69 Histoplasmosis

Histoplasma was initially described from a lesion in a horse by Rivolta in 1873, who named the organism Cryptococcus farciminosum. In 1905, Samuel Darling noted the presence of intracellular organisms in many tissues, including the lungs, of a patient suspected of succumbing to miliary tuberculosis (Darling 1906). Darling named the organism Histoplasma capsulatum, because it appeared to be an encapsulated protozoan-like organism. In 1912, mycologist Henrique da Rocha-Lima reviewed Darling’s slides and noted the cytological similarities between Darling’s Histoplasma organism and Cryptococcus farciminosum. Cryptococcus farciminosum was reclassified as Histoplasma farciminosum in 1934, and in 1985 it was again reclassified as a variant of Histoplasma capsulatum (var. farciminosum) (Weeks et al. 1985).

William De Monbreun cultured the organism from the blood of a child suffering from an unexplained febrile disease in 1934, and demonstrated it to be a dimorphic fungus (De Monbreun 1934). De Monbreun and others reported naturally occurring histoplasmosis in a dog in 1939, and subsequently demonstrated experimentally that clinically inapparent histoplasmosis occurred in dogs (De Monbreun 1939). De Monbreun and others speculated that animals might serve as the source of histoplasmosis in human beings. However, C.W. Emmons demonstrated in 1949 that Histoplasma capsulatum is a soil saprophyte, and that inhalation of aerosolized microconidia and mycelial fragments served as the source of infection (Emmons 1949).

The prevalence of histoplasmosis in endemic regions was estimated to be more than 50% based on positive skin tests for histoplasmin (Edwards et al. 1969). Active histoplasmosis has been identified in up to 50% of dogs in endemic regions based on culture at necropsy of healthy animals (Turner et al. 1972a). The case prevalence of disseminated histoplasmosis at a veterinary teaching hospital in an endemic region of the mid-western USA of 43 cases in cats and 12 cases in dogs per 100,000 hospital records per year has been estimated (Clinkenbeard et al. 1988; Kaplan 1973). Dogs and cats with outdoor exposure are reportedly at greater risk for histoplasmosis than those with minimal time outdoors. However, some completely indoor cats become ill with histoplasmosis (Davies and Troy 1996; Johnson et al. 2004). Young to middle-aged dogs of hunting and sporting breeds have historically been reported at greatest risk for acquiring histoplasmosis (Selby et al. 1981). Risk factors for cats have not been systematically studied.

Infection by Histoplasma capsulatum var. capsulatum is not contagious except in unusual situations. Rare cases of horizontal transmission have been reported. Horizontal transmission is associated with conjugal contraction of individuals with cutaneous lesions of the genitalia (Sills et al. 1973) and by solid organ transplantation of infected organs (Limaye et al. 2000). No documented cases of transmission from animals to human beings or vice versa have been reported. In contrast to Histoplasma capsulatum var. capsulatum, equine infection by Histoplasma capsulatum var. farciminosum is contagious and is transmitted by bites of contaminated flies or ticks as well as through skin traumatized with contaminated tack (Kohn 2006).

Inhalation of microconidia and mycelial fragments is the route of natural Histoplasma capsulatum var. capsulatum infection. Once inhaled the conidia are engulfed by neutrophils, macrophages and dendritic cells, in which they proliferate until Th1 immunity develops. Over the 3–5 days the conidia transform into yeasts, which are the pathogenic form of the organism (Procknow et al. 1960). The yeasts multiply within macrophages, which spread the infection to extrapulmonary sites.

During the second week of infection cell-mediated immunity develops and halts progression in man and murine experimental models of histoplasmosis. Th1 lymhocytes and several cytokines are responsible for an effective immune response, which results in death of the organism. Interleukin-12, IL-18, tumour necrosis factor-α (TNF-α) and interferon-gamma play important roles in acquired immunity in histoplasmosis.

Progressive disseminated infection in humans is associated with conditions that reduce cell-mediated immunity (Assi et al. 2007; Wheat et al. 1982). Key among these are HIV infection, use of corticosteroids and other immunosuppressive medications, including tumour necrosis factor inhibitors, and rare immunodeficiency states (Steiner et al. 2009; Zerbe and Holland 2005). However, individuals of certain animal species, including dogs (Fattal et al. 1961; Rowley et al. 1954; Smith et al. 1976) cats (Rowley et al. 1954) and bats (Greer and McMurray 1981; Tesh and  Schneidau 1967) fail to eradicate the infection in the absence of immunosuppression, suggesting inherent inability to develop immunity to H. capsulatum. Except for studies in murine and guinea pig models of histoplasmosis, immunity has not been studied extensively in animals.

Both natural and experimental histoplasmosis have been described in a variety of animals. While it has been possible to infect chickens, causing a localized infections of the feather (Tewari and Campbell 1965), histoplasmosis does not occur naturally in birds because of their higher body temperature. The natural disease caused by Histoplasma capsulatum var. capsulatum is most adequately described in dogs and cats. The prevalence of histoplasmosis is high in the endemic areas, as is indicated by the high percentage of dogs and cats with positive cultures from lungs or associated lymph nodes (Emmons et al. 1955; Emmons and Rowley 1955; Fattal et al. 1961; Turner et al. 1972a). The incubation period is approximately 7–14 days (Menges et al. 1954; Ward et al. 1979).

Most affected cats have disseminated histoplasmosis, a systemic disease, with clinical signs of weight loss, lethargy, fever, anorexia, and weakness (Clinkenbeard et al. 1987; Davies and Troy 1996). A review of medical records of 96 cats with histoplasmosis showed that, in contrast to previous reports, many cats with histoplasmosis are not presented for respiratory signs. In that study, the most common clinical signs of disease were lethargy, weakness, dehydration, pyrexia, and emaciation (67% of the cats). Thirty-nine percent of the cats had respiratory signs, 24% had ocular signs (chorioretinitis, anterior uveitis, retinal detachments), and 18% had evidence of skeletal involvement (Davies and Troy 1996). Skin lesions, oral ulcers, and diarrhoea are less common signs arising from focal tissue or organ dysfunction associated with disseminated histoplasmosis; mineralized pulmonary lesions are not commonly reported in cats (Clinkenbeard et al. 1987; Lamm et al. 2009; Pearce et al. 2007; Stark 1982; Vinayak et al. 2007). Bone marrow involvement is sometimes the only manifestation of histoplasmosis in cats. Thrombocytopenia, neutropenia, anemia, and mixed cytopenias are reported; evaluation of bone marrow aspirates or biopsies may be necessary for diagnosis (Clinkenbeard et al. 1987; Davies and Troy 1996; Gabbert et al. 1982; Hodges et al. 1994; Kerl 2003).

In dogs, pulmonary histoplasmosis can resolve with no sequelae, with mineralization of interstitial lung nodules and tracheobronchial lymph nodes, or with development of clinically apparent disseminated histoplasmosis (Burk et al. 1978; Schulman et al. 1999). Disseminated histoplasmosis is typically a subacute to chronic diarrhoeal disease of young to middle-aged dogs (Bromel and Sykes 2005; Clinkenbeard et al. 1989; Gingerich and Guptill 2008; Krohne 2000). Weight loss and anaemia are also common signs, fever, lymphadenopathy, hepatomegaly, splenomegaly, and respiratory signs are seen in up to 50% of dogs with systemic histoplasmosis. Focal signs or lesions other than diarrhoea are reported in fewer dogs with disseminated histoplasmosis, and these may include icterus, lameness, vomiting, oral ulcers, ocular lesions including retinal detachment and subretinal pyogranulomas which may lead to blindness, CNS signs, and rarely, skin lesions (Gingerich and Guptill 2008; Kabli et al. 1986; Kagawa et al. 1998; Krohne 2000; Nishifuji et al. 2005; Olson and Wowk 1981; Salfelder et al. 1965).

The clinical findings are generally similar, except that humans are more likely to have underlying conditions or be receiving medications that suppress their immunity, predisposing to progressive histoplasmosis. The clinical spectrum in humans may be somewhat broader than in animals, in part because of discovery of radiographic abnormalities as incidental findings during evaluation for other conditions. Also, manifestations causing mild symptoms, which are commonly noted in humans, may not be recognized in animals. The more common findings in humans include pulmonary or mediastinal manifestations and progressive disseminated disease.

In endemic areas, half to more than 80% of individuals have had histoplasmosis by age 20 (Edwards et al. 1969), and most are asymptomatic. About 5% of individuals develop a subacute pulmonary illness after low level exposure (Wheat 1989), and radiographs show enlarged hilar or mediastinal lymph nodes with patchy pulmonary infiltrates (Wheat et al. 1982). The illness usually is mild and evolves over several weeks to months before a diagnosis is established. Improvement occurs within a month (Brodsky et al. 1973), but fatigue may linger.

Pericarditis may occur as an immunological reaction to the adjacent and inflammed mediastinal nodes (Wheat et al. 1983). Outcome is excellent, with rare progression to constrictive pericarditis (Bilgi and Slesar 1976). Antifungal therapy is unnecessary (Picardi et al. 1976) unless the patient receives corticosteroids or has disseminated disease. Arthralga or arthritis, also occur as immunological reactions to the acute infection, and usually are associated with pulmonary complaints. The joint symptoms usually resolve in response to anti-inflammatory therapy (Medeiros et al. 1966).

An acute pulmonary illness follows heavy exposure. Patients usually present within two weeks of exposure with diffuse pulmonary involvement often causing respiratory difficulty (Loosli et al. 1952). Attack rates exceed 75% follow heavy exposure (Gustafson et al. 1981; Ward et al. 1979). Chest radiograms show diffuse reticulonodular or miliary pulmonary infiltrates, sometimes with mediastinal lymphadenopathy (Johnson et al. 1988). Some patients may manifest progressive extra pulmonary dissemination (Rubin et al. 1959). Although patients may recover without treatment (Rubin et al. 1959), the illness often is severe and recovery slow. Thus, treatment is advised. Adjunctive corticosteroid treatment may accelerate improvement in such cases, as the inflammatory response may contribute to the pathogenesis of the respiratory injury.

Chronic pulmonary histoplasmosis occurs in patients with underlying emphysema, and is characterized by cavitary upper lobe infiltrates similar to those typical of tuberculosis (Goodwin et al. 1976). The illness is chronic and progressive if left untreated.

Enlarged mediastinal lymph nodes may impinge upon the airways, pulmonary vessels or vena cava, or the oesophagus, occurring in < 10% of patients with pulmonary histoplasmosis (Coss et al. 1987). These findings may first present years after the initial infection, as a result of smoldering inflammation and necrosis in the involved node. Symptoms include chest pain, cough, haemoptysis, dyspnea and dysphagia (Schowengerdt et al. 1969). Although enlarged nodes usually shrink and symptoms resolve without treatment (Sakulsky et al. 1967), obstructive syndromes may be severe (Greenwood and Holland 1972), and the masses may persist for years.

Fibrosing mediastinitis is a fibrotic response to a prior episode of mediastinal histoplasmosis that is characterized by invasion and obstruction of mediastinal structures (Goodwin et al. 1972). The pathogenesis is thought to involve an excessive fibrotic response to Histoplasma antigens released from mediastinal lymph nodes. Obstruction may involve the superior vena cava, airways, pulmonary arteries or veins, or oesophagus. While the clinical impairments are chronic, they are not always progressive, and they do not respond to antifungal therapy. Surgery is associated with a high mortality, and is rarely indicated. Some patients may benefit from placement of stents in pulmonary or bronchial vessels, however.

Progressive disseminated histoplasmosis occurs mostly in immunodeficient patients and those at the extremes of age (Goodwin et al. 1980), and is characterized by a progressive infection involving extrapulmonary tissues. Fever and weight loss are the most common findings, accompanied by respiratory symptoms in most patients (Goodwin et al. 1980). Hepatomegaly, splenomegaly or lymphadenopathy occur in one to two-thirds of cases and, less frequently, adrenal, gastrointestinal, skin or mucosal lesions may be seen. The brain, spinal cord or meninges are involved in about 5% of cases (Wheat et al. 2005). Endocarditis is a rare complication of disseminated histoplasmosis (Bhatti et al. 2005). Any tissue may be involved. While the illness is progressive and ultimately fatal in 90% of cases (Sathapatayavongs et al. 1983), the untreated course may span a few months to several years, and some patients appear to recover spontaneously (Assi et al. 2007).

Several tests are useful for diagnosis of histoplasmosis in humans, including serology, fungal stain of tissues, detection of antigen in body fluids, and fungal culture (Wheat 2009). In humans with acute pulmonary and progressive disseminated histoplasmosis the highest sensitivity is with antigen detection, which is most useful in patients with acute or severe disease. Serology is most useful in mild cases. Diagnosis in veterinary medicine is currently primarily based on detection of Histoplasma organisms upon cytologic or histopathologic evaluation of tissue aspirates, biopsies, or fluid obtained by tracheal wash or bronchoalveolar lavage (Figs. 69.1 and 69.2). Fungal culture of tissues or fluids is also useful. Polymerase chain reaction (PCR) testing may be applied to tissue samples, and fungal stains are also used when evaluating tissue samples (Bromel and Sykes 2005; Gingerich and Guptill 2008; Greene 2006; Nishifuji et al. 2005; Ueda et al. 2003). Serologic testing for antibody detection is not highly sensitive or specific for veterinary patients with focal or disseminated disease. Testing of serum or urine for the presence of fungal antigen has recently been introduced, but is not yet validated for veterinary patients.

Histoplasma yeast measure 2–3 µm in diameter, and exhibit narrow-necked budding. In humans, fungal stain is less sensitive than antigen detection for diagnosis of disseminated and acute pulmonary histoplasmosis (Wheat 2009). One drawback of fungal stain is the requirement to perform invasive procedures to obtain specimens for evaluation. Also, the accuracy may vary depending upon the pathologist’s experience with recognition of Histoplasma yeast, as other yeast or staining artefact may be mistaken as Histoplasma and small numbers of yeasts may be easily overlooked.

Cultures are positive in many cases of disseminated histoplasmosis (Fattal et al. 1961; Turner et al. 1972b; Sathapatayavongs et al. 1983), but growth may be slow, requiring up to four weeks. The highest yield is from the lung, skin or mucosal lesions, or bone marrow, thus requiring invasive procedures to obtain specimens. In many cases in animals, cultures are not performed, however.

A galactomannan antigen in the cell wall of proliferating Histoplasma yeast is released into the tissues and blood, and excreted in the urine. This antigen can be detected in an enzyme immunoassay. Antigen was detected in 95–99% of humans with disseminated histoplasmosis (Connolly et al. 2007), and in 83% of those with acute pneumonia (Swartzentruber et al. 2009). The greatest sensitivity for diagnosis required testing both urine and serum (Wheat 2009). Also, a negative result does not exclude the diagnosis. In cases with negative results, follow-up specimens may be positive in patients exhibiting progressive illness. Serology or culture may be positive in patients with negative antigen results. Antigen detection was sensitive when used to test stored samples from a bottlenose dolphin with disseminated histoplasmosis (Jensen et al. 1998). Prospective studies are needed to define sensitivity of antigen detection in animals.

Antigen may be detected in the respiratory secretions in humans with pulmonary histoplasmosis (Hage et al. 2010), occasionally permitting diagnosis in patients with negative results in urine and serum. Antigen also may be detected in the cerebrospinal fluid of humans with meningitis, offering a helpful method to diagnose this elusive manifestation (Wheat et al. 2005).

The galactomannan found in histoplasmosis cross reacts with that found in blastomycosis (Spector et al. 2008; Connolly et al. 2007). Furthermore, the clinical findings and endemic distribution overlap. Thus, differentiation of the two mycoses may be difficult via antigen detection alone.

Antigen levels decline during treatment and increase with relapse, providing a tool for monitoring therapy in humans. Again, the use of antigen detection for diagnosis or monitoring of histoplasmosis in animals requires further validation.

In humans, serologic tests are positive in over 90% of cases of subacute and chronic pulmonary histoplasmosis and often provide the basis for diagnosis (Wheat 2009). Serology is less useful in disseminated and acute pulmonary histoplasmosis, where tests may be falsely negative due to underlying immunosuppression in disseminated disease and because of the one to two month delay for antibodies to develop following acute infection.

The lack of sensitivity and specificity of available serologic tests for diagnosis of active histoplasmosis in animals is highlighted in reports of canine and feline histoplasmosis. Seventeen of 26 animals with culture, cytology, PCR, or necropsy-proven histoplasmosis had negative serologic test results (agar-gel immunodiffusion or complement fixation tests) (Clinkenbeard et al. 1987; Clinkenbeard et al. 1988; Hawkins and DeNicola 1990; Hodges et al. 1994; Johnson et al. 2004; Kowalewich et al. 1993; Mackie et al. 1997; Mitchell and Stark 1980; Nishifuji et al. 2005; Noxon et al. 1982; Olson and Wowk 1981). In those reports, some animals with positive test results also had positive test results for other fungal pathogens; one animal with blastomycosis had a negative serologic test result for blastomycosis and a positive result for histoplasmosis, and two animals with negative tests for histoplasmosis had positive test results for other fungal pathogens. In a study evaluating naturally infected dogs, 51 dogs had positive Histoplasma cultures; 13 of those had positive serology for Histoplasma antigens, 10 were positive for Blastomyces, and 28 were positive for either Blastomyces or Histoplasma or both. Of 132 dogs with negative serologic tests, 23.5% had positive fungal cultures for Histoplasma, and of 125 dogs with positive complement fixation tests (using both Blastomyces and Histoplasma antigens), only 49% had positive fungal cultures for Histoplasma (Turner et al. 1972b).

Cross reactions by complement fixation may occur in patients with other endemic mycoses, and serum may be anti-complementary, preventing measurement of antibodies by complement fixation in some cases. H and M precipitin bands are specific for histoplasmosis and A precipitin bands for blastomycosis, assisting in differentiation of the two mycoses. However, the sensitivity of immunodiffusion is low (<20%) in dogs with blastomycosis (Spector et al. 2008) and <35% in dogs with histoplasmosis (Clinkenbeard et al. 1987, 1988; Hawkins and DeNicola 1990; Hodges et al. 1994; Johnson et al. 2004; Kowalewich et al. 1993; Mackie et al. 1997; Mitchell and Stark 1980; Nishifuji et al. 2005; Olson and Wowk 1981). Sensitivity may be improved using enzyme immunoassay, as described in blastomycosis, but the enzyme immunoassay may not be specific (Greene 2006; Spector et al. 2008).

Molecular methods have not been tested extensively for diagnosis of histoplasmosis in humans or animals. Several PCR targets have been utilized, including the 18S ribosomal RNA gene, the rRNA internal transcribed spacer region, M and H antigens, and a 100 kDa protein, HC100, that is necessary for intracellular survival (Bialek et al. 2001, 2002; Bracca et al. 2003; Imhof et al. 2003; Sandhu et al. 1995; Tang et al. 2006). Specificity and sensitivity of these assays has not been extensively evaluated, however. PCR may be falsely-negative in specimens in which the organisms are seen by histopathology (Bialek 2002). More recently PCR assay was found to be less sensitive than antigen detection when applied to urine (Tang et al. 2006) or other body fluids (Wheat, unpublished observation, 2009). There are currently no commercially available PCR assays for diagnosis of histoplasmosis (Kauffman 2009). Molecular methods have not been widely used for diagnosis of histoplasmosis in dogs and cats (Nishifuji et al. 2005; Ueda et al. 2003).

Amphotericin B is recommended for the first week or two in severe infections of small animals (Bromel and Sykes 2005; Gingerich and Guptill 2008; Grooters and Taboada 2003; Kerl 2003), as in humans (Wheat et al. 2007). Amphotericin B is recommended in humans who have more severe forms of histoplasmosis, requiring hospitalization. Liposomal amphotericin B (AmBisome) was more effective than the standard deoxycholate formulation, demonstrating a higher response rate (88% vs. 64%, respectively) and lower mortality (2% vs. 13%, respectively) in patients with progressive disseminated histoplasmosis (Wheat et al. 2007). Amphotericin B lipid complex may be preferred over liposomal amphotericin B because of lower cost. The deoxycholate formulation remains the standard formulation used in children. Amphotericin B is usually given for about a week, and then replaced with itraconazole in patients who have improved sufficiently to no longer require hospitalization. Nephrotoxicity remains a problem even with the lipid formulations of amphotericin B, and frequent monitoring of renal function, potassium and magnesium is required.

Amphotericin B lipid complex is currently recommended for small animal patients when amphotericin B treatment is required, rather than the deoxycholate form of the drug. The nephrotoxicity of the lipid complexed drug is significantly reduced compared with the deoxycholate form, giving it a markedly improved therapeutic index. However, monitoring of serum creatinine and urea nitrogen is still recommended (Grooters and Taboada 2003). Itraconazole is often administered concurrently with amphotericin B in small animals (Bromel and Sykes 2005; Gingerich and Guptill 2008; Greene 2006; Krohne 2000).

Itraconazole is recommended for individuals with mild to moderate disease and after clinical improvement with amphotericin B in humans with severe disease (Wheat et al. 2007). Itraconazole capsules require an acid pH for maximum absorption, and should be taken with food. The suspension does not require an acidic environment, and should be taken on an empty stomach. Itraconazole is usually given for a year or more in patients with progressive disseminated or chronic pulmonary histoplasmosis and for six to 12 weeks in patients with other forms of pulmonary histoplasmosis.

Itraconazole is the treatment of choice for most mild to moderately severe infections in small animals. The bioavailability of the oral solution for cats is greater than that of the capsule form (Boothe et al. 1997), and it may be possible to reduce the dose when using the oral solution. The recommended duration of treatment varies with each small animal patient, but most patients are treated for at least three to six months, and it is advised that treatment continue until at least 60 days past the complete resolution of clinical signs and radiographic or ocular lesions. Treatment may be required for a year or longer, particularly for widely disseminated disease (Greene 2006; Krohne 2000).

Itraconazole is cleared by cytochrome P450:3A4 metabolism, and blood levels may be affected by medications that interact with that enzyme. Itraconazole blood level measurement is encouraged during the second week of treatment in humans, and if treatment failure or relapse is suspected. Target blood levels are 1.0–10 µg/mL. Target blood levels are not well-defined for treatment of histoplasmosis in animals. In cats, steady state was reached after 14–21 days of treatment, and in dogs steady state was reached after 14 days (Boothe et al. 1997; Legendre et al. 1996). Itraconazole may cause a variety of adverse effects, most commonly anorexia, vomiting, diarrhoea, lethargy, skin ulceration, or increased hepatic enzyme activities. Gastrointestinal side effects may be related to high blood levels in humans (Lestner et al. 2009). Bilirubin and hepatic enzymes should be monitored during therapy (Legendre et al. 1996; Wheat et al. 2007).

Several other azoles are active in histoplasmosis and provide alternatives in those unable to take or who have failed itraconazole. Ketoconazole is infrequently used because it is less effective and causes more adverse effects than itraconazole, but its lower cost may be a reason to use it in some cases. Fluconazole is also less effective but may be used because of lower cost or reduced adverse effects compared with itraconazole. To prevent failure due to emergence of resistance to fluconazole (Wheat et al. 2001), doses of at least 10 mg/kg/day are recommended in human beings; similar data are not available for animals.

Posaconazole and voriconazole are more active than fluconazole and have been used successfully in humans with histoplasmosis (Wheat et al. 2007), but have not been evaluated extensively in animals and are expensive. These agents are reserved for patients unable to take or who have failed itraconazole and fluconazole. H. capsulatum also may become resistant to voriconazole (Wheat et al. 2006).

Several aspects of histoplasmosis caused by Histoplasma capsulatum organisms make control strategies difficult. The exposure levels to H. capsulatum var. capsulatum vary by locale within endemic regions. An attempt to identify heavily contaminated sites should be made for sites with a high historical or scientific potential for heavy Histoplasma contamination, particularly where earth moving, demolition, or renovation will likely expose people and animals to dust or other debris from the site. Large bird roosts, or bat habitats are candidates for heavily contaminated sites.

Most bird roosts identified to be contaminated have been in use for at least three years (Chick et al. 1981; Weeks and Stickley Jr. 1984); but once contaminated, however, the organism may persist in the absence of continued roosting. Contaminated sites should be posted to warn of the risk for exposure to histoplasmosis. Preventing long-term use of roosting areas could decrease environmental contamination (D’Alessio et al. 1965), but the efficacy of dispersion techniques has been questioned (Weeks and Stickley Jr. 1984). Soil decontamination with formalin has been advocated, but is inconvenient, hazardous to those involved and the environment, of uncertain effectiveness, and expensive. Physical removal of the contaminated material may be appropriate if the area of contamination is small. Decontamination procedures should follow published guidelines (Lenhart et al. 1997), which include the use of respirators and other personal protection equipment, to reduce their risk of exposure during the removal or decontamination of sites containing Histoplasma.

Persons at high risk for histoplasmosis should be informed of the probability of high exposure in certain environments. Health-care workers should be educated of the increased risk for contraction of histoplasmosis for immunocompromised individuals. Unfortunately, in many instances the avoidance of insidious exposure by high-risk individuals may not be possible, but advising high-risk individuals to avoid areas of probable high Histoplasma-contamination is warranted (Wheat 1992).

Several strategies are available for control of equine epizootic lymphangitis. These include the control of biting flies and ticks, practice of general hygiene and disinfection, and quarantine or euthanasia of infected animals (Gabal et al. 1983). A killed vaccine to protect from H. capsulatum var. farciminosum is reportedly available for horses in endemic areas (Kohn 2006).

The authors thank Dr. Craig Thompson for the excellent photomicrographs.