A novel Aspergillus galactomannan chemiluminescence immunoassay: Evaluation for diagnosis of invasive pulmonary aspergillosis in a multicenter controlled study

Abstract

The Aspergillus galactomannan chemiluminescence immunoassay (GM–CLIA) is a novel non-culture-based test applied in the detection of GM antigen in serum and bronchoalveolar lavage fluid (BALF). This multicenter, retrospective, and controlled study aims to investigate its diagnostic value for invasive aspergillosis (IA). A total of 246 serum and 228 BALF samples were obtained from IA patients (n = 229) and non-IA (n = 236) patients between March 2023 and July 2024. The performance of GM-CLIA was compared with that of Aspergillus GM sandwich enzyme immunoassay (GM-EIA) through Spearman rank correlation analysis and assessed via sensitivity, specificity, positive/negative predictive values (PPV/NPV), and receiver operating characteristic (ROC) curve analysis. Analytical performance was evaluated through interference, precision, limit of detection (LoD), and cross-reactivity experiments. The sensitivity, specificity, PPV, and NPV of serum GM–CLIA were 93.97%, 97.46%, 97.32%, and 94.26%, respectively. For BALF GM–CLIA, the corresponding values were 93.58%, 95.76%, 95.33%, and 94.17%. ROC analysis yielded high area under the curve (serum: 0.977; BALF: 0.982) with optimal cut-offs of 0.625 (serum) and 0.615 (BALF). The levels of GM by CLIA and EIA (rs = 0.838, P < 0.0001) strongly correlated. No interference was observed from hemoglobin, bilirubin, or triglycerides at certain concentrations. Repeatability, indoor, and inter-batch precision met predefined criteria, with no cross-reactivity in non-Aspergillus pathogens. The LoD for detecting GM–CLIA was 0.09 ng/ml. In summary, GM–CLIA was evaluated in both sample types, demonstrating high specificity and sensitivity in diagnosing IA. Thus, clinical implementation of CLIA may significantly improve diagnostic efficiency and patient outcomes.