Application of targeted metagenomic next-generation sequencing in pneumonia patients

Abstract

After the coronavirus disease 2019 (COVID-19) pandemic, the incidence rate of mixed infections, especially in critically ill patients with severe pneumonia, increases due to the immunity gap and has also been proven to be associated with mortality, increasing the difficulty in accurate and rapid diagnoses. Here, we evaluated the performance of targeted metagenomic next-generation sequencing (mNGS), including capture hybridization-based mNGS (chNGS) and multiplex PCR-based targeted mNGS (tNGS), in diagnosing pneumonia. Patients admitted to the Pulmonary and Critical Care Medicine, Beijing Aerospace General Hospital, and diagnosed as suspected pulmonary infections from April 2022 to March 2024 were retrospectively evaluated, and 110 patients were finally enrolled. According to the final comprehensive clinical diagnoses, there were 99 patients diagnosed with definite infectious diseases. Single infections accounted for 58.6% of these patients (58/99), while mixed infections occurred in about half of these patients (41.4%, 41/99) and were found in most of the death cases. Pseudomonas aeruginosa (n = 22), Pneumocystis jirovecii (n = 15), and severe acute respiratory syndrome coronavirus 2 (n = 13) were found to be the most common bacterial, fungal, and viral pathogens, respectively. Taking final comprehensive clinical diagnoses as the reference standard, the total coincidence rate (TCR) of chNGS can reach up to 64.1% (95% confidence interval [CI], 54.8%-73.3%), while the TCR of conventional methods was only 39.8% (95% CI, 30.4%-49.3%). The performance of tNGS was slightly superior to that of chNGS, while chNGS yielded more false-negative results, especially for viral detection. Additionally, chNGS combined with tNGS can improve the TCR to 81.3% (95% CI, 62.1%-100.0%).IMPORTANCEThis is the first report on evaluating the performance of capture hybridization-based metagenomicnext-generation sequencing (chNGS), multiplex PCR-based targeted mNGS (tNGS), and conventional methods in diagnosing pneumonia. Our findings emphasized the importance of chNGS and tNGS in diagnosing, managing, and ruling out infections, and an era of widespread application of regional tNGS in monitoring and diagnosing infections with high sensitivity and low economic burden on patients can be expected.

Keywords: capture hybridization; mixed infections; multiplex PCR; pneumonia; targeted metagenomic next-generation sequencing.

Fig 1

Flow diagram.

Fig 2

Infection type and pathogen profiles. (a) Sample types used for pathogen detection. (b) Different types of infections found in pneumonia. (c) Pathogen profiles. The number on the right side of the pathogen bar is the number of patients with the detection of the pathogen. SARS-CoV-2 represents severe acute respiratory syndrome coronavirus 2.

Fig 3

Performance comparison between chNGS and CTM. (a) Positive rates of chNGS and CTM. (b) Summaries of the number of cases with positive or negative results revealed by chNGS and CTM. “+” and “−” represent positive and negative results, respectively. (c) Performance of chNGS and CTM. PPV, NPV, and TCR represent positive predictive value, negative predictive value, and total coincidence rate, respectively.

Fig 4

Diagnostic value of chNGS and tNGS from the perspective of pathogens. The number on the pathogen bar is the number of the patients with true-positive or false-positive results.

Fig 5

Mixed infections and final outcomes of patients with different CURB-65 scores. Based on the CURB-65 scores, we divided the enrolled patients into mild (CURB-65 scores of 0–1), moderate (CURB-65 score of 2), and severe (CURB-65 scores of ≥3) groups. (I) Infection types in mild (a), moderate (b), and severe (c) groups. (II) Final outcomes in mild (a), moderate (b), and severe (c) groups.