Abstract
INTRODUCTION: Accurate diagnosis of Pneumocystis jirovecii pneumonia (PJP) in HIV patients remains challenging. This study compares metagenomic next-generation sequencing (mNGS) with PCR, GMS staining, and serum β-D-glucan (BG) assays for PJP detection and co-infection identification. METHODS: BALF samples from 34 HIV-positive PJP patients and 50 non-PJP controls were analyzed. Diagnostic performance metrics (sensitivity, specificity, NPV, AUC) and co-pathogen profiles were evaluated for mNGS versus conventional methods. RESULTS: mNGS and PCR both achieved 100% sensitivity. mNGS showed higher specificity (91.3% vs. 88%) and AUC (0.898 vs. 0.940 for PCR). Co-infections were detected in 67.6% of PJP cases by mNGS, including cytomegalovirus (41.2%), Epstein-Barr virus (29.4%), and non-tuberculous mycobacteria (14.7%). GMS and BG assays exhibited lower sensitivity (64.7% and 76.5%, respectively). DISCUSSION: mNGS offers superior specificity, accuracy, and co-infection detection compared to traditional methods. Its high NPV (100%) supports clinical utility in ruling out PJP. While resource-intensive, mNGS is a promising first-line diagnostic tool for HIV-associated PJP, particularly in polymicrobial infection settings.