Performance of galactomannan, beta-d-glucan, Aspergillus lateral-flow device, conventional culture, and PCR tests with bronchoalveolar lavage fluid for diagnosis of invasive pulmonary aspergillosis

Abstract

Galactomannan detection in bronchoalveolar lavage (BAL) fluid samples (GM test) is currently considered the gold standard test for diagnosing invasive pulmonary aspergillosis (IPA). The limitations, however, are the various turnaround times and availability of testing. We compared the performance of GM testing with that of conventional culture, an Aspergillus lateral-flow-device (LFD) test, a beta-d-glucan (BDG) test, and an Aspergillus PCR assay by using BAL fluid samples from immunocompromised patients. A total of 78 BAL fluid samples from 78 patients at risk for IPA (74 samples from Graz and 4 from Mannheim) collected between December 2012 and May 2013 at two university hospitals in Austria and Germany were included. Three patients had proven IPA, 14 probable IPA, and 17 possible IPA, and 44 patients had no IPA. The diagnostic accuracies of the different methods for probable/proven IPA were evaluated. The diagnostic odds ratios were the highest for the GM, PCR, and LFD tests. The sensitivities for the four methods (except culture) were between 70 and 88%. The combination of the GM (cutoff optical density index [ODI], >1.0) and LFD tests increased the sensitivity to 94%, while the combination of the GM test (>1.0) and PCR resulted in 100% sensitivity (specificity for probable/proven IPA, 95 to 98%). The performance of conventional culture was limited by low sensitivity, while that of the BDG test was limited by low specificity. We evaluated established and novel diagnostic methods for IPA and found that the Aspergillus PCR, LFD, and GM tests were the most useful methods for diagnosing the disease by using BAL fluid samples. In particular, the combination of the GM test and PCR or, if PCR is not available, the LFD test, allows for sensitive and specific diagnosis of IPA.