and those with mild–moderate disease (31 and 55%, respectively).[52] Other sterile body fluids in which antigen has been detected include pleural, pericardial, and peritoneal fluids (L.J. Wheat, unpublished).

Monitoring of antigen levels is a useful measure of treatment response since they decline with effective treatment and increase with relapsed infection. In a study of patients with HIV/AIDS undergoing treatment for disseminated histoplasmosis, serum antigen levels declined significantly within 2 weeks of treatment initiation, providing an early marker for treatment response.[53] In this study, amphotericin was associated with a faster decline in antigenemia than itraconazole but the latter was not associated with treatment failure. To effectively monitor response to antifungal therapy, antigen levels should be obtained at treatment onset, at 2 weeks, 4 weeks, then at 3-month intervals during the first year of therapy, prior to treatment discontinuation, at 6-months intervals for 1 to 2 years after treatment completion, and in the event of any symptom recurrence that suggests infection relapse. During monitoring of antigen levels, a lack of appropriate decrease or an increase is concerning for treatment failure. Reasons for treatment nonresponse include nonadherence to therapy or subtherapeutic antifungal levels from malabsorption or rarely, drug resistance.

Histoplasma antigen testing suffers from significant cross-reactivity with other endemic fungi with rates of up to 90% for Blastomyces dermatitidis, 100% for H. capsulatum var. duboisii, up to 67% for C. immitis/posadasii, up to 90% for Paracoccidioides brasiliensis, 100% for Sporothrix schenckii, and up to 94% for Talaromyces marneffei.[43] False-positives have been described in 15% of transplant patients receiving antithymocyte globulin.[54,55] However, in an endemic area, up to 75% of low positive results are clinically significant.[56]


Antibodies to H. capsulatum are produced 4 to 8 weeks after acute infection and may persist for years, limiting the utility for the diagnosis of acute histoplasmosis syndromes. Serology is most useful for the diagnosis of subacute, chronic, and mediastinal forms of histoplasmosis for which the sensitivity of antigen testing may be poor (Table 1). Although a majority of adults living in endemic areas are exposed to H. capsulatum at some point in their lifetimes, only a small percentage is positive for antibodies using commercial methods. In a study in Indianapolis, 55% of adults were skin test positive but only less than 4% were positive using commercial serologic methods.[57]

There are three currently available methods to detect H. capsulatum-specific antibodies including complement fixation (CF), immunodiffusion (ID), and enzyme immunoassay (EIA). The CF method is resulted as a titer, based on the extent of CF to antibody–antigen complexes. A fourfold rise in CF antibody titer from acute to convalescent sera and a single CF titer of 1:32 are highly suggestive of recent acute infection. In the event of negative CF testing, convalescent sera should be obtained 4 to 8 weeks after suspected infection to assess for seroconversion. Low positive titers occur in one-third of patients with active histoplasmosis. The ID assay detects the presence of H and M H. capsulatum antigens by precipitation onto agar gel. Presence of an M band is more common and indicates acute or remote infection. The M band persists for many years after acute infection so cannot distinguish between recent and remote infection. Detection of an H band is uncommon but confirms acute infection when present. The CF assay is more sensitive but less specific than ID (90 vs. 80% and 90 vs. 99%).[57] If certain criteria are met, serology provides evidence of acute infection. These include a fourfold rise in CF titers taken at least 2 weeks apart, detection of H band on ID assay, detection of M band by ID test after documented lack of M band on previous test, and detection of antibodies by single CF titer of ≥1:32.[58] Serologic tests on CSF may be positive in 80% of cases of CNS histoplasmosis. A probable diagnosis of meningitis can be made by detecting a CSF CF titer of ≥1:32 or an M band by CSF ID test. Using both ID and CF methods increases the overall sensitivity of detection. A recently described EIA-based method that permits detection of immunoglobulin M (IgM) and IgG antibodies separately demonstrated improved sensitivities over CF and ID assays. Combining serology and antigen testing may lead to improved sensitivity for acute pulmonary histoplasmosis.[59]

A high level of cross-reactivity occurs with other endemic fungal infections including paracoccidioidomycosis, coccidioidomycosis, and blastomycosis. Due to impaired antibody responses, serology is poorly sensitive in immunosuppressed patients. Seropositivity by ID and CF assays for immunocompromised patients was 63.2 and 70.3% respectively versus 85.7 and 77.8% in immunocompetent patients in a multicenter retrospective analysis.[28] The sensitivity is particularly low in recipients of solid organ transplantation (~30%) as compared with patients with AIDS[44,48] or those receiving TNF inhibitors.[45]

Molecular and Other Diagnostics

While molecular methods have become the de-facto gold standards for the diagnosis of many pathogens, H. capsulatum remains a notable exception. Polymerase chain reaction (PCR)-based testing offers many potential advantages including high sensitivity and specificity, rapid turnaround times, and less susceptibility to host factors. However, the sensitivity of laboratory-developed PCRs in published reports has ranged widely (33–100%) in comparison to culture, antigen testing, serology, and clinical criteria.[43] In addition, there are currently no FDA-approved PCR-based assays for histoplasmosis. Although consistency is problematic at this time, the diagnostic potential is clear. In one study comparing real-time PCR to culture, the majority of patients (10/11) with PCR-positive/culture-negative samples were confirmed to have histoplasmosis based on histopathology or cultures from other specimens.[60] With increased method standardization, PCR assays on direct specimens may occupy a larger role in the diagnosis of histoplasmosis. Broad-range PCR testing of internal transcribed spacer and ribosomal sequences is offered at some commercial laboratories though routine testing is low yield and costly.[61] Molecular diagnosis using rapid DNA probes is currently used for culture confirmation of suspected H. capsulatum colonies.[62] Other diagnostics for histoplasmosis are on the horizon. Next-generation metagenomic sequencing methods with the ability to detect H. capsulatum directly in clinical specimens are available and have been successfully used for diagnosis of histoplasmosis,[63] but performance characteristics are as yet unknown. A H. capsulatum interferon gamma release assay demonstrated high sensitivity and specificity, allowing discrimination between latently infected and noninfected individuals.[64] A matrix-assisted laser desorption/ionization time of flight mass spectrometry (MALDI-TOF MS) reference database for the identification of H. capsulatum has been developed; MALDI-TOF MS could reduce time to identification and also decrease risk to laboratory personnel cultivating the fungus.[65]