Clinical utility of metagenomic next-generation sequencing in suspected central nervous system infectious pediatric patients with empirical treatment: a cohort study

Background: Metagenomic next generation sequencing (mNGS) is becoming an increasingly available diagnostic method used to identify a broad range of pathogens. However, the optimal role of mNGS in clinical diagnostic schema remains uncertain, especially in pediatric patients suspected central nervous system (CNS) infection and treated with empirical antibiotic. The purpose of this study was to investigate the usefulness of cerebrospinal uid (CSF) mNGS in the pediatric patients. Methods: We performed a retrospective review of suspected CNS infection patients who had CSF mNGS test from April 2019 to December 2020. Results and clinical impact of mNGS test were collected. We investigated the usefulness of CSF mNGS in clinical impact and diagnosis. Results: We enrolled 57 pediatric patients with empirical treatment. A total of 39 CNS infection patients were diagnosed, and 27 patients were identied by mNGS that only 2 of 27 were co-identied by CSF culture. In all of the patients, 75.4% (45 cases) had changed (addition or de-escalation) in antimicrobial therapy according to the results of the mNGS test. By each CSF mNGS test as a whole, the positive and negative percent agreement were 69.2% (95% CI: 54.1-84.4%) and 61.1% (95% CI: 36.2-86.1%), respectively, and true negatives of negative mNGS tests were 50% (95% CI: 27.3 – 72.7%). Conclusion: In this study, CSF mNGS test improved the diagnosis of neurologic infections and adjusted antibiotic therapy in the vast majority of cases. Consequently, for patients with empirical treatment, CSF mNGS should be used more in pathogen diagnosis and clinical therapy.


Background
Central nervous system (CNS) infections are a cause of morbidity and mortality in children. Estimated incidence of acute encephalitis syndrome in children is 10.5/100000 [1]. For these patients, effective treatment is critical based on timely identi cation of causative pathogens [2]. But the pathogen of meningitis is not identi ed in approximately 50%-60% of patients [3][4][5]. The most commonly method for pathogen detection is cerebrospinal uid (CSF) culture. Traditional CSF culture can only identify approximately 30-40% of CNS infections [6]. Other methods for pathogen detection (such as tissue biopsy, Xpert MTB/RIF Ultra, loop-mediated isothermal ampli cation, and Filmarray meningitis/encephalitis Panel) can improve diagnostic ability [6,7], but they are restricted to a limited spectrum of pathogens. Routine microbiologic testing is often de cient to detect all neuroinvasive pathogens. Recent reports found that metagenomic next generation sequencing (mNGS) can be used to identify pathogens in cerebrospinal uid samples which were not found by conventional microbiological methods [8][9][10].
mNGS is a promising microbial detection technology for the diagnosis of infectious disease [11], because a wide-ranging spectrum of potential causes -bacterial, viral, fungal, and parasitic -can be independent identi ed by a single assay [12,13]. To date, several studies have proved the value of clinical CSF mNGS for nding out pathogens, improving diagnosis of Meningitis and Encephalitis [8,14,15]. However, it is unclear that the diagnostic performance of mNGS in pediatric patients with suspected CNS infection and empirical treatment. Therefore, we conducted a pediatric retrospective study to explore the clinical impact and performance of mNGS in CNS infection patients.

Patients and data
This study was a single-center, retrospective review of all patients with CSF mNGS testing at Children's Hospital of Hebei Province from April 5, 2019 to December 5, 2020. The Children's Hospital is a large tertiary teaching hospital. Most of the patients were transferred from other hospitals. Patients were enrolled based on a particular exposure (i.e., suspected CNS infectious patients without pathogenic evidence and received empirical treatment (Fig. 1)). Patients who were older than 18 years old were excluded. The patients' CSF was collected when a lumbar puncture was performed after admission. All of the patients' details were obtained from medical records. Clinical details were obtained, including demographic information, symptoms, clinical diagnosis, C-reactive protein (CRP), complete blood count, examinations of cerebrospinal uid (routine biochemical, culture, smear, PCR, antibody, mNGS), and treatment. Details about CSF mNGS test (BGI China) including results, interval (in days) from illness onset to CSF collection, lengths of empirical treatment (duration of antimicrobial use before the mNGS test), turnaround time (time from collection to return result), and clinical impacts (including new antimicrobial, antimicrobial de-escalation, no change) were evaluated. Clinically relevant organism identi ed from CSF mNGS was assessed relative to nal overall diagnosis (CNS infection versus non-CNS infection). All information was collected by two independent researchers. Any discrepancies between the two researchers were resolved by consensus or, if necessary, a third-party clinician adjudicated by consulting the medical record. All of patients were classi ed into two groups in the nal diagnosis: CNS infection group and non-CNS infection group. mNGS test was de ned as positivity if it identi ed an organism. Otherwise, the mNGS test was de ned as negative. In this way, CNS infection patients were divided into two groups (mNGS positive group and mNGS negative group). This retrospective study was approved by Children's Hospital of Hebei Province ethics committee (Approval number: 2018014). A signed, written informed consent form was attained from the patient's guardian. All methods were performed in accordance with the relevant guidelines and regulations. mNGS of CSF CSF specimens (0.5ml) were collected from patients according to standard aseptic procedures, snap frozen, stored at −20•C, and subjected to mNGS within 24h. The process of CSF mNGS consisted sample processing, nucleic acid (DNA and RNA) extraction, followed by library generation and bioinformatic pipeline analysis. After samples were received in the clinical laboratory, sample were processed by adding glass beads to CSF samples, followed by vigorous agitation. Then extracted DNA/ RNA was fragmented to yield 150-bp to 200-bp fragments, and DNA fragments were constructed through an end-repair method. Quality-controlled libraries were sequenced on a BGISEQ-500/50 platform (BGI-Tianjin, Tianjin, China).
For each sample, an average of 20 million reads was obtained. High-quality sequencing data were generated by removing low-quality and short (length <35 bp) reads, after that computational subtraction of human host sequences were mapped to a human reference genome (YH sequences and hg19) using Burrows-Wheeler alignment. Nonhuman sequences were mapped to categorization reference databases downloaded from NCBI (ftp://ftp.ncbi.nlm.nih.gov/genomes/), which included the genome sequences of 3446 bacterial species (45 species of mycoplasma/chlamydia and 104 species of Mycobacterium tuberculosis), 206 fungal species, 1515 viral species, and 140 parasites connected to human diseases.

Diagnostic assessment of mNGS
De nitive clinical diagnoses of the participants who completed the study were adjudicated by retrospective. The diagnosis of CNS infection was done according to the diagnosis of CNS infections written by Singhi et al [16]. The diagnosis of autoimmune encephalitis was carried out in accordance with the clinical diagnosis of autoimmune encephalitis published by Graus et al [17]. The diagnosis of epilepsy was made in accordance with the clinical diagnostic criteria for epilepsy [18].
The clinical utility of mNGS was evaluated through the following steps. Firstly, clinical characteristics were compared between CNS infections patients with positive and negative mNGS groups. Second, to evaluate whether mNGS test had a clinical impact. Clinical impact was de ned as the following conditions: 1) added new targeted antimicrobial, 2) de-escalation of antibiotic therapy, according mNGS test results. Finally, diagnostic performance of mNGS results was calculated on the basis of the following two approaches [19]. In the absence of a gold standard for mNGS results, positive percent agreement (PPA: agreement between mNGS test and CNS infection diagnosis) and negative percent agreement (NPA: agreement between negative mNGS test and diagnosis of non-CNS infection) were reported instead of sensitivity and speci city. In addition, the proportion of true positives out all positive mNGS ndings and proportion of true negatives out of all negative mNGS tests were described. mNGS tests often identi ed two or more organisms, which can be divided into two categories: clinically relevant and clinically irrelevant organisms depending on whether the organism had been recognized as causative agents. One method we counted one CSF sample as one test. An organism in one CSF sample test was clinically relevant, mNGS test was de ned as true positive. Another method was adopted: each organism identi ed was counted and assessed independently. This method provides more detail for mNGS ndings by separately assessing each organism.

Patient Characteristics
Between April 5, 2019, and December 5, 2020, a total of 60 patients were screened in this study (Fig. 1) Table 2).   Fig S2). Table 3 summarizes the clinical impact of each mNGS test. In all of the patients, 75.4% (45 cases) had changed (addition or deescalation) in antimicrobial therapy. In the mNGS positive and negative group, the proportions of clinical impact were 88.6% and 54.5% respectively.  Table S1). One mNGS test identi ed only clinically irrelevant organisms (Additional le: Table S2), and three mNGS tests showed both clinically relevant and irrelevant organisms (Additional le: Table S4). When each organism identi ed was analyzed independently, the positive and negative percent agreement of mNGS were 76.2% (95% CI: 65.4-87.0%) and 45.8% (95% CI: 24.3-67.3%), respectively, and the proportion of true positives out of all mNGS positives was 73.8% (Table 5).

Discussions
In this study, we evaluated the test performance characteristics of mNGS for diagnosing neurologic infections in a series of pediatric patients with empirical treatment before the time of enrollment. And we also assessed the clinical usefulness of mNGS. Thus, we described the real-life performance of CSF mNGS test in a di cult-to-diagnose pediatric patient for whom the essay is most likely to be performed. We assessed the clinical relevance of each mNGS test and each organism detected by mNGS, which substantially impacted the proportion of PPA for diagnosing neurologic infections (69.2% for per-test assessment versus 76.2% for per-organism assessment). In this study, the mNGS assay identi ed most potential pathogens in CNF infection patients. A total of 26 infections were diagnosed solely by metagenomic NGS. It is notable that 43 of these 57 cases had a clinical effect. These ndings demonstrated the potential usefulness of CSF mNGS test in these pediatric patients.
This study indicated that mNGS held diagnostic advantages over CSF culture for pediatric patients who had received empirical antimicrobial treatment. As there is currently no gold standard for mNGS results, PPA and NPA were reported instead of sensitivity and speci city. The clinical laboratory validated that PPA and NPA of CSF mNGS were > 95%, > 80% respectively, compared to the clinical microbiologic results [20]. Our results are consistent with previous literature reports, which showed that mNGS identi ed multiple pathogens successfully, and demonstrated that mNGS was most useful in patients with a CSF abnormality [15]. Additionally, an adult study found that mNGS provided a higher detection rate compared to culture in patients with empirical therapy in suspected CNS infection, and they indicated that empirical antibiotics use did not affect the detection rate of mNGS [21]. However, effective treatment could reduce the detection rate of CSF culture in meningitis (11%) [22]. The possible reasons for the above phenomenon are explained as follows. Culture needs the existence of livable pathogens and therefore is easily affected by the administration of antimicrobial treatment. On the other hand, mNGS requires only DNA fragments of microorganisms, which might be unaffected by treatment. Nevertheless, low concentrations of microorganisms may lead to false-negative results of mNGS [8]. This may also account for this difference observed in our studies, although we had no information on the organism concentrations. Our data found that CNS infection patients in the mNGS negative group had used antibiotics for a long time and had low in ammation markers like CRP and CSF WBC.
The diagnosis of this study exposed the challenge of data interpretation. In our study, three cases found clinically relevant and irrelevant organism, one case only found clinically irrelevant organism, and seven cases found false positive of mNGS in non-CNF infection (Additional le: Table S1). Autoimmune encephalitis was mainly established by clinical presentation, antibodies (CSF HSV PCR, VZV PCR, EBV PCR, anti-NMDAR antibody, anti-MOG antibody), and the response of treatment. Post-herpetic or EBV or other organisms. Autoimmune encephalitis could arise. A recent adult case report found that the anti-NMDAR antibody (IgG) was identi ed positively on day 11 after herpes simplex virus infection [23]. This possibility must be recognized by clinicians, especially those with symptoms that occurred after remission [24][25][26].
Our results suggest that mNGS should be an effective tool to guide clinical decisions for antibiotic treatment. When mNGS identi ed additional organisms or none organism, changes in therapy occurred in the majority (75.4%) of cases. Wilson et al. reported that in 13 patients identi ed by mNGS, 7 cases' medical treatment was adjusted according to mNGS results [8]. The results of mNGS can also be valuable even when no organism was detected, about half of the patients with mNGS negative (12/22) had antimicrobial de-escalation, considering ruling out co-infections.
There were some limitations in our study. First, our study had a relatively small sample size. This may be the cause of that many results observed a certain trend did not achieve statistical signi cance. Consequently, a larger sample size research needs to be conducted in further. Second, in these patients, medication dosage and time were not uniform. The inconsistent empirical treatment may come from different hospitals in the region. Thus, it appears that mNGS is advantageous in a real-world healthcare setting. Third, this is a single central retrospective study, in which selective bias in data collection, limit generalizability, and lack of antibiotic usage detail may restrict the wide application of our results in other hospitals.

Conclusion
Page 10/13 In summary, our major ndings included performance characteristics of CSF mNGS test with 76.2% of the organisms identi ed as clinically relevant and clinical effects with 75.4% cases (45/57) who had changed (addition or de-escalation) in antimicrobial therapy. Therefore, we suggest that mNGS should be used more in suspected CNS infection patients with antibiotic treatment in the future. Nonetheless, there is still a challenge to interpret data of mNGS for pediatric doctors in guiding clinical treatment. Despite the insights in our study regarding CSF mNGS test performance and utility, further research will be vital to recognize how to optimally integrate mNGS into the CNS infectious diseases diagnostic work up.

Availability of data and materials
The data sets generated and/or analyzed during the current study are not publicly available due to the data privacy requirements of the ethics committees, but are available from the corresponding author on reasonable request and approval from the ethics committees in all institutions.