The performance of laboratories enrolled in three of the Centers for Disease Control microbiology performance evaluation programs was studied in relation to the number of actual patient specimens tested by the laboratories. Laboratories were grouped according to the number of specimens tested weekly to compare performance among the groups. Results of the study showed that laboratories, as a group, that tested large numbers of specimens performed better than laboratories that tested small numbers. The infrequency of performing certain tests or identifying certain microbial species can be an important factor in laboratory error. One strategy for laboratories that cannot resolve internal problems associated with low testing volumes would be to limit their testing to procedures that are done well.

Previously published methods to produce Paracoccidioides brasiliensis antigens for serological tests have yielded antigens of inconsistent quality and have involved the use of special semisynthetic media and growth periods of 1 to 3 months to yield suitable reagents. A simple procedure that uses commercially available potato glucose agar and either SABHI broth (Difco Laboratories) or Trypticase soy broth (BBL Microbiology Systems) inoculated with the mycelial form of P. brasiliensis consistently yielded high-titer antigens in 2 weeks or less. This new method permits the almost exclusive production of an antigen identical to the specific E antigen described by Yarzabal (Yarzabal et al., Sabouradia 14:275-280, 1976) and the apparently equivalent specific antigen 1 described by Restrepo and Moncada (A. Restrepo and L. H. Moncada, Appl. Microbiol. 28:138-144, 1974). In the immunodiffusion test, the rapidly produced antigen demonstrated a sensitivity of 90% by detecting antibody in sera from 103 of 114 proven cases of paracoccidioidomycosis. The specificity of this antigen was 100% because none of 139 sera from patients with heterologous mycotic diseases demonstrated diagnostic precipitins against the P. brasiliensis antigen. In the complement fixation tests, the rapidly produced antigen was not as suitable as the one prepared by the method of Restrepo-Moreno and Schneidau (A. Restrepo-Moreno and J. D. Schneidau, Jr., J. Bacteriol. 93:1741-1748, 1967).

Anti-Histoplasma capsulatum sera are specifically identified when they react with histoplasmin to form h and m precipitin lines in the agar-gel immunodiffusion test. Ten scientists from nine nations (Argentina, Canada, England, France, Israel, Japan, Russia, United States of America and Venezuela) have evaluated candidate international reference sera and histoplasmin as international reference reagents for use in an agar-gel microimmunodiffusion test. Criteria for the acceptance of the test as an international reference method depended upon the determination of the sensitivity and specificity of the procedure when applied by each laboratory to 31 coded rabbit sera having diverse concentrations of antibody to the h or m antigens in the reference histoplasmin. Acceptance of a candidate lot of lyophilized histoplasmin and candidate reference antiserum was based upon the determination of the optimum reconstitution volumes of both reagents to give discrete and clearly delineated h and m precipitin lines identified by their position relative to the serum and antigen wells. The reagents were similarly tested by five laboratories using macroimmunodiffusion procedures. When the lyophilized histoplasmin was stored at 5°C for two years, the h antigen showed no loss of relative potency, whereas the m antigen showed a 20% loss. All of the ten participating laboratories recommended that both the candidate serum and candidate histoplasmin be accepted as World Health Organization (W.H.O.) International Reference Reagents for the identification of h and m precipitin lines. Eight of the ten recommended that the described microimmunodiffusion test be accepted as the W.H.O. Reference Method for the use of these reagents.

Sera from 71 patients with culturally proven nocardiosis were tested for precipitins against a pool of Nocardia asteroides and N. brasiliensis culture filtrates and against antigens from the supernatant of homogenized N. asteroides cells. A human nocardiosis case serum was used as a reference. Sera from 56 of the 71 cases were reactive with either the culture filtrate antigen, the homogenate antigen, or both antigens, resulting in an overall sensitivity of 79%. Sera from 35 of the patients (49%) were positive with the homogenate antigens, and 28 (39%) showed bands of identity with the reference serum. Sera from 50 nocardiosis cases (70%) were positive with the pooled culture filtrate antigens, and 29 (41%) produced bands of identity with the reference serum. Of 89 sera from patients with various systemic mycotic diseases, tuberculosis, or actinomycosis, 24 (27%) were positive with the nocardial homogenate antigens and 4 (4.5%) showed precipitin bands of identity. Thirty-five of the 89 sera (42%) were positive with the nocardial culture filtrate antigens, and 6 (6.7%) showed bands of identity. The majority of sera demonstrating false-positive reactions were from tuberculosis and actinomycosis cases. One of seven sera from well individuals produced a precipitin band with the culture filtrate antigen, but this was not a band of identity with reference serum. These antigens did not distinguish antibodies from patients with N. asteroides, N. brasiliensis, or N. cavia infections.

Cell walls isolated from the yeast forms of 2 strains each of Histoplasma capsulatum and Blastomyces dermatitidis were treated for up to 3 weeks with an enzyme complex containing chitinase and β-1:3-glucanase. The release of glucose, protein, and amino-sugars from the cell walls was monitored. Cell walls of H. capsulatum lost about 25% of their dry weight as glucose, about 19% as amino-sugar (mainly free N-acetyl glucosamine), and about 6% as protein. Cell walls of B. dermatitidis lost only about 1% of their dry weight as glucose, about 22% as amino-sugar (mainly chitobiose), and about 6% as protein. Electron microscopy of intact yeast-form cells showed them to be extensively degraded by the enzyme complex. Complement fixation (CF) tests indicated that the enzyme treatments caused no substantial change in the antigenic specificity or sensitivity of the cell walls. However, concentrated culture filtrates of B. dermatitidis and the soluble fraction of H. capsulatum cell walls after chitinase treatment, show...

The diagnosis of sporotrichosis can be time consuming. Serological procedures could facilitate the rapid and accurate diagnosis of this disease. A slide latex agglutination (SLA) test for sporotrichosis was developed and compared with the tube agglutination (TA), complement fixation (CF), and immunodiffusion (ID) tests in the serological study of 80 proven human cases of sporotrichosis representing the cutaneous, subcutaneous, and extracutaneous forms of the disease. In addition, the indirect fluorescent antibody (IFA) technique was applied to 61 case sera. In the SLA test, latex particles sensitized with culture filtrate antigens from the yeast form of Sporothrix schenckii (B 959) detected 94% of the cases, as compared to 96% of the cases detected by the TA test, 68% by the CF test, and 56% by the ID test. The IFA test detected 90% of the 61 cases. The SLA and ID tests were specific, showing no reactions with sera from 86 persons with no disease or with diseases other than sporotrichosis. Because of its sensitivity, specificity, ease of performance, and ability to provide results in 5 min, the SLA test is highly recommended for routine use in the clinical laboratory.

A newly developed latex agglutination (LA) test and a modified immunodiffusion (ID) test were evaluated. The antigen used was a homogenate of Candida albicans. A total of 167 antisera were employed in the evaluation. They included 36 sera from clinically well persons; 78 from patients with various clinical forms of candidiasis; 52 from patients with proven cases of aspergillosis, blastomycosis, coccidioidomycosis, cryptococcosis, histoplasmosis, nocardiosis, paracoccidioidomycosis, sporotrichosis, and tuberculosis; and one serum from a patient with toruloposis. Use of the LA test in conjunction with the ID test permitted the detection of more than 90% of 43 proven candidiasis cases. Of all the heterologous cases and normal human sera tested, LA reactions were noted with 3 of 10 cryptococcosis case specimens, 1 of 9 tuberculosis case specimens, and with the torulopsemia case serum. In contrast, the only heterologous serum reactive in the ID test was that from the patient with torulopsemia. Torulopsis glabrata and C. albicans antisera gave identical reactions in LA and ID tests with T. glabrata or C. albicans antigens. ID tests with selected antigens, however, permitted differentiation of rabbit and human T. glabrata antibody from that of C. albicans antibody. Six different precipitins were recognized with the C. albicans antigens. The occurrence of multiple precipitin lines and high LA titers was suggestive of severe candidiasis. The LA test, in contrast to the ID test, appeared to have prognostic value. Together, the LA and ID tests provided a simple, rapid, and accurate means of detecting and monitoring infections by species of Candida.