Diagnostic value of metagenomic next-generation sequencing of bronchoalveolar lavage fluid for the diagnosis of suspected pneumonia in immunocompromised patients

BACKGROUND: To evaluate the diagnostic value of metagenomic next-generation sequencing (mNGS) of bronchoalveolar lavage fluid (BALF) in immunocompromised patients for the diagnosis of suspected pneumonia in comparison with that of conventional microbiological tests (CMTs). METHODS: Sixty-nine immunocompromised patients with suspected pneumonia received both CMTs and mNGS of BALF were analyzed retrospectively. The diagnostic value was compared between CMTs and mNGS, using the clinical composite diagnosis as the reference standard. RESULTS: Sixty patients were diagnosed of pneumonia including fifty-two patients with identified pathogens and eight patients with probable pathogens. Taking the composite reference standard as a gold standard, 42 pathogens were identified by CMTs including nine bacteria, 17 fungi, 8 virus, 6 Mycobacterium Tuberculosis, and two Legionella and 19(45%) of which were detected by BALF culture. As for mNGS, it identified 76 pathogens including 20 bacteria, 31 fungi, 14 virus, 5 Mycobacterium Tuberculosis, four Legionella and two Chlamydia psittaci. The overall detection rate of mNGS for pathogens were higher than that of CMTs. However, a comparable diagnostic accuracy of mNGS and CMTs were found for bacterial and viral infections. mNGS exhibited a higher diagnostic accuracy for fungal detection than CMTs (78% vs. 57%, P < 0.05), which mainly because of the high sensitivity of mNGS in patients with Pneumocystis jirovecii pneumonia (PJP) (100% vs. 28%, P < 0.05). Nineteen patients were identified as pulmonary co-infection, mNGS test showed a higher detection rate and broader spectrum for pathogen detection than that of CMTs in co-infection. Moreover, Pneumocystis jirovecii was the most common pathogen in co-infection and mNGS have identified much more co-pathogens of PJP than CMTs. CONCLUSIONS: mNGS of BALF improved the microbial detection rate of pathogens and exhibited remarkable advantages in detecting PJP and identifying co-infection in immunocompromised patients. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s12879-022-07381-8.