Retrospective Multicenter Evaluation of the VirClia Galactomannan Antigen Assay for the Diagnosis of Pulmonary Aspergillosis with Bronchoalveolar Lavage Fluid Samples from Patients with Hematological Disease

Abstract

Galactomannan (GM) testing of bronchoalveolar lavage (BAL) fluid samples has become an essential tool to diagnose invasive pulmonary aspergillosis (IPA) and is part of diagnostic guidelines. Enzyme-linked immunosorbent assays (ELISAs) (enzyme immunoassays [EIAs]) are commonly used, but they have a long turnaround time. In this study, we evaluated the performance of an automated chemiluminescence immunoassay (CLIA) with BAL fluid samples. This was a multicenter retrospective study in the Netherlands and Belgium. BAL fluid samples were collected from patients with underlying hematological diseases with a suspected invasive fungal infection. Diagnosis of IPA was based on the 2020 European Organisation for Research and Treatment of Cancer (EORTC)/Mycoses Study Group Education and Research Consortium (MSGERC) consensus definitions. GM results were reported as optical density index (ODI) values. ODI cutoff values for positive results that were evaluated were 0.5, 0.8, and 1.0 for the EIA and 0.16, 0.18, and 0.20 for the CLIA. Probable IPA cases were compared with two control groups, one with no evidence of IPA and another with no IPA or possible IPA. Qualitative agreement was analyzed using Cohen’s k, and quantitative agreement was analyzed by Spearman’s correlation. We analyzed 141 BAL fluid samples from 141 patients; 66 patients (47%) had probable IPA, and 56 cases remained probable IPA when the EIA GM result was excluded as a criterion, because they also had positive culture and/or duplicate positive PCR results. Sixty-three patients (45%) had possible IPA and 12 (8%) had no IPA. The sensitivity and specificity of the two tests were quite comparable, and the overall qualitative agreement between EIA and CLIA results was 81 to 89%. The correlation of the actual CLIA and EIA values was strong at 0.72 (95% confidence interval, 0.63 to 0.80). CLIA has similar performance, compared to the gold-standard EIA, with the benefits of faster turnaround because batching is not required. Therefore, CLIA can be used as an alternative GM assay for BAL fluid samples.