66 Histoplasmosis

Histoplasmosis is an infection caused by Histoplasma capsulatum, a soil fungus. Exposure occurs by inhala­tion, and the primary infection is in the lung. The disease is usually benign and self-limited, despite a strong tendency for invasion of the bloodstream dur­ing the primary infection.

Distribution and Incidence

H. capsulatum has been isolated from soil of more than 50 countries. It is most common in temperate climates along river valleys and has been found in North, Central, and South America; India; South­east Asia; and rarely Europe.

By far the most heavily endemic area in the world is the east central United States, particularly the Mississippi and Ohio River valleys. It is most preva­lent in the states of Ohio, Kentucky, Indiana, Illi­nois, Kentucky, Tennessee, and Arkansas. Surround­ing states also have many infections.

Infection is almost universal in the most heavily endemic areas. Skin test surveys reveal that over 90 percent of persons living in some counties in the central United States have had histoplasmosis be­fore age 20 (Edwards et al. 1969). Based on skin test sιπveys, there are probably 40 to 50 million people in the central United States who have had histo­plasmosis, and there are several hundred thousand new cases each year. The number of serious infec­tions requiring diagnosis and treatment is very small, perhaps 1 or 2 percent of the total.

In contrast to the highly endemic central United States, only a handful of individual cases have been documented in Europe. Skin test surveys show about 0.5 percent of the population positive to histo- plasmin. In Italy, positive skin tests are slightly more common (about 2 percent) (Mantovani 1972).

Etiology and Epidemiology

H. capsulatum is a thermal dimorphic fungus. At 25°C, it grows on Sabouraud agar as a fluffy-white mycelium, which bears microconidia and also charac­teristic tuberculate macroconidia. The organism is free-living in nature in this form. At 37^oC, it grows as a small (2- to 4-p.m diameter) yeast. The organism is found in this form in infected tissue.

A minor disturbance of fungus-laden soil may scat­ter spores into the air. The microconidia are inhaled, causing infection. Within the lung the organism con­verts to the yeast phase, which is not infectious. Person-to-person transmission does not occur. So- called epidemics of histoplasmosis are more accu­rately point-source outbreaks.

Within a highly endemic area, the organism is widely but not uniformly distributed. Microfoci with high concentrations of organisms are found by chicken coops and starling roosts, and in caves in­habited by bats. The nitrogen-rich excrement of birds and bats provides a favorable growth environ­ment for the fungus. Exposure of small groups of people to high concentrations of organisms at such sites may result in outbreaks of symptomatic infec­tion. These are fairly easy to identify because a se­vere respiratory illness occurs simultaneously in a group of people who were together for a particular activity 14 days earlier. Extremely large outbreaks of longer duration have occurred during excavation for road or building construction. A good example is the community-wide outbreak of histoplasmosis that occurred during the building of a swimming pool and tennis court complex in Indianapolis, Indi­ana, infecting perhaps over 100,000 people (Wheat et al. 1981).

Most cases of histoplasmosis, however, are spo­radic and result from casual exposure to environmen­tal spores. Patients with sporadic illness probably inhale fewer spores and are more likely to be asymptomatic or minimally symptomatic. The vast majority of these infections are never recognized and are known to exist only as a result of skin test surveys.

Immunology

Cell-mediated immunity is crucial in host defense. Inhalation of spores causes patchy areas of pneumo­nitis. The spores are transformed into the yeast form, which is engulfed by macrophages. The yeasts multiply intracellularly with a generation time of 11 hours (Howard 1965). The regional lymph nodes are quickly involved, and hematogenous spread occurs. The fungus is cleared from the blood by reticuloendo­thelial cells throughout the body. Specific cell- mediated immunity develops and rapidly checks the infection in the lung and at distant sites. Granuloma formation and necrosis occur at sites of infection.

Humoral antibody also develops. Although circu­lating antibodies are the basis of many diagnostic tests, they have little importance in limiting the infection. Hyperimmune serum is of no benefit in experimental infections, and Iiypogammaglobulin- emic patients are not prone to progressive infections.

Clinical Manifestations and Pathology

Primary Pulmonary Histoplasmosis

This disease is asymptomatic at least half of the time. Symptomatic patients become ill about 2 weeks after exposure. They have an influenza-like illness with fever, chills, myalgias, headache, and a nonproductive cough. Rare manifestations include arthralgias, arthritis, and erythema nodosum. With or without symptoms, the chest roentgenogram may show patchy areas of pneumonitis and prominent hilar adenopathy.

A primary fungemia probably occurs in most cases. The calcified granulomas commonly found in spleens and livers of patients from endemic areas result from this primary, self-limited fungemia, not from progressive dissemination.

Following exposure to an unusually heavy inoc­ulum, a more diffuse pulmonary involvement may occur, with an extensive nodular infiltrate on the chest roentgenogram. Dyspnea may be added to the other symptoms, and symptoms are more severe and last longer. Most patients recover without treat­ment, but extreme cases may progress to respiratory failure.

The chest roentgenogram often returns to normal after a primary pulmonary infection, but a variety of residual abnormalities may be seen. Initial soft infil­trates may harden and contract, leaving one or sev­eral nodules. Central necrosis may result in a dense core of calcium (a “target” lesion), but this is not universal. Infrequently, alternate periods of activity followed by healing may result in concentric rings of calcium as the lesion slowly enlarges. Lymph node calcification, either in association with a paren­chymal nodule or as a solitary finding, is common. Finally, small “buckshot” calcifications may be scat­tered over both lung fields, a pattern characteristic of uneventful recovery after exposure to a heavy inoculum of organisms.

Primary histoplasmosis has several uncommon lo­cal complications within the chest. Involvement of the pericardium can cause a nonspecific acute pericar­ditis (Wheat et al. 1983). The inflammation probably represents a response to adjacent infection in the lung, as the cultures of pericardial fluid are usually sterile. Many cases are probably misdiagnosed as benign viral pericarditis. Rarely, the process pro­gresses to chronic constrictive pericarditis. More delayed and more serious, an extensive fibrosing pro­cess in the mediastinum can cause vascular compres­sion and result in the superior vena caval syndrome with edema of the head and upper extremities and development of superficial venous collaterals across the chest wall. Inflammation adjacent to the esopha­gus may cause a traction diverticulum. A lymph node impinging on a bronchus may cause a chronic cough. If a calcified lymph node erodes through a bronchus, it becomes a broncholith. Hemoptysis is a common clinical manifestation.

Chronic Cavitary Histoplasmosis

Although it may occur anywhere in the lung, chronic cavitary histoplasmosis usually involves both upper lobes and closely resembles reinfection tuberculosis in its roentgenographic appearance. The mechanism of infection, however, is not endogenous reactivation.

Disseminated Histoplasmosis

This condition refers to any progressive extrapul- monary infection. There is a range of infection with different tissue responses. At one extreme there are massive numbers of organisms in all reticuloendo­thelial organs with little tendency to granuloma for­mation. Clinical features include high fever, hepato- Splenomegaly, Iymphadenopathy, and pancytopenia due to bone marrow involvement. This type of dis­seminated histoplasmosis has been called the “infan­tile” form and may lead to death within days or weeks. Other patients, often older adults, have a more indolent illness, many months in duration, characterized by low or moderate fever, weight loss, and skin and mucous membrane lesions. Biopsies of involved tissues show well-formed granulomas simi­lar to sarcoidosis. Organisms are scanty and often are demonstrated only with special stains.

Disseminated histoplasmosis also occurs as an op­portunistic infection. The degree of granulomatous tissue response may vary from none to a consider­able amount and has prognostic value. If the bone marrow biopsy shows epitheloid granulomas or even recognizable aggregates of macrophages, the re­sponse to treatment is quite good. If the biopsy shows no granulomas, tissue necrosis, and a large number of organisms, the prognosis is very poor (Davies, McKenna, and Sarosi 1979).

Disseminated histoplasmosis often presents as a nonspecific systemic febrile illness rather than as a pulmonary infection.

Immunosuppressed patients probably get dissemi­nated histoplasmosis in two ways. If they inhale the organisms while immunosuppressed, the primary in­fection will progress, as documented in the India­napolis outbreak (Wheat et al. 1982). On the other hand, if they become profoundly immunosuppressed long after their primary infection, the disease may reactivate (Davies, Kahn, and Sarosi 1978). This is suggested by the systemic, nonpulmonary nature of the illness and is supported by the recent experience with the acquired immune deficiency syndrome (AIDS). Patients with past exposure to endemic ar­eas are developing disseminated histoplasmosis while living in nonendemic areas such as New York City (Huang et al. 1988), San Francisco, and Los Angeles.

Disseminated histoplasmosis may also present as a more localized infection. Examples include central nervous system histoplasmosis, meningeal histoplas­mosis, and isolated gastrointestinal histoplasmosis, which often involves the terminal ileum. All are extremely rare.

Diagnosis

The histoplasτnin skin test is a valuable epidemiologi­cal tool that has permitted mapping of the endemic area. However, it is worthless in individual case diagnosis. A positive skin test means only that the person is one of many millions who have had histoplasmosis, probably remotely and without se­quelae. It does not mean that a current illness under investigation is being caused by the fungus.

Primary histoplasmosis is usually not diagnosed at all. Sputum cultures are positive in less than 10 percent of cases. Most recognized cases are diag­nosed by serology. Serologic tests include immunodif­fusion (M and H bands) and complement fixation (yeast and mycelial antigens) tests. An M band by immunodiffusion is fairly specific. The H band is never found alone. When present with the M band (10 to 20 percent of the time), it adds further to specificity. Unfortunately, the immunodiffusion test is insensitive and appears slowly after primary infec­tion. Less than 25 percent of symptomatic patients have a positive test 2 weeks after the onset of clini­cal illness (Davies 1986). The complement fixation test is more sensitive but less specific. The titer against the mycelial antigen is not important be­cause it is almost always lower than the titer against the yeast antigen. A titer of 1 to 32 or higher against the yeast antigen is diagnostic if the clinical picture suggests histoplasmosis. Unfortunately, only 60 percent of patients have a positive test 2 weeks after the onset of clinical illness, and many have 1:8 or 1:16 titers (Davies 1986). What this means is the serologic tests are most useful for diagnosing pa­tients who have already recovered. Patients with rapidly progressive pneumonias not responding to antibacterial antibiotics need urgent diagnosis, espe­cially if respiratory failure is impending or actually develops. Some patients are diagnosed by serology. Others require lung biopsy for histopathological di­agnosis, because a negative serology cannot exclude the diagnosis, and empirical treatment for all possi­ble causes of progressive pulmonary infection is not possible.

Chronic cavitary histoplasmosis is easier to diag­nose. The pace is slower. Tuberculosis is suspected first, but the tuberculin skin test is negative and the sputum is negative for tuberculosis by smear and culture. Sputum cultures for H. capsulatum are usu­ally positive, as are serologic tests (immunodiffu­sion: 75 percent; complement fixation: 90 percent).

Disseminated histoplasmosis is difficult to diag­nose because the illness is so nonspecific. Serologic tests are positive in over half of cases and may pro­vide an important clue. Serodiagnostic tests are least helpful in the immunosuppressed because they are less sensitive and because the pace of the illness may be so fast that there is no time to wait for the results. Histopathological examination of tissue biop­sies is the method of diagnosis in most cases. Bone marrow biopsy is particularly valuable in febrile illnesses without localizing features. Special stains, such as one of the many modifications of the silver stain, are necessary to ensure visualization of the organisms. Cultures of blood, bone marrow, and other tissues and of body fluids may also give the diagnosis in some cases.

History and Geography

Infection with H. capsulatum was first described in April 1906 by Samuel Darling (1906). From an au­topsy of a laborer who died while working on the Panama Canal, he described a disseminated infec­tion of the reticuloendothelial system caused by an organism that he believed was protozoan. Macro­phages were filled with small organisms. Within a few years he reported two other similar cases. In 1913, H. da Rocha-Lima discussed the published pho­tomicrographs of Darling’s cases and speculated that the organism might be a fungus rather than a proto­zoan (Rocha-Lima 1913).

Another autopsy case was reported in Minnesota in 1926 by W. A. Riley and C. J. Watson (1926), who credited Darling with being the first to describe the infection in 1906. Later, however, there was some confusion as to whether a case report of R. P Strong (1906) from the Philippines in January 1906 had described histoplasmosis first. But in a personal let­ter many years later, Strong stated his belief that the case he had described was not histoplasmosis, but rather a rare human infection with Cryptococcus farciminosus, the cause of farcy in horses (Parsons and Zarafonetis 1945). Thus credit for the first case description remains with Darling, who recognized that the disease was previously undescribed and who named the organism and the illness.

After Riley and Watson’s paper, scattered autopsy reports of similar cases followed, mostly from the central United States. Then in 1934, the first pre- mortem diagnosis of such a patient was made by finding the characteristic intracellular organisms on a peripheral blood smear (Dodd and Tompkins 1934). W A. DeMonbreun (1934) isolated the infectious agent, proved that it was a ftmgus, and demon­strated its thermal dimorphism.

In January 1945, R. J. Parsons and C. J. Zara- fonetis (1945) reported 7 cases, reviewed 71 previous cases, and concluded that histoplasmosis was a rare systemic infection that was nearly always fatal. However, in the same year, A. Christie and J. C. Peterson (1945) and also C. E. Palmer (1945), using antigen derived from the first isolate, demonstrated that great numbers of asymptomatic persons in the central United States had been infected with the fungus. Furthermore, they showed that almost all tuberculin-negative persons with calcifications on chest roentgenogram had positive histoplasmin skin tests. Quickly the endemic area was mapped and fungus was isolated from soil in areas with high skin-test positivity. The new conclusion was that histoplasmosis is very common and almost invari­ably benign and self-limited. The fatal cases were rare and exceptional.

Most of the skin-test reactors in the early surveys had had asymptomatic or minimally symptomatic nonspecific infections. The retrospective discovery of a highly symptomatic but also self-limited form of primary histoplasmosis soon followed. Small groups of patients exposed to high concentrations of organ­isms, often in closed spaces, had been verified as victims of epidemics of an unknown but relatively severe acute pulmonary illness. An epidemiological investigation of one such outbreak, which occurred in 1944 and was reported 3 years later (Cain et al. 1947), demonstrated convincingly that H. capsu­latum had been the offending agent.

Upper-Iobe cavitary histoplasmosis resembling tu­berculosis was first described in Missouri in 1956 by M. L. Furculow and C. A. Brasher among Sanitorium patients being treated for tuberculosis. The mecha­nism of infection was most likely progressive pri­mary infection in an abnormal lung, and not reacti­vation (Goodwin et al. 1976).

In 1955, the isolation of amphotericin B was de­scribed by W. Gold and colleagues, and within a few years the drug was available for treatment of a wide variety of fungal infections. This drug, despite some toxicity, proved highly effective for histoplasmosis and remains the agent to which newer alternatives must be compared. Ketoconazole, a nontoxic oral imidazole, arrived in the 1980s. It is not as effective as amphotericin B but is reasonable therapy for mild to moderately ill patients with chronic cavitary dis­ease and indolent forms of disseminated disease.

With the increase in the use of glucocorticoids and cytotoxic drugs for malignant and nonmalignant dis­eases, disseminated histoplasmosis assumed increas­ing importance (Davies, Khan, and Sarosi 1978; Kauftnan et al. 1978). Endogenous reactivation was suspected as a mechanism in some cases because the illness presented as a nonspecific febrile illness. Treatment with amphotericin B was very effective if the diagnosis was made quickly and if the patient had some degree of cell-mediated immune response.

Finally, as previously noted, AIDS brought a new level of suppression of the cell-mediated immune sys­tem. The concept of endogenous reactivation received further support, for, unlike other immunosuppressed patients, even those AIDS patients who respond to treatment are not cured but require long-term sup­pressive therapy to prevent relapse of infection (John­son et al. 1986).

Scott F. Davies

Bibliography

Cain, J. C., et al. 1947. An unusual pulmonary disease. Archives OfInternal Medicine 79: 626—41.

Christie A., and J. C. Peterson. 1945. Pulmonary calcifica­tion in negative reactors to histoplasmin. American Journal of Public Health 35: 1131-47.

Darling, S. T. 1906. A protozoan general infection produc­ing pseudo tuberculosis in the lungs and focal necrosis of the liver, spleen and lymph nodes. Journal of the American Medical Association 46: 1283-6.

Davies, S. F. 1986. Serodiagnosis of histoplasmosis. Semi­nars in Respiratory Infections 1: 9-15.

Davies, S. F., M. Khan, and G. A. Sarosi. 1978. Dissemi­nated histoplasmosis in immunologically suppressed patients: Occurrence in non-endemic area. American Journal of Medicine 64: 94-100.

Davies, S. F., R. W. McKenna, and G. A. Sarosi. 1979. Trephine biopsy of the bone marrow in disseminated histoplasmosis. American Journal of Medicine 67: 617-77.

DeMonbreun, W. A. 1934. The cultivation and cultural characteristics of Darling’s histoplasma capsulatum. American Journal of Tropical Medicine 14: 93-125.

Dodd, K., and E. H. Tompkins. 1934. Case of histoplasmo­sis OfDarling in an infant. American Journal of Tropi­cal Medicine 14: 127-37.

Edwards, L. D., et al. 1969. An atlas of sensitivity to tuberculin, PPDB, and histoplasmin in the United States. American Review of Respiratory Disease 90 (Supplement) 1: 132.

Furcolow, M. L., and C. A. Brasher. 1956. Chronic progres­sive (cavitary) histoplasmosis as a problem in tubercu­losis sanitoriiuns. American Review of Tuberculosis and Pulmonary Disease 73: 609—19.

Gold, W., et al. 1955. Amphotericins A and B, antifungal antibiotics produced by a Streptomycete. Antibiotics Annual 579—86.

Goodwin, R. A., et al. 1976. Chronic pulmonary histoplas­mosis. Medicine 55: 413—52.

Grayston, J. T., and M. L. Furcolow. 1953. Occurrence of histoplasmosis in epidemics: Epidemiological studies. American Journal of Public Health 43: 665-76.

Howard, D. H. 1965. Intracellular growth in histoplasma capsulatum. Journal OfBacteriology 89: 518-23.

Huang, C. T., et al. 1988. Disseminated histoplasmosis in the acquired immunodeficiency syndrome: Report of 5 cases from a non-endemic area. Archives of Internal Medicine 2147: 1181—4.

Johnson, P. C., et al. 1986. Progressive disseminated histo­plasmosis in patients with the acquired immunode­ficiency syndrome. Seminars in Respiratory Infections 1: 1-8.

Kaufman, C. A., et al. 1978. Histoplasmosis in immunosup­pressed patients. American Journal of Medicine 64: 923-32.

Mantovani, A. 1972. Histoplasmosis in Europe. Annales de la Societe Beige de Medecine Tropicale 52: 421—34.

Palmer, C. E. 1945. Non-tuberculosis pulmonary calcifica­tion and sensitivity to histoplasmin. Public Health Reports 60: 513—20.

Parsons, R. J., and C. Zarafonetis. 1945. Histoplasmosis in man: A report of 7 cases and a review of 71 cases. Archives of Internal Medicine 75: 1—23.

Riley, W. A., and C. J. Watson. 1926. Histoplasmosis of Darling with report of a case originating in Minnesota. American Journal of Tropical Medicine 6: 271—82.

Rocha-Lima, H. da. 1913. Beitrag zur Kenntnis der Blasto- mykosen: lymphangitis epizootica und histoplasmo­sis. Zentralblatt fiir Bakteriologie, Parasitenkunde, Infektionskrankheiten; and Hygiene Seminars in Res­piratory Infections 1: 9-15.

Strong, R. P. 1906. A study of some tropical ulcerations of the skin with particular reference to their etiology. Philippine Journal OfScience 1: 91—115.

Wheat, L. J., et al. 1981. A large urban outbreak of histoplasmosis: Clinicalfeatures. Annals of Internal Medicine 94: 331-7.

1982. Risk factors for disseminated or fatal histoplas­mosis. Annals of Internal Medicine 95: 159—63.

1983. Pericarditis as a manifestation of histoplasmosis during two large urban outbreaks. Medicine 62: 110­18.