Tick-borne encephalitis (TBE) virus and the spirochete Borrelia burgdorferi are 2 well-known causes of meningoencephalitis, both transmitted via ticks. In children, they are among the most common infections affecting the central nervous system (CNS) in endemic areas.1,2 Although fundamentally different in etiologic agents, TBE and neuroborreliosis (NB, Borrelia infection engaging the CNS) may be similar in their clinical course. This also differs from the clinical picture seen in adults with unspecific symptoms possibly delaying diagnosis.3 Individuals with symptomatic TBE often present with varying degree of meningoencephalitis, ranging from mild meningitis to severe encephalitis or myelitis.1 It is one of the major causes of viral meningoencephalitis in endemic areas with a reported incidence in Swedish children around 3/100,000.4 NB is 10–20 times more prevalent5 and often manifests itself as meningitis, encephalitis or peripheral facial palsy in children.
The definite diagnosis of TBE and NB depends on detection of specific IgM and IgG antibodies in cerebrospinal fluid (CSF, for NB) or serum (TBE). However, there is no rapid analysis of specific antibodies for TBE and NB available in the acute setting, and results are usually received several days after the sample is taken, making it difficult to distinguish between these infections at an early stage. Negative results for Borrelia antibodies in CSF may be seen when analysis is performed early in the disease, with the risk of misdiagnoses or delaying treatment. CSF cell analysis could assist in differential diagnosis as the cellular profile in NB consists mainly of B cells, whereas TBE infections often are dominated by various subsets of T-lymphocytes.6 However, lymphocytic profiles cannot be analyzed by standard CSF methods and are therefore not available in clinical practice. Safely differentiating between TBE and NB in endemic areas in the acute setting is thus often difficult in individuals presenting with signs of CNS infection and a suspected tick bite. The delay in diagnosis that arises from waiting for antibody detection may increase the risk of children having their treatment delayed or receiving unnecessary treatment. Moreover, serology performed early in disease may show false-negative results. Taken together, better tests to distinguish NB from TBE and to detect CNS infection at an early stage, when antibodies or even pleocytosis may be absent, would have the potential to improve medical care.
Several studies have shown that the inflammatory response in the CNS differs depending on the etiology, inducing different chemokine and cytokine profiles. For example, the B-lymphocyte chemoattractant CXCL-13 has been shown to be elevated in NB in both adults and children and may even precede pleocytosis in CSF.7–9 In adult patients with TBE infection, proinflammatory markers such as interleukin 6 (IL-6) are strongly induced and appear also to be prognostic.10 In contrast, the anti-inflammatory cytokine IL-10 is detected at lower levels in CSF of adult TBE patients than in other viral encephalitides.11 These different patterns of cytokine/chemokine expression and the balance between pro- and anti-inflammatory activity may be of importance for outcome but may also be useful in early diagnosis. As cytokine analyses are becoming increasingly available in clinical routine care, this may offer additional information.
The aim of the present study was to compare the levels of cytokines and chemokines in CSF in children infected with NB or TBE and compare them to controls with the aim of finding patterns that may be used clinically for diagnosis.
MATERIALS AND METHODS
Cytokines and chemokines were measured in CSF from children with TBE and from children with NB. Children with TBE during 2004–2010 were retrospectively identified and retrieved from the local Swedish Institute for Infectious Disease Control in Stockholm in a previous study from 2016.12 CSF samples from children with NB during the same years were retrieved from Department of Microbiology at the Karolinska University Hospital in Stockholm. Specific TBE IgM/IgG and Borrelia IgM/IgG antibodies were determined by enzyme immunoassay at the Department of Microbiology at Karolinska University Hospital according to accredited procedures. For the diagnosis of TBE, specific TBE IgM antibodies with or without TBE IgG antibodies in serum in an individual with symptoms of CNS infection and pleocytosis in CSF were required. For the diagnosis of NB, a positive antibody index for CSF/serum IgM was requested. None of the children had a double infection with both TBE virus and Borrelia. As controls, children were acquired from a parallel prospective encephalitis study where they were included as potential encephalitis cases but excluded as they did not fulfill encephalitis criteria. The encephalitis criteria were1 signs of cerebral dysfunction either as (1) encephalopathy defined as altered consciousness, personality or behavioral changes lasting for more than 24 hours, or (2) abnormal electroencephalography findings compatible with encephalitis, plus at least one of the following: abnormal results of neuroimaging compatible with encephalitis, positive focal neurologic findings or seizures. All controls were negative for TBE and Borrelia. The lumbar puncture was performed in the acute phase of the illness before initiation of treatment. All samples were initially stored at −20°C at the microbiologic department and after retrieval to the study center samples stored at −70°C before analyses.
Analysis of Cytokines and Chemokines
A premade multiplex assays (Bio-Plex Pro; Life Science Bio-Rad, Solna, Sweden) was used for detection of IL1β, IL1-Ra, IL-4, IL-6, IL-7, IL-8, IL-10, IL-12 (p70), IL-13, IL15, IL-18, interferon-γ, monocyte chemotactic protein-1, interferon-inducible protein 10 and MIF. The panel was chosen as it contained proinflammatory cytokines that were considered relevant and that were likely to be available in clinical practice. The multiplex assay was run according to the manufacturer’s protocol using Luminex 200 (Luminex Corporation, Austin, TX). CSF samples were assessed as undiluted samples of a volume of 50 µL/ sample as previous preliminary experiments showed generally low concentrations of several of the measured cytokines in CSF. Due to limitations in volume, only single samples could be run. Standards were added to provide calibration curves. The calibration curves for each analyte were calculated using the Bio-Plex software. Values below the detection cutoff level were set at half the value of the lower cutoff, and values above the detection cutoff level were set at double the value of the upper limit.
Ethical approval was obtained from the local ethics committee in Stockholm before the start of the study.
All values are given as mean ± standard deviation or with interquartile ranges (IQRs). After verifying normality by Pearson test, statistical analysis was performed using 1-way analysis of variance. All reported P values are 2 sided and corrected for multiple comparisons within each group using Bonferroni correction. A P value of <0.01 was considered significant. Statistical analyses were performed with Prism Graph Version 8 (GraphPad Software Inc, San Diego, CA).
The studied cohort consisted of 37 children (15 female, 22 male) with TBE, 34 with NB (14 female, 20 male) and 19 controls (8 female, 11 male) with no significant sex differences among groups. Groups differed somewhat in age as children with TBE were significantly older than NB (median age: 138 months; IQR: 58 for TBE and median age: 108 months; IQR: 31 for NB, P < 0.01). Control children were significantly younger (median age: 14 months; IQR: 19.6) than both TBE and NB (P < 0.001).
CSF was obtained during the acute phase of disease and analyzed for cytokines and chemokines. All values are demonstrated in Table 1 and selected values of interest in Figure 1. The levels of the proinflammatory cytokine IL-6 was significantly increased in TBE cases when compared with both controls and NB cases, reaching a 50-fold increase in TBE cases. In contrast, IL-7 was expressed at significantly higher levels in NB when compared with both controls and TBE. For the chemotactic IL-8, there was a significant increase in NB cases when compared with TBE cases, but no significant differences were seen when compared with controls. The anti-inflammatory IL-10 was expressed at significantly higher levels in NB cases when compared with both TBE cases and controls, thus representing an inverted pattern when compared with IL-6. Similarly, levels of IL-13 were also higher in NB patients when compared with TBE cases and controls. There was also a significant IL-13 increase in TBE cases when compared with controls. IL-15 was significantly reduced in TBE cases versus controls, and a similar pattern was seen for monocyte chemotactic protein-1, where both TBE and Borrelia cases had reduced values. interferon-inducible protein-10 was increased in both TBE and Borrelia cases.
The ratio between IL-6 and IL-10 was calculated as an indicator of the proinflammatory and anti-inflammatory balance. This ratio was significantly increased in TBE cases, both compared with NB cases and controls. With a ratio set to 40, the positive predictive value for a TBE infection is 97% and the corresponding negative predictive value is 93%.
There were no significant differences measured for IL1-ra, IL-1β, IL-4, IL-12 p70, interferon-γ, IL-18 and MIF.
In the present study, we demonstrate that the intrathecal cytokine/chemokine profile in the acute phase may be used to identify children with active TBE or NB infection. Also, the studied infections have different inflammatory patterns that may serve as a biomarker to clinically distinguish between the 2 of them. The finding that TBE and NB activate different neuroinflammatory pathways upon infection underlines the different pathogeneses and prognosis and offers insight into possible differences in pathophysiologic mechanisms.
TBE is a viral disease, and no treatment other than supportive care is available. In contrast, NB is treated with antibiotics, intravenously in young children. The different treatment approaches warrant a rapid and correct diagnosis. After diagnostic lumbar puncture in a child with a suspected CNS infection, antibiotic treatment is often initiated if pleocytosis is present and discontinued several days later if antibody testing and cultures are negative. This approach, however, has several caveats. The development of antibodies in CSF and serum takes time, and children with CNS infections often present early in the disease course. It is well known that false-negative results on antibody testing are seen when serologic testing is performed early in NB.13 Another problem with early analysis of the CSF in CNS infections is that pleocytosis may not yet have developed.1,14 This may result in misinterpretation and a risk of missing children who have NB and who should be treated with antibiotics.
An ideal diagnostic testing would have a high sensitivity both for detecting disease in itself and also enable early distinction between NB and viral etiology. Furthermore, to avoid delayed treatment, identification of children with NB but with no CSF pleocytosis or detectable intrathecal antibodies at the time of testing is important. The chemokine and B-cell attractant CXCL-13 has been suggested as such a biomarker for NB with expression preceding the development of pleocytosis and the intrathecal antibody production.15 A study of 48 different cytokines in adults displayed elevated concentrations of all major T-helper cell-type cytokines in the CSF.16 In the present study, we found that the proinflammatory IL-6 was significantly higher in children with TBE compared with NB and controls. However, IL-6 expression appears unspecific as it is increased in the early stages of various kinds of neuroinflammation of infectious origin.17 Also, the lack of IL-6 in the presence of pleocytosis is unspecific, as it is found in noninfectious encephalitis, for example, N-Methyl-d-Aspartate Receptor encephalitis.18 The expression of IL-6 thus indicates an active proinflammatory state irrespective of etiology, but different etiologies cause different inflammatory patterns. This may be of importance for outcome, as a balanced anti-inflammatory response counteracting the pro-inflammatory cascade is believed to be important to avoid permanent neuronal damage.
The finding of high levels of IL-10 in NB is potentially important. These data are in line with previous publications showing that Borrelia spirochetes have the capacity to induce IL-10, as seen in CSF of adult patients with NB16 and Borrelia meningoradiculitis.19 The source of this IL-10 is most likely CNS monocytes. It is known that both alive and dead Borrelia spirochetes induce IL-10 in in vitro macrophage cultures and that the subsequent macrophage gene expression to a great extent can be mimicked by adding exogenous IL-10.20 IL-10 is considered an anti-inflammatory cytokine and affects the induction and repression of many genes in the inflammatory cascade.21 For example, transcription of the key proinflammatory cytokine IL-6 is repressed.21 It thus appears that the spirochete itself has the inherent capacity to limit inflammation. Another possibility is that lumbar puncture in NB patients was performed later in the infection than in TBE patients, and that the observations merely reflect a shift from early proinflammatory mediators to later anti-inflammatory cytokines. The in vitro data, however, support the first suggestion.
Furthermore, the different clinical course and prognosis in NB and TBE, respectively, indicate differences in activated pathophysiologic mechanisms. The prognosis of NB is considered favorable in nearly all patients regardless of prompt or delayed treatment, as well as in patient without treatment.22,23 In contrast, patients with TBE infection often suffer from long-lasting or permanent sequelae.24,25 Thus, it may be beneficial and protective with high levels of anti-inflammatory cytokines, such as IL-10 seen in NB. In comparison, children with TBE had low levels of IL-10 in our study, and similar results have been shown in adult patients with TBE.11 Of importance, IL-6 and IL-10 are already often available in clinical routine diagnostics. The balance between pro- and anti-inflammatory activity and the timing of these processes are likely to be of importance for the disease process and outcome following infections. Although complex in nature and difficult to characterize, this offers potential therapeutic targets for intervention.
The current study has a potentially important limitation in that the age differed significantly between groups. In particular, control children were younger than TBE and NB children. This may influence results if age is an important factor for baseline or reactive cytokine levels. However, data from our group on a prospective cohort of children with encephalitis do not indicate either factor as of essential importance (Unpublished data, Sofia Ygberg et al, 18th november 2019). We thus believe that the observed differences in cytokine expression are reliable, but they need to be verified in further studies. Furthermore, the studied cohorts are of a limited sample size and results need to be corroborated in further studies. In addition, the cytokine expression of an individual in response to Borrelia infection appears to be influenced by both the immune system of the infected individual and the Borrelia genospecies and strain causing the disease.26 Comparison between studies and cohorts may therefore be complicated for Borrelia, whereas no similar differences have been demonstrated for TBE.
In conclusion, the present study demonstrates that the intrathecal cytokine/chemokine profile may be used as a biomarker in children to identify CNS infections and differentiate etiologies. In particular, we suggest that the CSF IL-6/IL-10 ratio could be a valuable tool in differentiating tick-borne infections early in the acute phase.
1. Fowler A, Stödberg T, Eriksson M, et al. Childhood encephalitis in Sweden: etiology, clinical presentation and outcome. Eur J Paediatr Neurol. 2008;12:484–490.
2. Tveitnes D, Natås OB, Skadberg Ø, et al. Lyme meningitis, the major cause of childhood meningitis in an endemic area: a population based study. Arch Dis Child. 2012;97:215–220.
3. Hansson ME, Orvell C, Engman ML, et al. Tick-borne encephalitis in childhood: rare or missed? Pediatr Infect Dis J. 2011;30:355–357.
5. Södermark L, Sigurdsson V, Näs W, et al. Neuroborreliosis in Swedish children: a Population-based Study on incidence and clinical characteristics. Pediatr Infect Dis J. 2017;36:1052–1056.
6. Blom K, Cuapio A, Sandberg JT, et al. Cell-mediated immune responses and immunopathogenesis of human tick-borne encephalitis virus-infection
. Front Immunol. 2018;9:2174.
7. Barstad B, Tveitnes D, Noraas S, et al. Cerebrospinal fluid
B-lymphocyte chemoattractant CXCL13 in the diagnosis of acute lyme neuroborreliosis in children. Pediatr Infect Dis J. 2017;36:e286–e292.
8. Sillanpää H, Skogman BH, Sarvas H, et al. Cerebrospinal fluid
chemokine CXCL13 in the diagnosis of neuroborreliosis in children. Scand J Infect Dis. 2013;45:526–530.
9. Tjernberg I, Henningsson AJ, Eliasson I, et al. Diagnostic performance of cerebrospinal fluid
chemokine CXCL13 and antibodies to the C6-peptide in lyme neuroborreliosis. J Infect. 2011;62:149–158.
10. Kang X, Li Y, Wei J, et al. Elevation of matrix metalloproteinase-9 level in cerebrospinal fluid
of tick-borne encephalitis patients is associated with IgG extravassation and disease severity. PLoS One. 2013;8:e77427.
11. Günther G, Haglund M, Lindquist L, et al. Tick-borne encephalitis is associated with low levels of interleukin-10
in cerebrospinal fluid
. Infect Ecol Epidemiol. 2011;1.
12. Fowler Å, Ygberg S, Bogdanovic G, et al. Biomarkers in cerebrospinal fluid
of children with tick-borne encephalitis: association with long-term outcome. Pediatr Infect Dis J. 2016;35:961–966.
13. Ljøstad U, Skarpaas T, Mygland A. Clinical usefulness of intrathecal antibody testing in acute Lyme neuroborreliosis. Eur J Neurol. 2007;14:873–876.
14. Erdema H, Ozturk-Enginb D, Cagc Y, et al. Central nervous system
infections in the absence of cerebrospinal fluid
pleocytosis. Int J Infect Dis. 2017:65:107–109.
15. Henningsson AJ, Lager M, Brännström R, et al. The chemokine CXCL13 in cerebrospinal fluid
in children with lyme neuroborreliosis. Eur J Clin Microbiol Infect Dis. 2018;37:1983–1991.
16. Pietikäinen A, Maksimow M, Kauko T, et al. Cerebrospinal fluid cytokines
in lyme neuroborreliosis. J Neuroinflammation. 2016;13:273.
17. Dano ID, Sadou H, Issaka B, et al. Measurement of interleukin-6
in cerebrospinal fluid
for the diagnosis of bacterial meningitis. Pak J Biol Sci. 2016;19:185–190.
18. Ygberg S, Fowler Å, Wickström R. Cytokine and chemokine expression in CSF may differentiate viral and autoimmune NMDAR encephalitis in children. J Child Neurol. 2016;31:1450–1456.
19. Cepok S, Zhou D, Vogel F, et al. The immune response at onset and during recovery from Borrelia burgdorferi
meningoradiculitis. Arch Neurol. 2003;60:849–855.
20. Gautam A, Dixit S, Philipp MT, et al. Interleukin-10
alters effector functions of multiple genes induced by Borrelia burgdorferi
in macrophages to regulate lyme disease inflammation. Infect Immun. 2011;79:4876–4892.
21. Murray PJ. The primary mechanism of the IL-10-regulated antiinflammatory response is to selectively inhibit transcription. Proc Natl Acad Sci U S A. 2005;102:8686–8691.
22. Skogman BH, Glimåker K, Nordwall M, et al. Long-term clinical outcome after lyme neuroborreliosis in childhood. Pediatrics. 2012;130:262–269.
23. Krüger H, Kohlhepp W, König S. Follow-up of antibiotically treated and untreated neuroborreliosis. Acta Neurol Scand. 1990;82:59–67.
24. Fowler Å, Forsman L, Eriksson M, et al. Tick-borne encephalitis carries a high risk of incomplete recovery in children. J Pediatr. 2013;163:555–560.
25. Haglund M, Günther G. Tick-borne encephalitis--pathogenesis, clinical course and long-term follow-up. Vaccine. 2003;21(suppl 1):11–18.
26. Cerar T, Strle F, Stupica D, et al. Differences in genotype, clinical features, and inflammatory potential of Borrelia burgdorferi
sensu stricto strains from Europe and the United States. Emerg Infect Dis. 2016;22:818–827.