Work

The diagnostic and prognostic utility of blood metagenomic next-generation sequencing for invasive pulmonary aspergillosis

Year: 2026

Type: article

Abstract: ABSTRACT Differentiating invasive pulmonary aspergillosis (IPA) from colonization in patients with Aspergillus -positive blood metagenomic next-generation sequencing (mNGS) remains a clinical challenge. This study aims to evaluate the diagnostic and prognostic value of blood mNGS-derived fungal load (reads per million [RPM]) and two key serological biomarkers (galactomannan [GM] and 1,3-β-D-glucan [BDG]) in distinguishing these two entities. This retrospective study enrolled 95 patients with Aspergillus detected by blood mNGS, stratified into infection ( n = 60) and colonization ( n = 35) groups using modified EORTC/MSGERC criteria. We analyzed clinical characteristics, co-infection spectra, and serological biomarkers (GM and BDG). Diagnostic performance was evaluated via receiver operating characteristic (ROC) analysis, and prognostic factors for 28-day mortality were identified using least absolute shrinkage and selection operator-Cox regression. Distinct co-infection patterns were observed between groups: the infection group was dominated by polymicrobial co-infections, including clinically significant pathogens such as Acinetobacter baumannii , Klebsiella pneumoniae , Mucor spp., and Human cytomegalovirus ; in contrast, the colonization group primarily featured single viral co-infections. While mNGS effectively detected Aspergillus , RPM alone showed limited ability to discriminate infection from colonization, with area under the curves (AUCs) ranging from 0.406 to 0.657 across patient groups. The optimal RPM cutoff varied substantially by immune status, being highest in immunocompetent patients (RPM cutoff: 1.77). Diagnostic performance significantly improved when RPM was integrated with GM (AUC up to 0.900 at a cutoff of 0.36 optical density index) or BDG (AUC up to 0.881 pg/mL), particularly in immunocompetent individuals. RPM also correlated with albumin, hemoglobin, platelet counts, and lymphocyte counts (all P < 0.05). Multivariate analysis identified reversed halo sign (hazard ratio [HR] = 2.143), decreased ratio of partial pressure of arterial oxygen to fraction of inspired oxygen (PaO 2 /FiO 2 ; HR = 1.361), and elevated lactate dehydrogenase (HR = 1.055) as independent predictors of 28-day mortality. Blood mNGS demonstrates high sensitivity for detecting Aspergillus but requires integration with serological biomarkers to differentiate IPA from colonization. The RPM can offer prognostic utility. A multimodal strategy is crucial for early diagnosis and improving outcomes in high-risk patients. IMPORTANCE First large-scale validation of blood mNGS for invasive pulmonary aspergillosis diagnosis—this study represents the first sizable cohort systematically evaluating blood metagenomic next-generation sequencing (mNGS) for distinguishing invasive pulmonary aspergillosis from colonization, addressing a critical gap in non-invasive diagnostic approaches for critically ill patients. Comprehensive Aspergillus co-infection profiling—we identified distinct co-infection patterns, with the infection group showing significantly higher rates of polymicrobial infections, providing crucial insights into co-infection dynamics in Aspergillosis . Optimized diagnostic integration strategy—our findings demonstrate that while mNGS-derived reads per million alone show limited diagnostic value, their integration with serological biomarkers significantly improves performance, establishing a clinically relevant multimodal diagnostic framework. Robust prognostic stratification model—through least absolute shrinkage and selection operator-Cox regression, we established a validated prognostic model identifying reversed halo sign, decreased PaO 2 /FiO 2 , and elevated lactate dehydrogenase as independent predictors of 28-day mortality, providing clinically actionable tools for risk stratification.

Source: Microbiology Spectrum

DOI: 10.1128/spectrum.03384-25