Pneumonia is a common disease involving the alveoli and the distal bronchus of the lung tree in acute respiratory infection patients. Over the past few years, the widespread use of broad-spectrum antibiotics has increased the complexity of the drug resistance spectrum of CAP pathogens, rendering the diagnosis and treatment of CAP increasingly challenging. The pathogen diagnosis of severe pneumonia has always been a challenging issue that is especially common in patients with chronic diseases or who have been exposed to antibiotics and use ventilators. It has been demonstrated that the standard methods for detecting pathogenic microorganisms, which involve bacterial smears, cultures, or nucleic acid testing from pharyngeal swabs, sputum, or bronchoalveolar lavage fluid (BALF), exhibit restricted efficacy and are capable of identifying pathogenic microorganisms in only 25–40% of patients(11–13). This means that up to 50% of pneumonia patients have an unknown etiological diagnosis according to traditional etiological methods. Therefore, timely, fast and accurate identification of multiple pathogens remains invaluable for successful management of these patients. Clinicians have been persistently endeavoring to secure positive etiological outcomes, exploring the efficacy and implications of various specimen tests. Research indicates that blood cultures are not advisable even in severe cases of pulmonary infections, as the positive rate is extremely low (14, 15). The guidelines for pneumonia are continuously being updated. The new guidelines have established additional criteria for determining the necessity of conducting lower respiratory tract and blood specimen testing in cases of pulmonary infections. Specifically, the 2019 edition of the American IDSA/ATS guidelines advises against routine BALF and blood culture testing. Instead, it advocates for lower respiratory tract smear and culture, as well as blood culture pathogen detection, only in specific scenarios: for hospitalized patients with severe pneumonia, those at risk of MRSA and Pseudomonas aeruginosa infection (including those with past infections), and patients who have been hospitalized in the last 90 days and have undergone parenteral antimicrobial therapy(16). The updated recommendations in the guidelines are primarily motivated by the fact that cultures fail to alter patient outcomes, the relatively low positive rate of cultures, the potential for contaminated bacteria to contribute to inappropriate antimicrobial use, and the likelihood of prolonged treatment durations due to positive blood cultures. These factors highlight the inherent limitations of traditional pathogen detection methods. Nevertheless, the guidelines on the potential value of mNGS testing are lacking.
This study demonstrated that mNGS offers superior pathogen detection and significantly more positive results in BALF and blood mNGS than traditional cultures of corresponding specimens, which is in line with the findings of Fei Xie et al(10). Even compared with the combined detection methods employing traditional techniques such as BALF culture, PCR, and antibody detection, the pathogen detection rate achieved by mNGS in BALF stands notably higher at 84.5%, surpassing the rate of 26.8%(13). In cases of pulmonary infection, the culture results of BALF have always been greater than those of blood cultures, indicating a significant difference(17). We simultaneously conducted NGS on both BALF and blood samples from 56 patients, and the comparison of the results showed that the percentage of positive BALF samples was significantly greater than that of positive blood samples (80.36% vs. 48.21%). The results of this study also confirm the findings of Xu Chen's research(11). According to the published literature, the percentage of positive mNGS results was as high as 84%, which was greater than that of traditional culture methods for both BALF and blood, and antibiotic exposure did not affect the percentage of positive mNGS results, which was similar to the results of previous studies(18). For patients with pulmonary infection, BALF NGS is an important clinical detection method for identifying pathogenic bacteria.
Given the vast amount of sequence information obtained through mNGS and the diverse nature of pathogen species, the interpretation of these reports can be extremely difficult, increasing the risk of obtaining false positive results(19). To date, there have been no rigorous standards established to differentiate between pathogens detected by mNGS that are truly pathogenic, colonized, or merely false positives. To address this issue, we combined all pathogen detection results with potentially useful information that may aid in the diagnosis of pulmonary infection pathogens, such as serum immune tests, β-glucan tests, galactomannan antigen tests, and many other clinical examinations. Additionally, two deputy directors or higher-ranking doctors have conducted a comprehensive clinical interpretation of all test results to determine the authenticity and reliability of the final pathogen detection results, thereby avoiding false positive rates. To our knowledge, this study is the first to evaluate the diagnostic value of NGS using clinically accepted criteria as the gold standard. In our study cohort, the clinical acceptance rate of mNGS for BALF samples (45/56, 80.36%) was significantly greater than that for blood samples (27/56, 48.21%). Ten patients had negative BALF test results, which was inconsistent with the clinical findings. In one patient, Acinetobacter baumannii was detected but was colonized based on clinical evaluation. When the BALF and blood mNGS results were combined, the clinical acceptance rate (83.93%, 47/56) was greater than that of either method alone. For patients with respiratory infections, the positive rate and clinical acceptance of BALF culture are greater, making it more suitable for the clinical diagnosis of pneumonia. Additionally, this underscores the need for promptly identifying the source of infection in various locations and promptly collecting pathogens from the infected site for corresponding testing in patients with infection. Interestingly, our data revealed for the first time that the rate of positive blood NGS results was greater than that of traditional BALF pathogen detection (48.21% vs 39.58%). Patients with severe pneumonia are prone to bloodstream infections. Blood NGS is an alternative test when BALF specimens are not available. As bronchoscopy for collecting BALF samples is an invasive procedure, we recommend that patients with compromised cardiopulmonary function who may not tolerate this procedure should consider blood NGS testing as an alternative to promptly determine the etiology and minimize the risk of serious complications.
With the increasing clinical application of metagenomic testing, our understanding of the etiology of infectious diseases is likely to include a more comprehensive range of unculturable infectious agents and coinfections. A significant advantage of the metagenomic NGS method lies in its ability to identify numerous potential infectious agents in a single test, leading to improved clinical outcomes. In this study, 24 patients (42.86%) had mixed infections, and the optimal anti-infective treatment plan was adjusted based on these findings. Among the mixed infections, bacterial-fungal infections (44.3%) and bacterial-fungal-viral infections (35.7%) were relatively common(10). This finding suggests that a significant proportion of patients with pneumonia may have complex microbial infections involving the presence of multiple pathogens. Such codetections can provide valuable insights into the pathogenesis of the disease and may inform more targeted therapeutic strategies. Simultaneously, we observed that Legionella, tsutsugamushi disease, and viral reads are more prevalent in blood samples. Conversely, Klebsiella pneumoniae, Pseudomonas aeruginosa, and Acinetobacter baumannii are more readily detectable in BALF.
mNGS still has several limitations. Similar to traditional blood cultures, the yield of mNGS may be affected by the time of sample acquisition. A previous study by Grumaz et al(20). showed that the positive rate of mNGS results remained constant at different time points after sepsis onset, while the positive rate of blood cultures decreased at later time points. Another study also indicated that mNGS is less affected by prior antibiotic exposure. A study by Wang et al(21) demonstrated that mNGS exhibited greater sensitivity than traditional methods in individuals with a higher cumulative steroid dosage, but the positive rate gradually decreased with increasing cumulative steroid dosage. This leads us to speculate that the percentage of positive blood samples may be related to the patient's immune and inflammatory status. The length of hospital stay is also a factor that influences the positive detection rate(22). The results showed that compared to patients with negative blood NGS results, patients with positive blood mNGS results had lower body temperature, lower PCO2, and higher SOFA scores (p = 0.022, 0.028, and 0.038, respectively). These findings suggest that the internal environmental status of patients is affected by blood NGS and that blood NGS should be performed as soon as possible when there is significant organ dysfunction to clarify the etiology. However, there were no significant changes in the serological indicators detected by mNGS of the BALF specimens. This may be related to the fact that this study focused on lung infections, which primarily occur through respiratory tract transmission. The results of BALF may be influenced by factors such as specimen collection. However, because of the low number of reported cases in the literature and the limited number of available studies, further investigations are needed in the future.
This study has the following limitations. First, the sample size was relatively small. To ensure the consistency of paired blood and BALF samples, we excluded patients whose mNGS blood and BALF samples were collected on different days. Second, we did not distinguish between hospital-acquired and community-acquired pneumonia. For hospital-acquired pneumonia, the use of anti-infective agents often reduces the positive rates of both routine testing and mNGS to some extent. Third, considering that it is difficult to determine mixed viral infections, our study analyzed only bacteria and fungi, meaning that viruses that can also cause pneumonia were excluded, which may have affected the sensitivity of NGS. Fourth, as this study was retrospective, we analyzed only the impact of clinical laboratory indicators on NGS results and were unable to delve deeply into the operational impact of specimen collection.
Overall, by comparing and analyzing the etiological test results of blood and BALF samples from patients with lung infections, this study revealed that the etiologies of lung infections are often mixed. Among the current detection methods, BALF mNGS has the highest sensitivity for identifying nonviral pathogens in lung infections. At the same time, we suggest that when BALF samples cannot be obtained, blood NGS can also be a valuable detection method for identifying the etiology, especially when Legionella, tsutsugamushi disease, and viruses are suspected to be the causative agents. Additionally, although we analyzed clinical indicators that may affect negative NGS results, there are certain limitations, and further large-sample studies are needed to increase the rate of positive mNGS results.