ABSTRACT
Objective
Central nervous system (CNS) infections remain a significant cause of morbidity and mortality in children. This study aimed to evaluate the impact of cerebrospinal fluid (CSF), multiplex polymerase chain reaction (PCR) panel results on clinical decision-making and patient management in children who underwent lumbar puncture (LP) with a preliminary diagnosis of meningitis/meningoencephalitis.
Methods
Patients aged 1 month to 18 years who underwent LP for suspected CNS infection in our pediatric emergency or intensive care units between 2018 and 2023, and who had a CSF multiplex PCR meningitis/encephalitis panel performed, were retrospectively evaluated in terms of demographics, clinical presentation, laboratory parameters, and treatments. Patients younger than 1 month or older than 18 years, those who underwent LP for non-infectious indications, and those with ventriculoperitoneal shunts were excluded. Data were analyzed using SPSS version 24.
Results
The median age of the 144 patients was 2.7 (6.7) years, and 93 (64.6%) were male. At least one pathogen was detected by multiplex PCR in 35 patients (24.3%). Of these, 22 had viral agents (enterovirus in 9, HSV-1 in 4, HHV-8 in 2, HHV-7 in 2, VZV in 2, CMV in 2, and HHV-6 in 1), 11 had bacterial agents [Streptococcus pneumoniae (S. pneumoniae) in 7, Neisseria meningitidis in 3, and Haemophilus influenzae type b (Hib) in 1], and 2 had multiple agents (S. pneumoniae + Hib + HHV-6 in one case; enterovirus + HHV-6 in one case). No significant clinical differences were observed between viral and bacterial infections. In 51 patients (35.4%), treatment was modified based on PCR results, most often by discontinuing acyclovir (22.1%), antibiotics (7.6%), or both (3.5%).
Conclusions
In approximately one-third of cases, unnecessary antiviral or antibiotic treatments were discontinued based on PCR results, demonstrating the utility of molecular diagnostics in guiding clinical management. Especially in patients who had received antibiotics prior to LP, early pathogen detection via PCR may help reduce treatment costs, complications, and length of hospital stay.
INTRODUCTION
Central nervous system (CNS) infections are defined as inflammation of the meninges or brain parenchyma due to infectious causes, and they remain among the leading causes of morbidity and mortality in children. With the inclusion of conjugated Streptococcus pneumoniae and Haemophilus influenzae type B (Hib) vaccines in routine immunization programs, viruses have become the most common cause of meningitis worldwide, including in Türkiye1. Most cases of viral meningitis except those caused by herpes simplex virus (HSV) generally follow a benign clinical course and are treated symptomatically. Although the incidence of bacterial meningitis has decreased, it remains a serious disease because of the risk of irreversible neurological damage or death if not promptly treated, making it a medical emergency.
It is often difficult to clinically distinguish between viral and bacterial CNS infections at presentation. Both types of meningitis can manifest with similar symptoms in children, such as fever, headache, photophobia, and neck stiffness. In neonates and infants, the typical signs of meningitis may be absent. Furthermore, commonly used diagnostic parameters like cerebrospinal fluid (CSF) pleocytosis and acute-phase reactants [e.g., C-reactive protein (CRP) and white blood cell count] are insufficient to reliably distinguish viral from bacterial meningitis2. The inability to determine the etiology early often leads to prolonged and potentially unnecessary use of antibiotics and antivirals, as well as extended hospital stays.
Molecular diagnostic tests based on polymerase chain reaction (PCR) -particularly multiplex PCR panels- allow rapid and reliable identification of multiple neurotropic pathogens in CSF samples. These tests enable timely and accurate pathogen identification and initiation of appropriate treatment. Moreover, if a benign viral agent such as enterovirus (EV) is identified, PCR can inform early clinical decision-making and potentially prevent unnecessary antimicrobial use 2-5.
The primary objective of this study was to evaluate the impact of CSF-PCR panel results on clinical decision-making and patient management in children who underwent lumbar puncture (LP) in pediatric emergency and intensive care units with a preliminary diagnosis of meningitis/meningoencephalitis. The secondary objective was to determine the frequency of viral and bacterial agents detected by the CSF-PCR panel and to assess whether clinical or laboratory features could predict the likely pathogen.
MATERIALS and METHODS
This retrospective study included patients aged one month to 18 years who underwent LP for suspected CNS infection in our pediatric emergency or intensive care units between 2018 and 2023, and had a CSF multiplex PCR Meningitis/Encephalitis panel (Bio-Speedy Meningitis/Encephalitis RT-PCR MX-17, Bioeksen, Istanbul, Türkiye) performed. We recorded each patient’s demographic and clinical characteristics, laboratory findings, and treatment modalities.
Patients younger than one month or older than 18 years, those who underwent LP for non-infectious reasons (e.g., suspected intracranial hypertension, evaluation of seizures, metabolic workup, Guillain-Barré syndrome, autoimmune or vasculitic disease, malignancy, subarachnoid hemorrhage), and those with ventriculoperitoneal shunts were excluded.
Patients were classified into two groups based on the CSF-PCR results: viral or bacterial. Patients with mixed pathogens were assigned to the group that best matched their clinical presentation, as determined by the treating physician. The two groups were then compared in terms of clinical and laboratory characteristics. For acute-phase reactants, we defined the cut-off values as procalcitonin (PCT) >0.5 ng/ml and CRP >5 mg/L. For CSF analysis, pleocytosis was defined as >10 leukocytes/mm3 for infants aged 1-3 months and >5 leukocytes/mm3 for children older than 3 months. Elevated CSF protein was defined as >75 mg/dL for infants aged 1-3 months and >45 mg/dL for children older than 3 months. Low CSF glucose was defined as <40 mg/dL (or a CSF/serum glucose ratio <0.6) for infants aged 1-3 months, and <50 mg/dL (or a ratio <0.5) for children older than 3 months6.
Children who have received at least two doses of conjugated Streptococcus pneumoniae vaccine and three doses of Haemophilus influenzae type B vaccine are considered 'fully vaccinated'.
Statistical Analysis
Statistical analyses were performed using SPSS version 24.0 (IBM, Chicago, IL, USA). Categorical variables were compared using the chi-square test. Continuous variables were compared using the Student’s t-test or the Mann-Whitney U-test, depending on the distribution. A p-value <0.05 was considered statistically significant.
Ethics Committee Approval: This study was approved by the Istanbul Medeniyet University Goztepe Suleyman Yalcin City Hospital Clinical Research Ethics Committee (decision no: 2023/0604, date: 20.09.2023).
RESULTS
The median age of 144 patients was 2.7 years, with a standard deviation of 6.7 years, and 64.6% were male. Fourteen patients (9.8%) were incompletely vaccinated or had unknown vaccination status. At least one pathogen was detected by the multiplex PCR panel in 35 patients (24.3%). The demographic and clinical characteristics of patients with a detected pathogen are shown in Table 1. Of these, 22 had viral agents (EV in 9, HSV-1 in 4, Human herpes virus (HHV-8 in 2, HHV-7 in 2, Varicella zoster virus (VZV) in 2, Cytomegalovirus (CMV) in 2, and HHV-6 in 1), 11 had bacterial agents (S. pneumoniae in 7, Neisseria meningitidis in 3, and Hib in 1), and 2 had multiple agents (S. pneumoniae + Hib + HHV-6 in one case; EV + HHV-6 in one case) (Figure 1). Of the two patients with mixed infections, one was categorized in the viral group and the other in the bacterial group based on clinical presentation. The median turnaround time for CSF-PCR results was 24 hours (range 3-118 hours).
Among the patients with an identified pathogen (n=35), fever was present in 91.4%, vomiting in 57.1%, seizures in 34.3%, altered consciousness in 31.4%, signs of meningeal irritation in 28.6%, and headache in 25.7%. There were no statistically significant differences in presenting symptoms between the viral and bacterial cases (Table 1).
Laboratory comparisons revealed that patients in the bacterial group had significantly higher leukocyte counts, CRP, and PCT levels than those in the viral group (p=0.001, p<0.001, and p<0.001, respectively). In the CSF analysis, pleocytosis, elevated protein, and low glucose levels were significantly more frequent in the bacterial group compared to the viral group (p=0.027, 0.001, and 0.011, respectively) (Table 2).
Among 12 patients in the bacterial pathogen group, 9 (75%) were fully vaccinated and 7 (58.3%) had received antibiotics prior to LP. Only 2 (16.7%) of these 12 patients had positive CSF and/or blood culture results (Table 3). None of the PCR-negative patients had a positive culture.
In 51 of 144 patients (35.4%), the CSF-PCR results led to a change in clinical management. The most frequent modifications were discontinuation of acyclovir (22%), discontinuation of antibiotics (8%), and discontinuation of both (3%) (Figure 2). Among the 35 patients with a detected pathogen, 20 (57.1%) had their treatment altered based on the PCR findings. In terms of treatment duration, cases in the viral group had a significantly shorter antibiotic course, with a median of 5 days compared to 10 days in the bacterial group. Meanwhile, cases in the bacterial group had a significantly shorter antiviral (acyclovir) course, with a median of 2 days compared to 14 days in the viral group (p<0.001 for both) (Table 4).
DISCUSSION
In this study of 144 pediatric patients who underwent LP for suspected CNS infection, a multiplex PCR panel identified a viral or bacterial pathogen in approximately one-quarter of cases. Based on the CSF-PCR results, empiric treatments were modified in about one-third of all patients and in nearly half of those in whom a pathogen was detected. These modifications led to shorter antibiotic courses in viral meningitis cases and reduced antiviral (acyclovir) use in bacterial meningitis cases. Our findings thus demonstrate the clinical utility of multiplex CSF-PCR panels in guiding patient management.
Identifying the causative agent in CNS infections is critical for determining appropriate treatment and predicting prognosis. However, clinical features often overlap between viral and bacterial infections, and the sensitivity and specificity of individual acute-phase reactants (e.g., CRP, PCT) are low in identifying specific pathogens. Therefore, it is recommended to interpret these markers alongside clinical scoring systems7, 8. In our study, we found no significant differences in clinical presentation between viral and bacterial cases, but laboratory markers such as CRP, PCT, and leukocyte count were significantly higher in the bacterial group (p<0.001 for all). Similarly, CSF analysis showed that pleocytosis, elevated protein, and decreased glucose were significantly more common in bacterial meningoencephalitis than in viral cases (p=0.027, 0.001, 0.011). While these findings suggest that laboratory parameters may aid in the differential diagnosis, the gold standard remains the identification of the specific pathogen via molecular methods or culture.
Recently developed multiplex PCR panels provide rapid, highly sensitive, and specific identification of causative microorganisms. For example, a multicenter evaluation of the FilmArray ME panel reported 100% sensitivity for most bacterial pathogens and over 99% specificity9. Published positivity rates for CSF multiplex PCR range from 18.8% to 32.8%10-12, aligning with our positivity rate of 24.3%. Differences in detection rates may be due to variability in clinical indications for LP and differences in the PCR panels used.
Viruses are currently the leading cause of meningitis worldwide. In pediatric studies using PCR for the etiologic diagnosis of meningoencephalitis, viral pathogens were detected in 79.7% of cases in the study by Ayhan et al.12, 84.5% in the study by Mizuno et al.13, and 53.7% in the study by Bal et al.14 In our study, viral pathogens accounted for 65.7% of PCR-positive cases, which is consistent with these reports. Globally, EVs are responsible for nearly 85% of viral meningoencephalitis cases15. EVs generally cause a mild clinical course and result in aseptic meningitis3, 4, 16. Although HSV-1 is encountered less frequently, it is the most common cause of sporadic necrotizing encephalitis17.
Findings from different regions show some variations in the prevalent viral agents. In a multicenter pediatric study from Japan, parechovirus (45%) and EV (43%) were the most common causes of viral CNS infections, whereas a study from Türkiye found EV (23.5%), adenovirus (22%), and HHV-6 (22%) to be the most common,13, 18; HSV-1 positivity in the latter study was 5.9%. In our study, EV was detected in 43.5% of viral cases, followed by HSV-1 in 17.4%, a finding which aligns with the literature. Differences between studies may result from variations in seasonality, socioeconomic status, indications for LP, patient age, and geographic location.
Although the prevalence of bacterial meningitis has declined in countries with routine S. pneumoniae and Hib vaccination programs, it remains a major concern due to its high morbidity and mortality19. While CSF culture is the gold standard for diagnosing bacterial meningitis, prior antibiotic use often decreases culture yield. Thus, multiplex PCR testing has become increasingly important. Studies from various countries report bacterial pathogen detection rates via multiplex PCR ranging from 15.5% to 23% 13, 20. In Türkiye, Ayhan et al.12 reported a bacterial detection rate of 20.6%, and Bal et al.14 reported a rate of 36.3%. Our study found a rate of 34.3%. The higher rates in studies from Türkiye compared to other regions might be attributable to differences in vaccination coverage, population characteristics, or study periods.
In patients with suspected meningitis, antibiotic treatment is often initiated before LP is performed, complicating culture-based diagnosis. Additionally, performing LP in children can be technically challenging, and the CSF volume obtained may be insufficient for culture. In a multicenter Turkish study by Ceyhan et al.19, among 645 patients with bacterial pathogens detected by PCR, 74% had received antibiotics before LP, and the culture positivity rate was only 16.2%. Similarly, in our study, of 12 patients with bacterial pathogens detected by PCR, 7 (58.3%) had received antibiotics before LP, and only one had a positive culture. Overall, bacterial growth in CSF and/or blood cultures was identified in just 2 cases (16.7%). These findings highlight that multiplex PCR panels can detect bacterial meningitis pathogens even in patients who have already received antibiotics, allowing for earlier targeted treatment and potentially reducing morbidity and mortality. However, PCR results should always be interpreted in the context of the clinical presentation, as some pathogens may not be included in the panel and false-positive or false-negative results, though rare, can occur.
As a general principle, children with suspected CNS infection should be treated empirically for bacterial meningitis or HSV encephalitis until those diagnoses are excluded. This approach can lead to unnecessary prolonged use of antibiotics or acyclovir, longer hospital stays, and increased healthcare costs, especially in cases that ultimately prove to be benign viral infections4. Previous studies have shown that rapid PCR testing of CSF can reduce unnecessary antimicrobial use and shorten hospital stays1, 4. For example, in pediatric patients with a positive CSF EV PCR result, the duration of IV antibiotic therapy was shortened by a median of 1.5 days, hospital stay decreased from 71.5 to 42 hours, and discharge occurred about 5 hours after the result became available. In contrast, PCR-negative or untested patients often received longer empirical treatment4. In our study, therapeutic modifications were made in 51 patients (35.4%) based on the PCR results. The most common change was discontinuation of acyclovir (22.1% of patients), followed by antibiotics (7.6%) and both (3.5%). These changes significantly shortened the duration of antibiotic therapy in viral cases and the duration of antiviral (acyclovir) therapy in bacterial cases (p<0.001 for both). The widespread use of multiplex PCR panels in emergency settings may help identify benign viral cases early and allow prompt discontinuation of unnecessary treatments, thereby reducing hospital stays and healthcare costs.
The impact of PCR results on clinical decision-making also depends on the test turnaround time. When results are available within 24 hours, studies report that antibiotic use can be reduced by approximately 20%, yielding significant cost savings3. In our study, the median PCR turnaround time was 24 hours (range 3-118 hours), which is relatively long. While this turnaround did not facilitate early discharge from the emergency department, it did allow for shorter durations of unnecessary antimicrobial use. In the future, the use of faster molecular diagnostic methods may enable earlier discontinuation of empirical treatments and safe discharge from the emergency department3, 4.
Study Limitations
This study has some limitations. It was a single-center retrospective study with a relatively small sample size, and we did not perform confirmatory testing (e.g., sequencing or separate PCR assays) for the pathogens detected by the panel. Prospective studies with larger cohorts and broader testing panels are needed to further evaluate the clinical and cost-effectiveness of multiplex PCR testing in CNS infections.
CONCLUSION
CSF multiplex PCR testing is a valuable diagnostic tool for children with suspected CNS infections, as it can rapidly identify causative pathogens and guide early therapeutic decisions. Our findings indicate that PCR results enabled the discontinuation of unnecessary antimicrobial treatments in many cases, thereby shortening treatment durations and potentially reducing hospital stays and healthcare costs. Wider implementation of rapid multiplex PCR panels in clinical practice, along with further large-scale studies, may help improve the management and outcomes of pediatric CNS infections.
