Abstract
Laboratory diagnosis of blastomycosis relies on a combination of methods, including antigen detection. We assessed the performance of analyte-specific reagents from Gotham Biotech (Portland, ME) for quantitative detection of Blastomyces dermatitidis galactomannan (GM) in urine using an enzyme immunoassay (EIA) compared to the Blastomyces quantitative EIA from MiraVista Diagnostics (Indianapolis, IN). Residual urine from 232 unique patients previously tested by the MiraVista assay was evaluated using the Gotham EIA, which showed 97.4% (74/76), 100% (156/156), and 99.1% (230/232) positive, negative, and overall agreement, respectively. Correlation between the quantitative B. dermatitidis antigen levels by the Gotham and MiraVista EIAs was low (R(2) = 0.20). Medical records were available for 36 of the 232 patients, among whom four had confirmed blastomycosis and both the Gotham and MiraVista EIAs were positive. Nine of these patients had histoplasmosis, and the Gotham and MiraVista EIAs yielded negative results in 44.4% (4/9) and 22.2% (2/9) of cases, respectively. Both assays were negative in the remaining 23 patients. After laboratory implementation of the Gotham EIA, chart reviews were performed on the first 50 unique patients (51 samples) tested by the assay in our hospital. Among these, 3/50 (6%) samples were positive by the Gotham EIA, including two samples from a patient with culture-confirmed blastomycosis and one from a patient with histoplasmosis (also positive by the MiraVista Blastomyces EIA). All remaining patients were negative by the Gotham EIA and had alternative diagnoses. Our findings show comparable performance between the Gotham and MiraVista quantitative EIAs for detection of B. dermatitidis GM in urine.