Evaluation of metagenomic next‐generation sequencing diagnosis for invasive pulmonary aspergillosis in immunocompromised and immunocompetent patients

Abstract Background Invasive pulmonary aspergillosis (IPA) can occur in both immunocompromised and non‐immunocompromised hosts, and early diagnosis of IPA is difficult. Metagenomic next‐generation sequencing (mNGS) is a novel non‐migratory pathogen detection method; however, utilising this method for IPA diagnosis is challenging due to the current lack of a unified clinical interpretation standard following Aspergillus detection using mNGS. Objectives To investigate the accuracy of IPA diagnosis by positive bronchoalveolar lavage fluid (BALF) mNGS results in immunocompromised and immunocompetent patients. Methods We retrospectively included patients with confirmed pulmonary infections having a BALF mNGS result of Aspergillus reads ≥1. We compared the accuracy of using mNGS for IPA diagnosis in patients with different immune statuses based on the revised EORTC/MSG criteria. Results Overall, 62 mNGS Aspergillus ‐positive patients were divided into two groups: with (41) and without IPA (21). In univariate logistic regression analysis, immunocompromised function, fever, halo sign on CT image, and multiple masses or nodules were associated with mNGS Aspergillus ‐positive IPA diagnosis. In multivariate logistic regression analysis, immunocompromised function (OR = 6.68, 95% CI: 1.73–25.87, p = .006) and a halo sign (OR = 7.993, 95% CI: 2.07–30.40, p = .003) were independent risk factors. The concordance rate of IPA diagnosis was significantly higher in immunocompromised patients [82.1% (23/28)] than in non‐immunocompromised patients [52.9% (18/34); p = .016]. Conclusions For immunocompromised patients, a combination of mNGS testing and lung CT imaging can be used for IPA diagnosis. However, caution is required in IPA diagnosis based on positive mNGS results in non‐immunocompromised patients.