Diagnostic Usefulness of Galactomannan Assay for Invasive Aspergillosis

Background : Invasive aspergillosis (IA) is an important cause of morbidity and mortality among immunocompromised patients. However, IA is difficult to diagnose, especially in the aforementioned patient group. Recently, galactomannan assay (GMA) using enzyme immunoassay has been introduced in Korea. We evaluated its diagnostic usefulness in the diagnosis of IA and we analyzed the results according to the underlying diseases. Materials and Methods : All patients who underwent GMA during the period from October 2007 to June 2008 were evaluated retrospectively. According to the criteria of European Organization for Research and Treatment of Cancer/Mycoses Study Group, IA was classified into four clinical categories: 'proven', 'probable', 'possible', and 'non' IA. Patients with 'proven' and 'probable' IA were used as the reference standards for IA. GMA was performed using Platelia Aspergillus EIA (Bio-Rad, Hercules, CA, USA). Optical density index >= 0.5 was considered positive. Results : Of the 144 patients who underwent GMA, two patients were classificed as 'proven' IA and sixteen patients were 'probable' IA. Overall sensitivity, specificity, positive predictive value, and negative predictive value of the assay were 50% (95% confidence interval [CI95], 26-74%), 88% (CI95, 81-93%), 38% (CI95, 19- 59%), and 93% (CI95, 86- 97%), respectively. Among the 99 patients with underlying hematologic diseases, GMA showed 67% (CI95, 35- 90%) sensitivity and 89% (CI95, 80-94%) specificity, whereas in 45 patients with underlying diseases other than hematologic diseases, sensitivity and specificity of the assay were 17% (CI95, 0- 64%) and 87% (CI95, 73- 96%), respectively. Conclusions : GMA showed high specificity, irrespective of the patient population. However, sensitivity of GMA was low and the assay was less sensitive in patients with underlying diseases other than hematologic diseases than in patients with hematologic diseases.