Prospective case‐control study of enterovirus detection differences in children’s cerebrospinal fluid between multiplex PCR and real‐time RT‐PCR assay

Abstract Background Viral encephalitis is common in childhood. It is an acute brain parenchymal inflammation caused by a variety of viral infection, and enterovirus accounts for the majority. Due to atypical clinical manifestations, pathogenic testing is important for assisting clinical diagnosis. The purpose of this study was to evaluate the performance of the multiplex PCR assay compared with quantitative real‐time PCR for enterovirus detection. Methods A prospective case‐control study was performed involving 103 pediatric patients suspected for viral encephalitis and cerebrospinal fluid (CSF) samples were collected and tested for 9 pathogens using multiplex PCR assay during April to November in 2018. In parallel, an aliquot of samples was tested for enterovirus infection by real‐time PCR assay. Results There were 85.4% children were confirmed as viral encephalitis on discharge, the remaining ones were diagnosed as other CNS diseases, such as epilepsy. The specificity of the two methods was the same as that of the clinical diagnosis, but the sensitivity and consistency with clinical diagnosis of multiplex PCR were both higher than the real‐time PCR. Besides of enterovirus, multiplex PCR could also detect coinfection of enterovirus with Epstein‐Barr virus and mumps virus. Conclusion Results of multiplex PCR method are more consistent with the clinical diagnosis and are superior to real‐time PCR for detecting enterovirus in CSF.


| Study samples
Obtained by lumbar puncture, CSF samples were collected and used for routine CSF biochemical tests and culture, the remaining samples were stored at −80°C for molecular analysis.
After treatment and observation, patients who were finally diagnosed with other central nervous diseases were enrolled into the control group to evaluate the molecular diagnostic assay.

| DNA/RNA extraction
A total of 200 µL CSF samples were used to extracted and purified nucleic acid by extraction kit (HGT, Ningbo, China) on an automated extraction workstation Smart LabAssist-16/32 (TANBead, Taiwan, China). The extracts were immediately used as template for PCR amplification or stored below −20°C.

| Detection of pathogens by multiplex PCR and qPCR
The one-step RT-PCR was fulfilled with the ABI Verity 96 Thermal Cycler. The PCR products were added to a 96-well plate, prepared for capillary electrophoresis (CE), and fragment analysis by applying

| Statistical analysis
Chi-square test was used on the SPSS 13.0.1 statistics package (SPSS Inc, Chicago, USA). Agreement of the results between molecular assay and discharge diagnosis was assessed using Kappa statistics (κ value 0.21-0.4 fair, 0.41-0.6 moderate, 0.61-0.8 substantial, and 0.81-1 almost perfect). 10 P < .05 was considered statistically significant.

| Study population
A total of 103 CSF specimens enrolled in this study (

| Clinical concordance with multiplex PCR and real-time PCR
A moderate agreement (κ value = 0.447) was observed between the discharge diagnosis and multiplex PCR results, but a fair agreement (κ value = 0.329) was observed between the discharge diagnosis and real-time PCR. In the CSF from one case, both multiple PCR and RT-PCR tests showed positive EV, but the discharge diagnosis was epilepsy. In addition, 21 and 30 cases were diagnosed as viral encephalitis without certain pathogen detection using these two methods, respectively ( Table 3). The false-negative rate shown by multiplex PCR was significantly lower than that of real-time PCR (P = .013).

| Co-detection by multiplex PCR
Besides of EV, multiplex PCR assay also identified other viruses, including 1 EBV and 4 MuV. Seven mixed infections (EV and EBV) were also identified by multiplex PCR (Table 4).

| D ISCUSS I ON
Nucleic acid amplification techniques (NAATs) such as real-time PCR and multiplex PCR have been widely used to identify pathogens in infectious central nervous system diseases. 11,12 These NAATs prevent misdiagnosis in children with normal cellularity, normal protein levels, or without hypoglycorrhachia, whose CSF PCR tested positive for EV. 13,14 These data highlight the need to perform PCR in CSF of children despite the normal results of the traditional tests. However, Only a few reports have described the performance of NAATs in microbiological testing in pediatric patients suspected of viral encephalitis. 6,15,16 The clinical application of single-targeted NAATs is limited due to the insufficient CSF volume and small number of detection channels. On the other hand, methodological studies indicated that multiplex PCR may be less sensitive than the corresponding single-targeted real-time PCR due to the imbalance in amplification efficiency between diverse targets. 8 Therefore, it is necessary to use CSF from children with viral encephalitis to compare the differences between the two methods.
In this study, we compared the detection of EV in 103 CSF speci- Febrile convulsion 2 1.9 Purulent meningitis 2 1.9 Neurosis 2 1.9 Autoimmune encephalitis 2 1.9 Intracranial hypertension 1 1.0 Systemic inflammatory response syndrome 1 1.0 Central nervous system demyelination 1 1.0 Acute tonsillitis 1 1.0 this "false-negative" may be that the encephalitis is caused by a pathogen other than the target in test kit. Alternatively, the concentration of pathogens in CSF may be too low to permit detection. In addition to false-negative cases, the false-positive ones were also observed, as one patient was diagnosed with epilepsy and both multiplex PCR and real-time PCR showed positive enterovirus results. It is well known that CNS infection is the main risk factor for epilepsy. 22 Approximately 42% of infants with enterovirus infection present with severe seizures. 23 When status epilepticus is accompanied by encephalitis, the prognosis is worse than etiologies infection, 24 Therefore, for such patients, it is more necessary to understand the pathogens in CSF.
Furthermore, in our and others' studies, the presence of mixed pathogen is remarkable. We observed 6.8% (7/103) coinfection as EV and EBV, and Kahraman et al found that 9.1% (3/33) CSF samples were simultaneously positive for 2 pathogens. 25 Shin et al found a case was positive for L monocytogenes and EBV by multiplex PCR, but only positive for L monocytogenes by conventional PCR. 21 These data suggest that multiplex PCR methods may increase the isolation rate of pathogens in central nervous system infections. Further research is needed to investigate the clinical relevance of this coinfection result.

| CON CLUS IONS
EV was the most identified virus causing meningitis in children. It is needed to applicate viral PCR testing in clinical. In this study, we observed a higher sensitivity and a higher consistency of clinical diagnosis of multiplex PCR compared with single-target real-time PCR.
The results of rapid multiplex PCR testing can be used to guide antimicrobial therapy and may result in reduced antimicrobial exposure in children with viral encephalitis.

ACK N OWLED G M ENTS
We sincerely thank the parents and children who volunteered to participate in this study. The study would not have been possible without the excellent support from clinical staff from the No.1 Neurology Department at our hospital of Hebei Province.

CO N FLI C T O F I NTE R E S T S
The authors declare that they have no competing interests.

AUTH O R S' CO NTR I B UTI O N S
LW and SZS designed the study and take responsibility for the entire process; DPY and FC conducted literature search, data extraction, quality assessment, and draft writing; JJL, YHG, and FY collected and analyzed the data; XPZ and WJW edited the article. All authors have read and approved the final article.

E TH I C S A PPROVA L A N D CO N S E NT TO PA RTI CI PATE
The study was approved by the Children's hospital Hebei Province Ethics Committee (number 2 018 002). The legal guardian(s) or parent(s) of the children provided written informed consent for sample collection and clinical record review.

DATA AVA I L A B I L I T Y S TAT E M E N T
The datasets generated and/or analyzed during the current study