The accuracy of serum galactomannan assay in diagnosing invasive pulmonary aspergillosis

Abstract

Galactomannan (GM) antigen is an aspergillus specific antigen that is released during the growth phase of invasive aspergillosis. We aimed to find the optimum cutoff and accuracy of serum Galactomannan assay in immunocompromised patients. Immunocompromised patients diagnosed with invasive pulmonary aspergillosis (IPA) based on the European Organization for Research and Treatment of Cancer/Invasive Mycosis Study Group (EORTC/MSG) with three levels of certainty proven, probable and possible, referred for GM antigen measurement at Immunology, Asthma and Allergy Research Institute (IAARI) from 2006 to 2009 and if they met the criteria were enrolled in this study. Totally 49 patients with IPA were enrolled in our study. According to EORTC/MSG, patients categorized into three levels of certainty: They were diagnosed as 'proven' invasive pulmonary aspergillosis 16(32.7%), 'probable' 18(36.7%) and 'possible' 15(30.6%). The most common host risk factor was solid tumors 17(34.7%). The accuracy of Galactomannan assay increased from 0.5 to 2 cutoffs. The optimum sensitivity and specificity obtained at the index cutoff of ≥1.5 for diagnosis of "proven" IPA; which were respectively, 69.2% and 72.2%. Other cutoffs had high variance between sensitivity and specificity for diagnosis of IPA. The calculated cutoff gained by receiver operating characteristic (ROC) analysis for detecting proven IPA was 1.5. Intermediate accuracy of serum GM test in conjunct with clinical findings would help early IPA detection among immunocompromised patients.