Authors propose best urine antigen testing protocol for initial detection of histoplasmosis
January 23 2019
The initial symptoms of histoplasmosis are often non-specific and, though the disease generally responds well to antifungal therapy early in its progression, it can develop into a serious illness if undetected. Therefore, rapid and accurate diagnosis is necessary.
The current practice of most clinicians is to request Histoplasma antigen detection by ELISA on urine (UAg) as an initial screen for histoplasmosis, along with additional laboratory testing. Less often, physicians order antigen detection tests on serum (SAg) instead, or in addition, to the UAg assay. Previous studies of UAg and SAg detection have shown that their sensitivity and specificity vary depending on the type and degree of Histoplasma infection, with UAg reportedly demonstrating higher sensitivity. Dr Diane Libert and colleagues have retrospectively analysed 21,248 UAg and SAg Histoplasma tests, to determine whether SAg detection plays a valuable role in the diagnosis of histoplasmosis.
Of the total studied, 601 paired (requested for the same patient within 1 week) UAg and SAg tests were identified. Of the paired tests reviewed, 98% agreement between the results was determined. Medical records were available for 8 of the 11 discordant results, from which 6 of the UAg results were categorised as false positives, one as a true positive and one as a false negative.
Overall, UAg and SAg detection performed comparatively, with greater sensitivity demonstrated by the UAg assay - a useful characteristic for screening tests. The authors recommend that the most practical and cost-effective approach for clinicians is to request UAg detection as an initial screen for histoplasmosis, and to confirm positive results with SAg detection, as well as reviewing additional laboratory studies if required.
Libert D, Procop GW, Ansari MQ. Histoplasma Urinary Antigen Testing Obviates the Need for Coincident Serum Antigen Testing. Am J Clin Pathol. 2018 Mar 7;149(4):362-368. doi: 10.1093/ajcp/aqx169.