Background: We evaluated the performance of a galactomannan (GM) assay in bronchoalveolar lavage (BAL) fluid compared to serum samples for the diagnosis of invasive pulmonary aspergillosis (IPA) in patients with hematological diseases.
Methods: Two hundred and fifty-five bronchoscopies were performed on 230 patients. Bronchial and alveolar samples from BAL fluid as well as serum samples were analyzed in the GM assay.
Results: Twenty-eight cases of IPA (11%) were diagnosed. The sensitivity, specificity, positive predictive value, and negative predictive value of the GM assay using a cut-off of 0.5 were 57.1%, 99.3%, 94.1%, and 92.5%, respectively, for the alveolar sample; 44.0%, 99.3%, 91.7%, and 91.4%, respectively, for the bronchial sample; and 60.7%, 100%, 100%, and 92.9%, respectively, for serum. The highest sensitivity (78.6%) with good specificity (98.6%) was obtained with a 'triple detection' of GM in bronchial, alveolar, and serum samples. Neutropenia and antifungal therapy for only 24h increased the sensitivity, while antifungal treatment for ≥ 2 days decreased assay performance. Moreover, a trend towards a higher volume of aspirated fluid in GM-negative BAL (p=0.092) was observed.
Conclusions: In contrast to recently published data, we found only moderate sensitivity, but high specificity and high positive predictive value of the detection of GM in BAL fluid. In addition, neutropenia, antifungal therapy, and BAL standardization affected GM assay performance.
Copyright © 2011 International Society for Infectious Diseases. Published by Elsevier Ltd. All rights reserved.