Search

False-positive serologic tests for histoplasmosis (H) and blastomycosis (B) are common in populations from endemic areas. In order to determine the significance of false-positive test results, we reviewed the final diagnoses of all patients whose sera were submitted to our laboratory for radioimmunoassay (RIA) and immunodiffusion (ID) during a 3-yr period. Of the 263 patients whose sera were examined, 29 (11%) had H or B; 41 (17.5%) of the remaining 234 patients had false-positive test results. Of these 41 patients, 31 were positive for H alone, and 10 had antibodies to both H and B. All three patients with false-positive ID tests for histoplasmosis also had positive titers (greater than or equal to 1:16) on RIA. No patient had a false-positive ID result for blastomycosis. The percentage of patients in each of five major diagnostic categories with and without false-positive serologic tests was similar (p greater than 0.05). The majority of patients had pulmonary infections, almost half of which were granulomatous infections other than H or B; this reflects the clinical indications for requesting fungal serologic tests. A positive fungal serology is not useful in suggesting the presence of a pulmonary disease other than H or B in patients from an endemic area suspected of having a pulmonary mycosis.

A highly specific enzyme immunoassay (EIA) was developed for the serodiagnosis of blastomycosis. Among the earliest sera available from eight active cases of blastomycosis, seven were reactive by EIA, three by immunodiffusion, and three by complement fixation. The one seronegative case was associated with acquired immunodeficiency syndrome (AIDS). No cross-reaction was observed with sera from 12 patients with active histoplasmosis or from five patients with active coccidioidomycosis or from 23 healthy persons. In contrast, all patients' sera cross-reacted in conventional EIAs based on commercial immunodiffusion antigens. The blastomycosis-specific EIA seems to offer sensitivity without compromising specificity. Rheumatoid factors, which could cause false-positive reactions, were controlled for.

Currently available serologic tests for histoplasmosis and blastomycosis have limited sensitivity. In order to develop a more sensitive screening test, the radioimmunoassay (RIA) technique was adapted to detect circulating antibodies to Histoplasma mycelial, Histoplasma yeast, and Blastomyces yeast antigens. Results were compared with those of complement fixation (CF) using serums from 4 groups of subjects. Group 1 (n = 104) consisted of consecutive healthy blood donors from an endemic area. Group 2 (n = 8) had active histoplasmosis and Group 3 (n = 12) had active blastomycosis. Group 4 (n = 30) had other chronic lung diseases. Among Group 1 subjects, elevated titers to at least 1 antigen were found in 13% using RIA and 21% using CF. Only 50% of patients with active mycoses (Groups 2 and 3) had elevated CF titers to a homologous antigen, whereas the RIA test was positive in 19 of 20 patients (95%). Elevated titers to 1 or more antigens was found in Groups 1 and 4, and frequent cross-reactions among the antigens occurred with both tests, indicating low specificities for both RIA and CF. The RIA is a sensitive and rapid screening test for histoplasmosis and blastomycosis, and a negative RIA using the 3 antigens tested makes the diagnosis of active infection very unlikely in immunologically competent patients.

Using a new enzyme immunoassay (EIA) and standard immunodiffusion (ID) and complement fixation techniques for antibody to the A antigen of Blastomyces dermatitidis, we tested serum from 27 patients with blastomycosis diagnosed histopathologically or by culture; 20 with diagnoses made during 1981 through 1983 (Group A) and 7 during 1974 through 1976 (Group B). We also studied 30 control subjects with Mycoplasma pneumoniae infection (17 subjects), histoplasmosis (6 subjects), coccidioidomycosis (1 subject) and no known disease (6 subjects). Detectable antibody by all 3 tests was present more often for Group A than for Group B, possibly because of delay in testing. Enzyme immunoassay was the most sensitive test; a 1:8 or greater titer was present in acute-phase serum of all Group A patients tested, and a 1:64 or greater titer was associated with disseminated disease (p = 0.003). A positive ID was also more common in disseminated (88%) than in localized (33%) disease. Enzyme immunoassay titers of 1:16 were present in 4 control subjects, 3 with histoplasmosis. The 100% predictive values of a negative EIA and positive ID suggest that these tests are useful for serologic screening (EIA) and serologic confirmation (ID) of suspected blastomycosis, particularly in disseminated disease. Enzyme immunoassay titers of 1:32 or greater strongly support the diagnosis and titers of 1:16 or less may indicate localized disease or be nonspecific. None of the serologic tests predicted disease progression or reactivation.

Of 78 patients with blastomycosis, 3 patients had received glucocorticoid therapy prior to diagnosis and 3 others had an underlying hematologic malignancy (chronic lymphocytic leukemia, chronic myelogenous leukemia, multiple myeloma). The clinical picture in these 6 patients was similar to blastomycosis in nonimmunosuppressed patients (in contrast to histoplasmosis or coccidioidomycosis in immunosuppressed patients in whom a distinct clinical syndrome is often seen). The patients presented with chronic pulmonary infiltrates or with isolated skin ulcers. The response to therapy was good if the diagnosis was made early. Blastomycosis can occur in immunosuppressed patients. However, the spectrum of underlying illness is not that seen in opportunistic histoplasmosis or coccidioidomycosis, where patients with T-cell defects predominate. Possible explanations for the rarity of blastomycosis in more classically T-cell-immunosuppressed patients are discussed.

A 73-year-old man with malignant lymphoma who was undergoing treatment with systemic chemotherapy presented with an acute, febrile illness associated with cough and sputum production. The patient developed bilateral, diffuse, pulmonary interstitial infiltrates with rapidly progressive respiratory insufficiency. The diagnosis was made by demonstrating Blastomyces dermatitidis in bronchial brush and transbronchial lung biopsy specimens obtained through fiberoptic bronchoscopy.

This study evaluates the usefulness of enzyme immunoassay (EIA) as a screening test for serum antibodies to Histoplasma capsulatum and Blastomycoses dermatitidis and compares the results using this assay with those using complement fixation (CF), immunodiffusion (ID), and radioimmunoassay (RIA), in 12 patients with active histoplasmosis, 12 with active blastomycosis, 28 with other acute or chronic lung diseases and 25 healthy blood donors from an endemic area; EIA was as sensitive as RIA and more sensitive than CF and ID in the 24 patients with active fungal infections. The specificity of EIA was equal to that of the other serologic tests. We conclude that EIA is a useful screening test for serum antibodies to H. capsulatum and B. dermatitidis that avoids some of the problems associated with other sensitive assays such as RIA. A negative EIA result is evidence against invasive histoplasmosis or blastomycosis and suggests an alternative diagnosis.

Blastomycosis is not generally recognized to be a self-limited pulmonary infection. We report 13 patients with blastomycosis who followed a self-limited course. Presenting complaints were usually those of an acute pulmonary infection with fever, productive cough, and pleurtiic chest pain. The duration of symtpoms before diagnosis was usuallms were variable and not diagnostic. The blastomycin skin test and complement-fixing serologies to blastomycin were generally not helpful. In all patients the diagnosis was made by either cultural or visual identification of the organism from sputum, bronchial washings, or pleural fluid. All patients were improving both clinically and by chest roentgenograms by the time the diagnosis was reached, and thus therapy was withheld. Follow-up of these 13 patients ranged from 5 months to 8 yr (mean, 43 months), and in no instance has there been any evidence of reactivation of the illness.

Double immunodiffusion (DID) is a useful, inexpensive antibody screening test that is more specific than complement fixation for blastomycosis. Seven of 10 patients with proved blastomycosis had positive results on DID assays for antibody to the A antigen of Blastomyces dermatitidis. Sputum cytologic examination and skin or other tissue biopsies may be diagnostic if the organism is demonstrated. Culture remains the sina qua non of diagnosis, and a negative DID test result for A precipitin does not rule out this infection. Follow-up serologic studies of patients in remission and in relapse should be undertaken to determine the evolution of the immune response to the A antigen.

The adult respiratory distress syndrome (ARDS) developed in 2 patients with blastomycosis. In one, ARDS apparently followed the rupture of an infected subcarinal lymph node into a bronchus. Extensive involvement of alveoli with Blastomyces dermatitidis was noted at autopsy in both patients. Four additional cases of blastomycosis with definite or presumed ARDS are reviewed. The clinical course of these 6 patients was characterized by a prolonged prodrome (median, 36 days), consistent with pulmonary blastomycosis, and a rapid demise (median, 5 days) associated with ARDS.

A 53-year-old man experienced relapse of pulmonary blastomycosis 9 years after having been treated successfully with amphotericin B and responded satisfactorily to retreatment. Relapse of blastomycosis after adequate treatment with amphotericin B is apparently uncommon. However, published reports include relatively few patients who have had long-term follow-up.

Two patients with second relapses of pulmonary blastomycosis after successful treatments with adequate doses of amphotericin B were retreated with miconazole. One patient had experienced an acute intrabronchial spread of disease to the contralateral lung before miconazole therapy. Marked clinical and roentgenographic improvement occurred with treatment, followed by mycologic relapse 1 month later. The other patient sustained a more indolent relapse of disease. A favorable response to treatment with miconazole was demonstrated by improvement in abnormalities noted on a chest roentgenogram and a [67]gallium-citrate scan.