Intrathecal expression of IL-5 and humoral response in patients with tick-borne encephalitis

Ticks and Tick-borne Diseases xxx (xxxx) xxx–xxx

Contents lists available at ScienceDirect

Ticks and Tick-borne Diseases journal homepage: www.elsevier.com/locate/ttbdis

Intrathecal expression of IL-5 and humoral response in patients with tickborne encephalitis Sambor Grygorczuk, Piotr Czupryna⁎, Sławomir Pancewicz, Renata Świerzbińska, Maciej Kondrusik, Justyna Dunaj, Joanna Zajkowska, Anna Moniuszko-Malinowska Department of the Infectious Diseases and Neuroinfections, Medical University in Białystok, Poland

A R T I C LE I N FO

A B S T R A C T

Keywords: Tick-borne encephalitis Pleocytosis Serologic response B-lymphocytes IL-5

Aim: The aim of the study was to assess the role of an early specific humoral response in human infection with a tick-borne encephalitis virus (TBEV) and the role of IL-5 as its potential mediator and marker. Materials and methods: The retrospective study involved a cohort of 199 patients diagnosed with TBE, in whom anti-TBEV IgM and IgG antibody titers were analyzed on admission and compared with clinical presentation and basic laboratory parameters. The prospective study included 50 TBE patients in whom IL-5 serum and CSF concentration was measured with ELISA on admission in the TBE neurologic phase and in selected patients before discharge, at follow-up or in samples obtained before the neurologic phase onset. Results: The serum anti-TBEV IgM correlated with good clinical outcome and the CSF anti-TBEV IgM with more pronounced CSF inflammation on admission, but also with its more complete resolution on follow-up. The serum anti-TBEV IgG correlated with milder presentation and better outcome. Concentration of IL-5 was increased in CSF but not in the serum of TBE patients. IL-5 concentration index on admission favored its intrathecal synthesis. IL-5 did not correlate significantly with clinical presentation and specific IgM and IgG titers. Conclusions: Specific anti-TBEV IgM systemic and intrathecal response and IgG systemic response are protective, together favoring milder presentation, better outcome and resolution of central nervous system (CNS) inflammation. IL-5 is expressed intrathecally in TBE, but its pathogenetic role remains unclear.

1. Introduction Tick-borne encephalitis (TBE) is an acute infectious disease involving central nervous system (CNS), caused by a zoonotic flavivirus (tick-borne encephalitis virus, TBEV) which is transmitted by Ixodes ticks. It is endemic in the temperate zone of Asia, Eastern and Central Europe, where several thousand cases are reported annually (Mansfield et al., 2009; Randolph, 2001). Infection with the European sub-type of TBEV may be asymptomatic or result in a mild, flu-like disease (Gustafson et al., 1992). If the virus penetrates into CNS, the disease may take form of uncomplicated meningitis or neurologic involvement (meningoencephalitis or meningoencephalomyelitis) of variable severity, from relatively mild to life-threatening and/or causing permanent neurologic deficits (Czupryna et al., 2011; Kaiser, 2002; Mickiené et al., 2002; Schellinger et al., 2000). The pathogenesis of nervous tissue damage is complex and seems to result from a combination of the TBEV direct effect on neurons and local inflammatory response. The neural cells are highly susceptible to TBEV infection and when infected die either by apoptosis or necrosis (Růžek et al., 2009) while TBEV-

⁎

infected astrocytes become a potent source of pro-inflammatory mediators, with possible pathologic consequences to the nervous tissue (Palus et al., 2014). Animal models and clinical studies involving TBE patients show a prevalent intrathecal cellular response including Th1 CD4+ lymphocytes and cytotoxic CD8+ lymphocytes, likely involved in the disease immunopathogenesis and neuronal damage (Gelpi et al., 2006; Holub et al., 2002; Jeren and Vince, 1998; Růžek et al., 2009). Although B lymphocytes are scarce in CSF compared to T cells (Jeren and Vince, 1998), data suggest that a peripheral and intrathecal humoral response to TBEV may play an important protective role. Serologic IgM response is early and highly specific in different flavivirus infections, for example in dengue its onset coincides with the defervescence and the end of viremia (Innis et al., 1989). Analogously, IgM and IgG antibodies are usually detectable in serum at the onset of the neurologic phase of TBE, while in CSF specific IgM can be detected in about 50% of patients early in the neurologic phase and in practically all 10 days after the first symptoms of CNS involvement (Holzmann, 2003). In a study by Günther et al. (1997) specific IgM was detectable in serum at the beginning of the neurologic phase and peaked in CSF in

Corresponding author at: Żurawia 14, 15-540, Białystok, Poland. E-mail address: [email protected] (P. Czupryna).

https://doi.org/10.1016/j.ttbdis.2018.03.012 Received 19 November 2017; Received in revised form 10 March 2018; Accepted 11 March 2018 1877-959X/ © 2018 Elsevier GmbH. All rights reserved.

Please cite this article as: Grygorczuk, S., Ticks and Tick-borne Diseases (2018), https://doi.org/10.1016/j.ttbdis.2018.03.012

Ticks and Tick-borne Diseases xxx (xxxx) xxx–xxx

S. Grygorczuk et al.

the second week after the onset with the intrathecal production documented in >90% of patients. An elevated IgG index pointing to a high intrathecal immunoglobulin synthesis often persists into the convalescent period of the disease (Günther et al., 1997). In CNS infections with other flaviviruses the serologic response may be indispensable for the control of the infection, as shown in a mouse model of West Nile virus (WNV) infection and in patients with Japanese encephalitis (JE) (Diamond et al., 2003a,b; Libraty et al., 2002). The studies on TBE patients suggest the protective role of the early serologic response in periphery (Atrasheuskaya et al., 2003; Günther et al., 1996; Günther et al., 1997; Kaiser and Holzmann, 2000; Toporkova et al., 2008), while the significance of the intrathecal synthesis of the specific antibodies is less clear, as their high titer and prolonged presence tend to associate with more serious neurologic involvement (Günther et al., 1997; Kaiser, 2002). It may be suspected that the delayed and/or originally ineffective intrathecal serologic response contributes to more severe neurologic phase, with worse control of the infection and predominant immunopathogenic cellular response, but more data on the association of the serologic response with the clinical presentation and outcome are needed to confirm that hypothesis. The role of intrathecal B lymphocytes presence and activation status has not been directly studied in TBE so far. These cells constitute only up to a few percent of all CSF lymphoid cells and are scarce in the CNS infiltrates in flavivirus encephalitis, but their association with the clinical manifestation and outcome is not known (Grygorczuk et al., 2016; Holub et al., 2002; Jeren and Vince, 1998). CXCL13, chemokine acting predominantly on B lymphocytes, seems to play an important role in the pathogenesis of acute autoimmune and enteroviral encephalitis (Kothur et al., 2016) and is characteristically up-regulated in CSF of patients with neuroborreliosis. However, although detected as increased in serum and CSF of patients with TBE, the increase of its concentration is relatively modest and does not form a clear chemotactic gradient toward CSF, making its role as a marker of B cell-dependent response dubious in this setting (Cerar et al., 2013; Pietikäinen et al., 2016; Zajkowska et al., 2011). Here we have attempted to study another cytokine involved in B cell-dependent immunity, IL-5. In humans IL-5 plays an important role in the immune response, inflammation and disease control, including, although not limited to, specific humoral responses (Takatsu, 2011). IL-5 is a factor that induces terminal differentiation of B cells to Ig-secreting cells. Stimulation of Il-5 induces rapid tyrosine phosphorylation of cellular proteins including the βc, SH2/SH3-containing proteins such as Vav and Shc, Btk and Btkassociated molecules, JAK1/JAK2 and STAT1/STAT5, PI3K and MAP kinases that activate downstream signaling molecules (Horikawa et al., 2001; Huang et al., 2006; Kagami et al., 2000; Kouro et al., 1996; Martinez-Moczygemba et al., 2007; Meads et al., 2010; Ogata et al., 1998; Satoh et al., 1995; Takaki et al., 1994; Takatsu, 2011; Zahn et al., 2000). Il-5 induces CD38-activated splenic B cells to differentiate into immunoglobulin M-secreting cells and undergo μ to γ1 class switch recombination (CSR) at the DNA level, resulting in immunoglobulin G1 (IgG1) production (Horikawa et al., 2001). Its intrathecal expression in CNS infections and role in their pathogenesis remains largely unknown. The aim of our study was to assess the role of IL-5 expression and early serologic response in the pathogenesis of TBE. Therefore, we retrospectively studied the relation between the serologic response and selected clinical and laboratory parameters in a large group of patients with TBE. Following that, expression of IL-5 was evaluated prospectively in a smaller group of selected patients.

Fig 1. The specific anti-TBEV IgM score in serum (S, empty bars) and cerebrospinal fluid (CSF, light grey bars) on admission to hospital in patients with TBE with typical biphasic (+); (n = 85) and monophasic (−); (n = 93) clinical presentation. Shown are the median (horizontal line), quartiles (box) and maximum values (whiskers). * − the significant difference with p < 0.05.

University of Białystok, Poland, from September 2010 to December 2014 was screened for patients with the established diagnosis of TBE with the revision of their clinical and laboratory data. We identified 243 patients, of whom 20 were excluded because of the uncertain diagnosis (serologic testing towards TBE not performed or all the available results negative), leaving 223 patients with a documented specific anti-TBEV IgM response in at least one serum or CSF sample (on admission or seroconversion in follow-up samples). Further 24 patients in whom the initial serologic examination was not performed and the diagnosis was based solely on the follow-up serology were excluded. The final study group consisted of 199 patients. 2.1.2. Procedures The time from symptoms onset, the course of the disease at home, as well as the clinical presentation and severity on admission and complications during hospitalization, the outcome on discharge and during follow-up visits, and the results of laboratory examinations were extracted from patients' records. The disease severity was scored from 0 to 6 using an arbitrary pre-defined scale designed for the purpose of the study, in which 0 corresponded to a flu-like infection, 1 to an uncomplicated meningitis, 2 − mild CNS involvement with no altered mental status nor paresis (limited, transient neurologic symptoms, e.g. Babinski sign, mild ataxia, paresthesia, tremor), 3 − relatively mild encephalitis with lethargy, drowsiness, altered affect, mild monofocal paresis, gait disorders, but patients in logical contact and able to walk; 4 − moderately severe encephalitis with disorientation, multifocal and/ or severe focal neurologic symptoms, generalized seizures; 5 − loss of consciousness; 6 − coma or death. Patients were also stratified according to the course of the disease (typical biphasic with a distinct initial flu-like peripheral phase or monophasic), presentation in the neurologic phase (meningitis, meningoencephalitis, meningoencephalomyelitis), abnormalities in mental status (defined as mild − slowliness, lethargy; moderate − disorientation, agitation, psychotic symptoms; severe − lack of any logical contact or loss of consciousness), presence of paresis, as well as outcome on discharge from hospital or during planned follow-up visits within weeks after discharge (defined

2. Materials and methods 2.1. Retrospective study 2.1.1. Patients Clinical registry of diagnoses of patients hospitalized in the Department of Infectious Diseases and Neuroinfections of the Medical 2

Ticks and Tick-borne Diseases xxx (xxxx) xxx–xxx

S. Grygorczuk et al.

Fig. 2. The association of the specific anti-TBEV IgM in serum and CSF of TBE patients with age. A. The negative correlation of the serum anti-TBEV IgM score on admission (horizontal axis) with the age (vertical axis). Data from individual patients are show with points, the linear fit with a continuous line and 95% confidence interval with dashed lines. The strength and statistical significance of the correlation are presented on the plot. B. The age of patients with the specific anti-TBEV IgM detectable (+); (n = 77) and undetectable (−); (n = 103) in CSF on admission. Shown are median (vertical line), quartiles (box) and minimum/maximum values (whiskers). The level of statistical significance of the tendency for a higher age in (−) is presented on the plot.

2.2. Prospective study

as complete recovery, persistent subjective symptoms or persistent objective neurologic/psychiatric sequelae). The CSF data came from diagnostic lumbar punctures performed on admission (although in individual patients, mostly with mild or atypical presentation, delayed by up to 6th day after admission) and control lumbar punctures performed in an early convalescent period in most of the patients, typically directly before discharge 12–16 days after admission (up to 21st day). The results of peripheral blood laboratory examinations simultaneous with lumbar punctures were selected for analysis.

2.2.1. Patients The consecutive patients hospitalized in the Department of the Infectious Diseases and Neuroinfections of the Medical University of Bialystok in 2013–2015 with a suspicion of TBE were invited to participate in the study. The final inclusion into a TBE group required a history of a recent tick bite or exposure to ticks in TBE endemic areas, clinical symptoms consistent with a neurologic phase of TBE (acute febrile illness with a prominent headache and meningeal signs), documented CSF pleocytosis of at least 15 cells/μl and the detection of the specific anti-TBEV IgM antibodies in at least one serum or CSF sample (on admission or follow-up). Patients with a probable alternative diagnosis, clinically important co-infection or a known significant immune deficiency (HIV infection, immunosuppressive treatment, active malignancy) and no willing to participate were excluded. The study group consisted of 50 patients with CNS infection caused by TBEV (19 women, 31 men, aged from 18 to 70 years, mean 44 years) Seven patients with aseptic meningitis of non-TBE etiology were studied as a meningitis control group; 6 of them were hospitalized during Echovirus 30 meningitis outbreak and had consistent, mild clinical presentation. The blood control group consisted of 8 healthy blood donors. Non-inflammatory control CSF samples (n = 8) were obtained from patients who were hospitalized with the suspicion of meningitis but in whom neuroinfection was eventually excluded. The subjects gave their informed consent for the participation and the study was accepted by the Bioethics Committee of the Medical University of Bialystok (Ref. no of the approval decisions: R-I-002/225/2015).

2.1.3. Laboratory examinations Peripheral leukocytosis and differential, C-reactive protein concentration and serum albumin concentration as well as basic CSF parameters (total pleocytosis, neutrophil and lymphocyte count, total protein and albumin concentration) were routinely studied in hospital laboratory with standard laboratory techniques on Cobas Integra 400 analyzer. Anti-TBEV IgM and IgG antibodies were routinely measured in admission serum and CSF of patients hospitalized with a suspicion of meningitis and re-evaluated in the convalescent samples in patients with initial negative results but pending clinical suspicion of TBE. The commercial diagnostic Enzygnost Anti-TBE/FSME IgM and Anti-TBE/ FSME IgG kit from Siemens (Munich, Germany) was used throughout the study period following the standard procedure. 2.1.4. Data analysis The serologic response against TBEV was analyzed and correlated with clinical (manifestation: meningitis, encephalitis, meningoencephalitis; severity, presence of altered mental status or paresis) and basic laboratory parameters. CSF albumin quotient (Qalb) was calculated as albumin CSF/albumin serum in 32 patients, in whom the serum albumin concentration was available. The data were analyzed with Statistica 12 software with the use of non-parametric tests and p < 0.05 was considered statistically significant. The distribution of values of clinical scores assessing presentation and severity in groups of patients with and without specific IgM detected in the admission CSF was assessed with chi-square test.

2.2.2. Procedures The patients underwent from 2 to 4 lumbar punctures during the diagnostic process and follow-up. For the needs of the study the blood and CSF samples were obtained simultaneously, together with the material drawn for clinically indicated laboratory examinations at the time of lumbar punctures. The first sample (examination I) obtained on admission was studied in all 50 patients, the second after 10–16 days in 13 patients to assess the IL-5 concentration dynamics. The third pair of samples was obtained during later follow-up examinations in the 3

Ticks and Tick-borne Diseases xxx (xxxx) xxx–xxx

S. Grygorczuk et al.

Fig. 3. The score of the specific anti-TBEV IgM in serum (S, empty bars) and cerebrospinal fluid (CSF, light grey bars) on admission to hospital in patients with different clinical presentation and outcome of TBEV infection. Shown are the median (horizontal line), quartiles (box) and maximum values (whiskers). A. Patients with different clinical forms of TBEV infection: meningitis (M, n = 84), meningoencephalitis (ME, n = 95) and meningoencephalomyelitis (MEM, n = 17); * − the trend for higher score in MEM significant with p < 0.05. B. Patients presenting with (+); (n = 22) and without (−); (n = 177) paresis. * − the significant difference with p < 0.05; C. Patients with full recovery on discharge from hospital or during the first control visit (1, n = 122), with protracted subjective complaints (2, n = 36) and with residual neurologic or psychiatric deficits (3, n = 15). * − significantly higher score in (1) when compared with (2) and (3) combined, p < 0.05.

assess the intrathecal synthesis as IIL-5 = (IL-5 CSF/IL-5 serum)/Qalb. The data were analyzed with Statistica 12 software with non-parametric tests and p < 0.05 considered significant.

convalescent period (5–8 weeks after initial hospitalization) in 5 patients who underwent additional follow-up examinations because of the slow or no normalization of the CSF inflammatory parameters in examination II time point.

3. Results 2.2.3. Laboratory examinations Basic laboratory and serologic examinations were performed with same methods and procedures as described in the “Retrospective study” section. The venous blood for IL-5 detection was obtained into EDTA-coated tubes and stored in 5 °C before being centrifuged and serum collected. All the serum and CSF samples were frozen within 24 h after collection, stored in −80 °C and thawed simultaneously directly before study. Concentration of IL-5 was measured with commercial ELISA kit from SunRed (Shanghai, China). The procedure was performed strictly following the manufacturer’s instructions. The assay sensitivity according to the manufacturer’s data is 0.158 pg/ml.

3.1. The retrospective study 3.1.1. Patient clinical data The patients were studied after a median time of 10 days since the first symptoms of infection and 4 days after the onset of fever. No patient reported any previous anti-TBEV vaccination. There were 3 patients with a flu-like infection, 84 with meningitis (M), 95 with meningoencephalitis (ME), and 17 with meningoencephalomyelitis (MEM) of variable severity. Of patients with neuroinvasive disease (M, ME, MEM) 85 reported a biphasic and 93 a clinically monophasic disease. Of patients diagnosed with ME/MEM, the disease severity was scored as 2 in 24, 3 in 50, 4 in 35, only one patient was scored 5 and two − 6 (there were no fatal cases). In the statistical analysis, patients scoring 4–6 were pooled in a single group. The mental status was not altered in 140 patients, while 21 presented with changes assessed as mild, 32 − moderate and 6 as severe (4 were conscious, but without verbal contact

2.2.4. Data analysis CSF albumin quotient (Qalb) was calculated as described above for the retrospective cohort in 34 patients in whom the sufficient data were available. The index of IL-5 concentration in CSF was calculated to 4

Ticks and Tick-borne Diseases xxx (xxxx) xxx–xxx

S. Grygorczuk et al.

Table 1 Tick-borne encephalitis patients included in the prospective study. No

Age

Sex

Clini-cal form

CSF parameters during the neurologic phase (examination I)

Anti-TBEV IgM antibody score on admission

Anti-TBEV IgG antibody score on admission

pleocytosis (cells/μl)

protein (mg/dl)

albumin (mg/dl)

Qalb x 100

serum

CSF

serum

CSF

1 2

69 27

F M

M M

23 235

41.1 67.7

33.13 51.2

7.05 12.49

4.70 8.99

0 0

21.8 26.1

0 0

3

61

M

M

48

54.6

38.5

NA

8.77

0

27.4

0

4

53

F

M

120

59

58.3

13.56

9.37

0

0

0

5

43

F

ME

144

47

33

8.25

7.83

0

0

0

6 7 8 9

68 24 36 70

F M F F

ME M ME ME

241 142 236 171

47.3 116.4 130 101.7

36.5 87.36 99.9 85.7

9.36 17.83 24.37 22.55

3.77 9.03 8.62 NA

0 3.09 2.62 NA

31.3 18.1 0 NA

0 18.7 0 NA

10

20

M

M

71

25.3

16.5

3.84

6.63

0

21.6

0

11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53

41 42 27 26 30 27 51 35 50 41 38 23 53 32 65 64 58 20 52 18 43 65 50 45 31 53 50 44 65 43 62 66 55 28 28 30 55 70 44 24 18 45 57

F M F M M M M F M M F F F M M F M F M M F M M F M M M M M M F M M M M M F F M M F F M

M M M M M ME M M ME ME M M M ME ME M ME ME M ME ME M ME ME MEM ME ME M ME ME ME ME MEM ME ME M ME M M M M ME MEM

103 44 424 260 194 16 126 42 115 30 386 73 40 58 58 49 63 215 118 103 56 46 146 1196 85 33 32 51 94 352 146 8 67 258 176 122 64 30 141 76 266 80 51

65 31 86.7 65 77.3 46 37.9 26.8 124.4 66.3 90.2 53.9 39.2 42.7 73.1 52.7 70.2 92.8 66.5 58.6 45.2 84.7 53.1 129.9 49.8 57.4 109.2 42.5 56.3 95.7 67.4 37.7 103 100.6 48.2 64 65.9 42.2 60.3 61.4 50.4 45.7 91.5

47.8 20.2 56.8 49.3 58.2 NA NA NA 97.3 NA 74.6 45.2 31.3 34.8 49.7 31 52.5 57.39 47.9 39.3 29.32 60.03 35.7 94.3 35.1 40.9 86.5 29.8 40.1 28.6 44.9 24 85.6 72.9 33.6 48.6 48.9 28.6 43.7 47.5 38 36.1 68.1

11.2 5.61 12.91 10.96 11.64 NA NA NA NA NA NA 9,42 NA NA NA NA NA NA NA 9.14 7.52 16.68 8.71 NA NA NA 20.60 6.34 9.11 6.65 NA 5.45 19.91 17.78 7.30 11.30 13.97 8.17 10.93 10.80 10.27 8.80 17,03

NA 8.15 9.85 8.94 7.53 7.28 5.64 6.73 9.02 NA 5.84 9.56 10.24 9.92 8.67 10.55 3.97 11.96 10.60 12.97 10.64 6.22 9.98 3.14 10.39 4.53 4.37 9.86 8.06 9.56 6.41 9.78 4.61 9.61 5.70 6.18 8.08 10.48 8.90 7.40 7.76 9.55 6.17

NA 0 3.16 1.40 NA 3.13 0 0 5.34 NA 0 0 0 0 0 1.13 1.42 7.74 0 1.82 0 0 2.24 0 2.54 0 0 0 0 1.40 0 0 0 0 0 0 0 1.18 NA 0 0 NA 3.01

NA 23.0 12.1 0 36.2 37.5 0 0 11.1 NA 86 10.9 33 49.7 33.4 12.5 0 0 16.4 33.9 33.8 546.5 33.5 0 39.7 21.8 17 48,7 31.8 12.6 17 86.1 10.7 17.5 0 38.4 22.8 43.1 10.2 20.2 33.2 17 9.7

NA 0 0 0 NA 0 0 0 0 NA 10.6 0 0 0 0 0 0 0 0 0 0 23.4 0 0 0 0 9.1 0 12.7 0 0 0 0 0 0 0 0 0 NA 0 0 NA 10

M – meningitis. ME – meningoencephalitis. MEM – meningoencephalomyelitis. NA – not available. Qalb – CSF/serum albumin quotient.

5

Remarks

additional peripheral phase sample additional peripheral phase. examination II and III samples additional (pre-examination I) early neurologic phase sample additional (pre-examination I) early neurologic phase and examination II samples additional examination II and III samples additional examination II sample additional examination II sample additional examination II sample additional examination II and III samples additional examination II and III samples additional examination II sample additional examination II sample additional examination II sample additional examination II sample additional examination II sample

only early neurologic phase sample

only early neurologic phase sample

only early neurologic phase sample

Ticks and Tick-borne Diseases xxx (xxxx) xxx–xxx

S. Grygorczuk et al.

disease severity or mental status. However, patients who eventually fully recovered had higher admission serum anti-TBEV IgM score than those who developed residual signs or symptoms (p < 0.05) (Fig. 3). The disease presenting as MEM was associated with the presence of the anti-TBEV IgM in CSF (p < 0.05, Table 2). The CSF anti-TBEV IgM median score was higher in patients with MEM than in M or ME and in patients presenting with paresis than in the rest of the cohort (p < 0.05) (Fig. 3). Other clinical parameters (presence of M versus ME, severity and mental status scores) did not associate with either presence or score of anti-TBEV IgM in CSF. The serum anti-TBEV IgM score correlated with peripheral leukocytosis (R = 0.22, p < 0.01) and neutrophil count (R = 0.22, p < 0.01), but not with the lymphocyte count. It did not correlate with any of the CSF inflammatory parameters measured neither on admission nor during follow-up except for the highly significant negative correlation with Qalb (R = −0.62, p < 0.001) (Fig. 4). Patients positive for anti-TBEV IgM in CSF tended to have higher neutrophilia (p = 0.055) and in the CSF IgM-positive group the specific IgM score correlated further with the peripheral neutrophil count (R = 0.25, p < 0.05). The CSF inflammatory parameters were elevated in the CSF IgM-positive patients compared to the CSF IgM-negative ones: the difference was highly significant for pleocytosis (p < 0.01), lymphocyte count (p < 0.01), albumin (p < 0.01) and total protein concentration (p < 0.0001). In CSF IgM-positive patients the antibody score correlated with CSF total protein (R = 0.32, p < 0.01) and albumin (R = 0.37, p < 0.01) concentration and tended to correlate with CSF lymphocyte count. The convalescent CSF albumin and protein concentrations did not depend on admission serology, while for the cellular parameters the original trend reversed, with median values lower in patients with CSF anti-TBEV IgM present on admission (p < 0.05 for pleocytosis, p = 0.01 for lymphocyte count) (Fig. 5). The correlation with CSF albumin could indicate the specific IgM infiltrating CSF from the periphery through the damaged blood-brain barrier (BBB). However, Qalb did not differ between the groups of CSF IgM-negative and CSF IgM-positive patients and did not show any trend for a correlation with CSF IgM score in the later, suggesting that the anti-TBEV IgM score in CSF reflected primarily an intrathecal synthesis (Fig. 5).

Table 2 The association of the clinical presentation of the infection with TBEV with the presence of the specific anti-TBEV IgM antibodies in the acute neurologic phase cerebrospinal fluid (CSF) sample. Anti-TBEV IgM Presentation

positive

negative

M ME MEM

30 (39%) 36 (47%) 11 (14%)*

50 (46%) 55 (51%) 3 (3%)

M – meningitis. ME – meningoencephalitis. MEM – meningoencephalomyelitis. TBEV – tick-borne encephalitis virus. * Significantly higher frequency in the seropositive group with p < 0.05.

and 2 unconscious). Twenty two patients presented with cranial or spinal paresis. Eventually, 122 patients were discharged from hospital without any residual symptoms and neurologic deficits or were found asymptomatic during control visits in following weeks, 38 had subjective complaints weeks to months after discharge (headache, myalgia, tinnitus, irritability, impaired memory) but without evident alteration of the mental and neurologic status and 15 presented with objective long-term neurologic or psychiatric sequelae (paresis, altered behavior, cerebellar syndrome). 3.1.2. Serologic IgM data Only two patients were seronegative on admission and both seroconverted in a follow-.up examination. One of them presented with a flu-like infection and the other with a clinically unremarkable ME (severity score 3). Of 180 patients examined for the presence of anti-TBEV IgM in CSF on admission, the result was positive in 77 (43%). The anti-TBEV IgM score in serum did not correlate significantly with the time from the onset of symptoms, but was higher in patients with a history of biphasic than in monophasic disease (p < 0.05) (Fig. 1). There was no association of the disease duration or biphasic course with the anti-TBEV IgM presence or score in CSF .. Anti-TBEV IgM score in serum correlated negatively with age (R = −0.16, p < 0.05) and patients with no anti-TBEV IgM detectable in CSF on admission tended to be older (p = 0.059) (Fig. 2). There were no significant differences of the median serum antiTBEV IgM score between patients with different clinical manifestation,

3.1.3. Serologic IgG data Specific anti-TBEV IgG was studied in sera of 194 patients and in CSF of 174 study cohort patients. Of these 156 (80%) patients were IgGFig. 4. The negative correlation between the serum score of the specific anti-TBEV IgM antibodies (horizontal axis) and the cerebrospinal fluid (CSF) albumin quotient (Qalb, vertical axis) in TBE patients. Data from individual patients are show with points, the linear fit with a continuous line and 95% confidence interval with dashed lines. The strength and statistical significance of the correlation are presented directly on the plot.

6

Ticks and Tick-borne Diseases xxx (xxxx) xxx–xxx

S. Grygorczuk et al.

Fig. 5. The association of the blood and cerebrospinal fluid (CSF) inflammatory parameters with the presence and score of the specific anti-TBEV IgM antibodies in CSF of TBE patients on admission to hospital. Plots A, C, D and E compare the CSF IgM-negative (−) (n = 103) and CSF-IgM-positive (+) (n = 77) patients and show the median (horizontal line), quartiles (box) and minimum/maximum not-extreme values (whiskers) for the parameters compared between these two groups. A. Peripheral blood leukocytes on admission: total leukocytosis (C), lymphocytosis (L) and neutrophilia (N); # − the difference with p = 0.055; B. The correlation between the CSF specific IgM score in the IgM-positive patients (horizontal axis) and the peripheral blood neutrophil count (vertical axis). Data from individual patients are shown with points, the linear fit with a continuous line and 95% confidence interval with dashed lines. The strength and statistical significance of the correlation are presented directly on the plot. C, D. CSF inflammatory parameters on admission to hospital (I) and during follow-up examination before discharge (II): total cytosis (C), lymphocyte count (L), neutrophil count (N) expressed in cells/μl in C, total protein (P) and albumin (A) expressed in mg/dl in D. The extreme values were not shown for clarity. ** − significantly higher in the IgM-positive group with p < 0.01; **** − significantly higher in the IgM-positive group with p < 0.0001; ^ − significantly higher in IgM-negative group with p < 0.05; ^^ − significantly higher in the IgM-negative group with p = 0.01; E. The albumin quotient Qalb in a subgroup of patients in whom the serum albumin concentration was available: the lack of a difference between (−); (n = 16) and (+) (n = 11). F. The lack of correlation between the specific anti-TBEV IgM score in CSF and Qalb in a subgroup of patients with the positive result of the admission CSF serologic test (+ in E). Data from individual patients are show with points, the linear fit with a continuous line and 95% confidence interval with dashed lines. NS − non-significant.

7

Ticks and Tick-borne Diseases xxx (xxxx) xxx–xxx

S. Grygorczuk et al.

anti-TBEV IgG presence or score in serum or CSF. However, unlike with IgM, there was an apparent trend for a higher Qalb in CSF IgG-positive patients, which could have been rendered non-significant by a small number of patients available for Qalb analysis.

Table 3 The association of the clinical presentation of the infection with TBEV with the presence of the specific anti-TBEV IgM antibodies in the acute neurologic phase cerebrospinal fluid (CSF) sample. Anti-TBEV IgG

3.2. The prospective study Presentation

positive

negative

M ME MEM

66 (43%) 78 (51%) 9 (6%)

16 (41%) 15 (38%) 8 (21%)*

3.2.1. Patients’ clinical data There were 23 patients with M, 25 with ME and 2 with MEM. Eleven of ME and one MEM patient had a relatively mild presentation without evidently altered mental status or severe neurologic deficits, while the remaining 14 ME and 1 MEM patient had a moderately severe disease. There were 12 patients with altered consciousness and 6 with paresis as well as 24 patients with biphasic and 26 with monophasic presentation. None of the patients had been vaccinated against TBEV. In two patients the first lumbar puncture was performed in the peripheral phase of the disease and showed no CSF abnormalities. They were originally diagnosed with unspecific viral infection, but later developed the symptoms of the neurologic phase which prompted another lumbar puncture. In two others the original CSF was obtained very early in the neurologic phase and presented with minor abnormalities (pleocytosis of 11 and 25 cells/μl) − these patients subsequently deteriorated clinically and had another lumbar puncture performed within the following week, with deterioration of the CSF inflammatory parameters. The second samples taken during the maximum extent of CSF inflammation from these 4 patients were included in “examination I” batch, while 4 “early” sample pairs were also evaluated, however analyzed separately. Additional sample pairs obtained from 3 other patients (not initially included in the study group, two with meningitis and one with MEM) before fully developed neurologic phase were studied as well, together forming and additional “pre-neurologic phase” subgroup.

M – meningitis. ME – meningoencephalitis. MEM – meningoencephalomyelitis. TBEV – tick-borne encephalitis virus. * Significantly increased frequency in a seronegative group with p < 0.05. Table 4 The association of paresis in patients infected with TBEV with the presence of the specific anti-TBEV IgG antibodies in the acute neurologic phase serum sample. Anti-TBEV IgG Paresis

positive

negative

– absent – present

142 (93%) 11 (7%)

29 (74%) 10 (26%)**

TBEV – tick-borne encephalitis virus. ** Significantly increased frequency in a seronegative group with p < 0.01.

positive in serum and 25 (14%) in CSF on admission. As expected, patients with no anti-TBEV IgG in serum tended to have shorter history of clinical symptoms (p < 0.06, not shown). The presence and score of anti-TBEV IgG in serum and CSF did not differentiate the patients with a biphasic versus monophasic course. There was also no association of anti-TBEV IgG presence and score with age. Serum anti-TBEV IgG tended to associate with milder clinical presentation. The seronegative patients were significantly more likely to have MEM (p < 0.05) and to present with paresis (p < 0.01) compared to the seropositive ones (Tables 3, 4). In the seropositive group there was a trend towards lower antibody score in patients with more severe clinical form of TBE (p = 0.051, a significant difference in a direct comparison between M and ME with p < 0.05), with a higher clinical severity score (p = 0.06) and with altered consciousness. The difference between patients with normal and with altered mental status (the later pooled together irrespective of severity) was significant (p < 0.05) (Fig. 6A–C). The presence of the anti-TBEV IgG in CSF did not associate with clinical features. The analysis of sub-group differences in the CSF IgGpositive subgroup was hampered by small number of patients, but there was significantly lower anti-TBEV IgG score in ME in comparison with M (p < 0.05) (Fig. 6D). The absence of anti-TBEV IgG in serum, but not in CSF, associated with an incomplete recovery, both as residual subjective complaints (p < 0.01) and objective neuropsychiatric deficits (p < 0.001) (Table 5). In the seropositive patients, the anti-TBEV IgG median scores in serum and CSF did not further differ in patient groups stratified by outcome. The CSF neutrophil count was lower in patients positive for antiTBEV IgG in serum (p < 0.05), while other basic laboratory parameters did not differ significantly (Fig. 7). Patients with anti-TBEV IgG present in CSF had higher CSF lymphocyte count (p < 0.01), total protein (p < 0.05) and albumin (p < 0.05) concentration but lower CSF neutrophil count (p < 0.001) (Fig. 8). In seropositive patients none of these parameters correlated further with anti-TBEV IgG scores. The convalescent CSF parameters did not depend on the initial IgG-class serologic response. The CSF albumin quotient Qalb was not significantly linked to the

3.2.2. Serologic data The serology results on admission were not available in 3 sera and 6 CSF samples, but all these patients were documented to be IgM-positive before discharge from hospital. All the studied patients, including the ones hospitalized before the peak neurologic phase, had specific IgM detectable in serum on admission. Admission CSF was positive for specific IgM in 1 out of 4 patients (25%) examined in the early neurologic phase and in 14 of 43 patients (33%) examined during a fully developed neurologic phase. Anti-TBEV IgG was detectable in 3 early neurologic phase serum samples (60%) and in 37 out of 45 remaining sera (82%). It was detected in 6 CSF samples, including 5 out of 43 obtained in the developed neurologic phase (12%) and a weak positive result in a single early neurologic phase sample. The clinical and serologic data of TBE patients included in the prospective study are listed in Table 1. 3.2.3. IL-5 concentration The median concentration of IL-5 in serum and CSF is shown in Fig. 9A. In serum IL-5 concentration tended to be lower in TBE patients than in healthy controls, significantly for examination I samples (p < 0.05). To the contrary, there was a tendency for an increased IL-5 concentration in CSF in comparison with non-inflammatory CSF samples, present in the pre-neurologic phase (p < 0.001), non-significant during examination I and again significant in examination II and III time-points (p < 0.0001 and p < 0.01, respectively). Apparently, there was a tendency for higher serum and CSF IL-5 concentrations in examination II and III compared to examination I, but the difference was not statistically significant and the trend seems to be data artifact resulting from the patient selection. The subgroups of patients included in examination II and III had significantly increased IL-5 in CSF at examination I time-point as well (p < 0.01) and presented with a gradual decrease of serum and CSF IL-5 concentrations in 8

Ticks and Tick-borne Diseases xxx (xxxx) xxx–xxx

S. Grygorczuk et al.

Fig. 6. The score of the specific anti-TBEV IgG in serum (empty bars in plots A-C) and cerebrospinal fluid (light grey bars in plot D) on admission to hospital in seropositive TBE patients with different clinical presentations. Shown are the median (horizontal line), quartiles (box) and maximum values (whiskers) in the anti-TBEV IgG-positive patients. A. Patients with different clinical forms of TBEV infection: meningitis (M, n = 66), meningoencephalitis (ME, n = 78) and meningoencephalomyelitis (MEM, n = 9); * − the significant difference with p < 0.05 in a direct comparison of M and ME groups, the p = 0.051 value for the trend for all three groups in nonparametric ANOVA is presented directly on plot. B. Patients with a value of a disease severity score (as defined in “Material and methods) from 1 (n = 66) through 2 (n = 17), 3 (n = 42) to ≥4 (n = 28); the p = 0.06 value for the trend for all four groups in non-parametric ANOVA is presented directly on the plot. C. Patients with normal mental status (n = 107) and with consciousness disorders defined as mild (n = 16), moderate (n = 25) or severe (n = 4) during the hospital stay; * − the significant difference with p < 0.05 for comparison with all the patients with consciousness disorders analyzed as a single group; D. Patients with meningitis (M, n = 15) and meningoencephalitis (ME, n = 10), * − the significant difference with p < 0.05.

subsequent samples (p < 0.01 for the trend with ANOVA and for direct comparison between examination I and II, p < 0.05 for comparison between examination II and III, data for examination II shown in Fig. 9B). The selection bias may result from the facts that patients who underwent the second and especially the third lumbar puncture were qualified to them basing on clinical grounds, due to the persistent clinical symptoms or laboratory features of inflammation. It suggests that higher IL-5 levels were characteristic for patients with a tendency for a delayed clinical and laboratory improvement, although our remaining data do not completely support this association. The IL-5 concentration in admission serum and CSF of patients with aseptic non-TBE meningitis did not differ significantly from either control or TBE group. The median IL-5 concentration index in TBE patients at the examination I time-point was 8.7 (individual values from 2.0 to 35.8) suggestive of its intrathecal synthesis. IL-5 concentration in CSF did not

Table 5 The association of the clinical outcome of the infection with TBEV with the presence of the specific anti-TBEV IgG antibodies in the acute neurologic phase serum sample. Anti-TBEV IgG Outcome

positive

negative

– recovery – subjective symptoms – objective deficits

101 (76%) 25 (19%) 7 (5%)

15 (43%)^^ 13 (37%)** 7 (20%)***

TBEV – tick-borne encephalitis virus. ^^ Significantly decreased frequency in a seronegative group with p < 0.01. ** Significantly increased frequency in a seronegative group with p < 0.01. *** Significantly increased frequency in a seronegative group with p < 0.001.

9

Ticks and Tick-borne Diseases xxx (xxxx) xxx–xxx

S. Grygorczuk et al.

Fig. 7. The association of the cerebrospinal fluid (CSF) inflammatory parameters with the presence of the specific anti-TBEV IgG antibodies in serum of TBE patients on admission to hospital. Total cytosis (C), lymphocyte count (L) and neutrophil count (N) on admission to hospital (I) and during a follow-up examination before discharge (II) are presented in serum IgG-negative (−) (n = 38) and serum IgG-positive (+) (n = 156) patients, expressed in cells/μl. The median is shown with a horizontal line, quartiles with a box and minimum/maximum not-extreme values with whiskers. The extreme values were not shown for clarity. * − significantly higher in the IgG-negative group with p < 0.05.

Fig. 8. The association of the cerebrospinal fluid (CSF) inflammatory parameters with the presence and score of the specific anti-TBEV IgG antibodies in CSF of TBE patients on admission to hospital. A, B. CSF inflammatory parameters on admission to hospital (I) and during a follow-up examination before discharge (II): total cytosis (C), lymphocyte count (L), neutrophil count (N) expressed in cells/μl in A, total protein (P) and albumin (A) expressed in mg/dl in B, compared between the CSF IgG-negative (−) (n = 149) and CSF-IgG-positive (+) (n = 25) patients. Shown are the median (horizontal line), quartiles (box) and minimum/maximum not-extreme values (whiskers). The extreme values were not shown for clarity. * − significantly higher in the IgG-positive group with p < 0.05; ** − significantly higher in IgG-positive group with p < 0.01; *** − significantly lower in the IgG-positive group with p < 0.001; C. The albumin quotient Qalb in a subgroup of patients in whom the serum albumin concentration was available: the difference between (−) (n = 19) and (+) (n = 6) groups did not reach the level of a statistical significance. D. The lack of correlation between the specific anti-TBEV IgG score in CSF and Qalb in a subgroup of patients with the positive result of the admission CSF serologic test (+ in C). Data from individual patients are shown with points, the linear fit with a continuous line and 95% confidence interval with dashed lines. NS − non-significant.

10

Ticks and Tick-borne Diseases xxx (xxxx) xxx–xxx

S. Grygorczuk et al.

Fig. 9. The concentration of IL-5 in serum (S) and cerebrospinal fluid (CSF) of patients with tick-borne encephalitis (TBE), aseptic non-TBE meningitis (AM, n = 7) and in control non-inflammatory samples (C, n = 8); for TBE the samples were obtained at the peripheral phase or during the onset of the neurologic phase (0, n = 7), during the acute neurologic phase (I, n = 50, coinciding with the hospital admission in most of the patients), during early convalescent period 10–16 days after admission (II, n = 13) and during follow-up examinations 5–8 week after initial admission (III, n = 5) in selected patients. All AM samples were obtained on admission to hospital. Concentration expressed in pg/ml. Shown are the median (horizontal line), quartiles (box) and maximum values (whiskers). A. Data from the whole study and control groups; ^ − lower than simultaneously in C with p < 0.05; ** − higher than simultaneously in C with p < 0.01; *** − higher than simultaneously in C with p < 0.001; **** − higher than simultaneously in C with p < 0.0001; B. The decrease of the IL-5 concentration in serum (S) and CSF of the 13 TBE patients with both neurologic phase (I) and early convalescent (II) samples available; & − the significant difference with p < 0.05.

inflammatory parameters were relatively weak, but there was a significant negative correlation with blood lymphocytosis (R = −0.44, p < 0.01 for IL-5 in serum, R = −0.49, p < 0.001 for CSF) (Fig. 11) and a weak positive correlation with CSF pleocytosis (p < 0.05 for IL-5 in serum and CSF versus total pleocytosis, for IL-5 in serum versus CSF neutrophil count and IL-5 in CSF versus CSF lymphocyte count; not shown). There was no correlation with the CSF protein, albumin and Qalb and no detectable influence of the admission IL-5 values on the CSF parameters during follow-up (not shown). There was no correlation of either serum or CSF IL-5 concentration with the specific anti-TBEV IgM antibody score in serum and CSF and no difference of IL-5 concentration between the patients who were CSF IgM-positive and IgM-negative during examination I (Fig. 12). There was also no correlation of IL-5 concentrations with anti-TBEV IgG scores and no significant differences of IL-5 expression between serum and CSF IgG-positive and IgG-negative patients (Fig. 13).

correlate with Qalb, also pointing against a substantial influx from serum into CSF (not shown). The concentrations of IL-5 and values of IIL-5 in the subgroups of patients with different clinical features are shown in Fig. 10. The IL-5 concentration in serum and CSF on admission did not depend on the time from the disease onset, did not differ between patients with monophasic and biphasic disease and between patients with M and ME. When patients were stratified according to severity, the median serum IL-5 concentration was the highest in the intermediate group with the mild neurologic involvement and decreased both in patients with uncomplicated meningitis and with moderate to severe ME/MEM (p < 0.01 for the trend). The serum IL-5 concentration was lower in patients with altered consciousness (p = 0.01). The CSF concentration tended to decrease with the increasing severity (p = 0.068) and the CSF IL-5 index tended to be lower in patients presenting with paresis (p = 0.087) but the study lacked statistical power to confirm especially the latest trend (Fig. 10). The correlations of IL-5 concentration in serum and CSF with 11

Ticks and Tick-borne Diseases xxx (xxxx) xxx–xxx

S. Grygorczuk et al.

Fig. 10. The concentration of IL-5 in serum (S) and cerebrospinal fluid (CSF) expressed in pg/ml (left panels) and the index of the IL-5 CSF concentration IIL-5 (right side panels, calculated as in ‘Methods’) in TBE patients during the acute neurologic phase of the disease, stratified dependent on the clinical manifestation and severity. Shown are the median (horizontal line), quartiles (box) and maximum values (whiskers) or the values for individual patients (points) in case of very smalls groups (MEM in A − not included in a formal statistical analysis; IIL-5 in the paresis group in B). A. Clinical presentation: M − meningitis (n = 23), ME − meningoencephalitis (n = 25), MEM − meningoencephalomyelitis (n = 2) B. Presence (+) (n = 6) or absence (−) (n = 44) of paresis; # − the difference with p = 0.087; C. Presence (+) (n = 12) or absence (−) (n = 38) of the altered mental status; ** − the significant difference with p < 0.05; D. Clinical severity scored as 1 = no neurologic abnormalities (uncomplicated meningitis) (n = 23); 2 = mild CNS involvement (n = 12); 3 = moderate to severe CNS involvement (n = 15); ** − the trend for a higher concentration in (2) significant with p < 0.01; # − a tendency for a decrease with severity with p = 0.068. The size of the groups is given for the left side (concentration) panels and could be lower in the IIL-5 charts because of the missing data used to calculate IIL-5.

4. Discussion We analyzed the early peripheral and intrathecal expression of the specific anti-TBEV IgM and IgG antibodies in a cohort of well clinically characterized TBE patients. We found the peripheral serologic IgM response to be associated with a complete recovery and to correlate with an attenuated BBB disruption. Patients with a clinically overt peripheral phase of the disease had stronger IgM humoral response by the time of the neurologic phase, with no impact on clinical outcome. The role of intrathecal specific IgM was less clear, as it was associated with paresis and with a myelitic presentation of the disease, as well as with more pronounced inflammatory changes in CSF, but also correlated with a quicker resolution of the CSF cellular infiltrate in follow-up examinations. The systemic IgG response was detectable in the majority of patients and associated with milder presentation and better clinical outcome. On the other hand the CSF anti-TBEV IgG was detected in a small fraction of patients only, in association with more intensive CSF inflammatory changes and BBB disruption and with no evident clinical or prognostic significance. Our results are consistent with previous animal and clinical studies, which have suggested that a humoral response could be protective by controlling the peripheral infection and limiting CNS entry of TBEV, while the intrathecal response could protect against a neuronal damage, but may be also a hallmark of an intense and protracted CNS inflammation in a severe infection. The results from animal models show that a humoral response plays an important protective role in infections caused by Flaviviridae, including TBEV (Palus et al., 2013; Bréhin et al., 2008; Diamond et al., 2003a,b; Elsterova et al., 2017). The administration of the intravenous immunoglobulin containing specific TBEVneutralizing antibodies may dramatically increase survival in mice challenged with TBEV (Elsterova et al., 2017). One of the most evident differences between TBE-sensitive and TBE-resistant mouse strains is an early appearance and a high final titer of the anti-TBEV neutralizing antibodies in sera of the resistant animals, compared to the delayed and weak response in the susceptible strain (Palus et al., 2013). Studies addressing a correlation of the peripheral humoral response with the clinical presentation in TBE patients also point to its protective effect, especially at the early stage of infection (Atrasheuskaya et al., 2003; Günther et al., 1997; Kaiser and Holzmann, 2000). In the study of Günther et al. (1997) low specific antibody titers in the second week of the disease were associated with moderate to severe encephalitis, and according to Kaiser and Holzmann (2000) low serum levels of TBEVneutralizing antibodies correlate with severe clinical course of the disease. Interestingly, both in animal models of flavivirus encephalitis and in human TBE studies the IgM-class response was more evidently related to the favorable clinical course than its IgG counterpart (Diamond et al., 2003a; Günther et al., 1997; Toporkova et al., 2008). In a study of Toporkova et al. (2008) the serum specific IgG titer on the 5th-7th day after disease onset did not correlate with the clinical 12

Ticks and Tick-borne Diseases xxx (xxxx) xxx–xxx

S. Grygorczuk et al.

Fig. 11. The negative correlation of the concentration of IL-5 in the cerebrospinal fluid (CSF) of patients in the acute neurologic phase of TBE (horizontal axis, expressed in pg/ml) with the simultaneous peripheral blood lymphocyte count (vertical axis, expressed as a number of cells/μl). Data from individual patients are shown with points, the linear fit with a continuous line and 95% confidence interval with dashed lines. The strength and statistical significance of the correlation are presented directly on the plot. Two extreme cases have been removed from the plot for clarity, but were included in the analysis.

the control of the infection (Günther et al., 1997; Kaiser and Holzmann, 2000; Kaiser, 2002) and our results are generally consistent with these observations, although the findings of relatively high titers in a small group of CSF IgG-positive patients with meningitis are unexpected in that context. Our detection of the association of myelitis with an early intrathecal IgM expression is more puzzling, although to some extent it parallels the results of the study of Günther et al. (1997) in which the persistent intrathecal synthesis of the specific IgG was characteristic for patients with neurologic involvement, especially with myelitis. Previously, we observed higher protein concentration and neutrophil counts in CSF of patients with MEM compared to other clinical forms of TBE (unpublished results). The data on intrathecal IgM are another piece of evidence suggesting a specific pathogenesis of the spinal involvement, different from the more common meningeal and encephalitis presentations of TBE. We found the peripheral and intrathecal IgM response to be weaker in older TBE patients. This tendency was expectable, as the diversity of the B-cell repertoire is known to decrease with age, in spite of the generally preserved B cell numbers in periphery (Weksler and Szabo, 2000). The age is also one of the markers of poor prognosis in TBE, with higher frequency of the severe neurologic manifestations in the elderly (Czupryna et al., 2011; Kaiser, 2002). As the peripheral humoral response seems protective in TBEV infection, its observed decrease with age could offer a causative explanation of the increasing susceptibility. The strength of our retrospective study was limited by a very small number of IgM-seronegative patients on admission. The diagnostic criteria prompted us to exclude patients with no serologic confirmation of TBE, some of whom could in fact have been infected with TBEV but responded with the production of the specific antibodies with delay. Another limitation was a very small number of patients from both the extremes of the clinical spectrum of TBE, with the severe manifestation of encephalitis as well as with the flu-like infection, not allowing for their systematic comparison with the rest of study cohort. It is tempting to speculate that in the mild flu-like TBEV infection the peripheral serologic response is a factor preventing the disease progression into the neurologic phase, which could not be verified with our patient group. Our study was not designed to assess the intensity of the intrathecal versus peripheral synthesis of anti-TBEV antibodies precisely, as the records of hospitalized patients lacked the necessary data. This

severity or outcome. In comparison, possibly because of evaluating a relatively large patient cohort, we observed a consistent association of the systemic anti-TBEV IgG response, as assessed a few days after the onset of the disease, with mild presentation and good outcome. The unfavorable course of TBE in patients with a weak or delayed peripheral humoral response could be mediated by several mechanisms. According to Palus et al. (2013), the humoral response in resistant mice controls the infection in the periphery, thus preventing neuroinvasion, but also may have a secondary modulatory effect on the intrathecal inflammation. The higher viremia resulting from weaker peripheral response could facilitate TBE migration across BBB and increase severity of CNS involvement. TBEV is able to cross BBB by causing a persistent infection and replicating within the brain microvascular endothelial cells (Palus et al., 2017), a process that could be enhanced by higher viremia. Severe manifestations of TBE could also result from harmful pro-inflammatory effects of shifting immune balance from humoral towards Th1 and cytotoxic responses. As suggested by animal models of WNV encephalitis and limited clinical observations in TBE, the CNS invasion may be facilitated by the BBB inflammatory disruption (Arjona et al., 2007; Grygorczuk et al., 2017; Wang et al., 2004) and our current results show that the later may be indeed more severe in patients with a weaker systemic humoral response in IgM class. With that respect, Toporkova et al. (2008) have previously detected an inverse correlation between serum concentration of a pro-inflammatory cytokine IL-6 and specific anti-TBEV IgG antibody titers. The role of the intrathecal humoral anti-TBEV response is less clear, although animal data suggest it may be protective and counterbalance the pathogenic cytotoxic and inflammatory responses (Palus et al., 2013). According to Günther et al. (1997) the early intrathecal specific IgM expression decreased the probability of the encephalitic syndrome, but the following studies by Kaiser and Holzmann (2000), Kaiser (2002) did not confirm that association and even found the specific intrathecal IgM synthesis to be more frequent in patients with more severe TBE. Our results offer a possible explanation of this discrepancy, as antiTBEV IgM associated both with a more active intrathecal inflammation and with quicker improvement, suggesting it may constitute a part of the protective response to a severe CNS infection. On the other hand, the intrathecal expression of anti-TBEV IgG has been previously described as the late phenomenon rather to the immunopathology than to 13

Ticks and Tick-borne Diseases xxx (xxxx) xxx–xxx

S. Grygorczuk et al.

Fig. 12. The lack of significant dependence of the anti-TBEV IgM expression on the IL-5 concentration in TBE patients. A. Lack of the difference of the median IL-5 concentration in CSF (expressed in pg/ml) between patients with (+) (n = 15) and without (−) (n = 32) specific anti-TBEV IgM detectable in CSF. Shown are the median (horizontal line), quartiles (box) and extreme values (whiskers). B. Like A, but presenting IL-5 concentration in serum. C. Concentration of IL-5 in serum (expressed in pg/ml) on horizontal axis, specific anti-TBEV IgM score in serum on vertical axis. Data from individual patients are show with points, the linear fit with a continuous line and 95% confidence interval with dashed lines. No significant correlation present. D. Like C, but a specific anti-TBEV IgM score in CSF is shown on vertical axis. E. Like C, but a specific anti-TBEV IgM score in CSF is shown on vertical axis and IL-5 concentration in CSF on horizontal axis. Excluding patients with undetectable anti-TBEV IgM did not change the results presented in D and E. Excluding 5 patients who were studied serologically earlier than the rest of the group, at the onset of the neurologic phase, did not change the results either.

diagnostic kit only, which was a practical solution in a large group of patients, but the detailed study of anti-TBEV neutralizing antibodies could give more insight into the effectiveness of the humoral response. We have confirmed the intrathecal expression of IL-5 in TBE

might have especially affected the detection of the anti-TBEV IgG in CSF, which tended to correlate with markers of BBB disruption and thus could be influenced by the leakage of the specific antibodies from serum. Finally, the serologic response was assessed with a commercial 14

Ticks and Tick-borne Diseases xxx (xxxx) xxx–xxx

S. Grygorczuk et al.

Fig. 13. The lack of significant dependence of the anti-TBEV IgG expression with the IL5 concentration in TBE patients. A. Lack of significant difference of the median IL-5 concentrations in serum (S, empty bars) and cerebrospinal fluid (CSF, light grey bars) between patients with (+) (n = 40) and without (−) (n = 10) specific anti-TBEV IgG detectable in serum. Shown are the median (horizontal line), quartiles (box) and extreme values (whiskers). B. Like A, but for patients stratified according to the presence (+) (n = 6) or absence (−) (n = 41) of the specific anti-TBEV IgG in cerebrospinal fluid. C. Concentration of IL-5 in serum (expressed in pg/ml) on horizontal axis, specific anti-TBEV IgG score in serum on vertical axis. Data from individual patients are shown with points, the linear fit with a continuous line and 95% confidence interval with dashed lines. No significant correlation present. Excluding data from patients with undetectable serum anti-TBEV IgG and/or a single outlying data point did not change the result. Excluding 5 patients who were serologically studied earlier than the rest of the group, at the onset of the neurologic phase, did not change the results either.

TBEV IgM or IgG antibody titers, meaning that the activity of this cytokine in TBE is probably not critical for boosting humoral responses. This contradicted our initial hypothesis, as we have decided to assess expression of IL-5 considering it as a potential mediator and/or marker of a humoral response in TBE, based on its effect on B cell differentiation and maturation (Horikawa et al., 2001; Takatsu, 2011). The negative result means that the correlates and predictors of the effective humoral response against TBEV remain to be defined by future studies. The lack of correlation between IL-5 and humoral immunity, however, to some degree parallels the results obtained in an animal study by Palus et al. (2013), who compared the response to TBEV in mouse strains of variable susceptibility. The expression of mRNA for CD19, the marker of B lymphocytes, was higher in the brain tissue of the resistant mice, consistent with their protective role that could be exerted through an intrathecal synthesis of the neutralizing antibodies. Interestingly, IL5 mRNA was expressed only in the brains of sensitive animals, which in the same time failed to develop an effective humoral response, and in concert with pro-inflammatory cytokines, suggestive of its contribution rather to the immune-mediated CNS damage than to neutralizing humoral immunity (Palus et al., 2013). Our clinical data suggest a more complicated association of peripheral and intrathecal IL-5 expression with the course of TBE, with possible protective effects. Both the animal model and our results show IL-5 may be importantly involved in TBE pathogenesis, but the mechanisms of its activity and balance of protective and harmful effects must be clarified in further studies.

patients, peaking before or at the onset of the neurologic phase and gradually decreasing thereafter. The relation of IL-5 expression with the clinical presentation was not straightforward, as it seemed to associate both with a mild clinical presentation and with a tendency for protracted intrathecal inflammation. It also correlated negatively with the peripheral blood lymphocyte count and positively with CSF pleocytosis. Previously, there were few and not fully consistent data on IL-5 expression in infections with neurotropic flaviviruses and on its intrathecal expression in neuroinfections in general. IL-5 was undetectable in serum of most of the recipients of the attenuated JE virus vaccine (Zhang et al., 2012), but T CD4+ cells from the persons vaccinated against TBEV were able to secret it under stimulation with TBEV proteins in vitro (Gomez et al., 2003). Holub et al. (2006) did not detect IL-5 in CSF of 37 patients with aseptic meningitis, including TBE, while Cerar et al. (2013) found it at low, but detectable levels in serum and CSF in a group of 39 TBE patients. In a study by Pietikäinen et al. (2016), IL-5 was undetectable in most of the TBE CSF samples studied. Our results not only show an intrathecal expression of IL-5 but also for the first time hint at its role in the pathogenesis and influence on leukocyte distribution. The patients hospitalized in the neurologic phase of TBE usually present with temporarily lowered blood lymphocyte count, which has not been systematically studied so far (Grygorczuk et al., 2016). Our current data show that the increased IL-5 expression may be involved in a migration of lymphocytes into CSF and their accumulation there, with the simultaneous depletion or suppression in periphery, although the underlying mechanism remains to be explained. We did not find a correlation of IL-5 levels with the specific anti15

Ticks and Tick-borne Diseases xxx (xxxx) xxx–xxx

S. Grygorczuk et al.

5. Conclusions

Innis, B.L., Nisalak, A., Nimmannitya, S., Kusalerdchariya, S., Chongswasdi, V., Suntayakorn, S., Puttisri, P., Hoke, C.H., 1989. An enzyme-linked immunosorbent assay to characterize dengue infections where dengue and Japanese encephalitis cocirculate. Am. J. Trop. Med. Hyg. 40, 418–427. Jeren, T., Vince, A., 1998. Cytologic and immunoenzymatic findings in CSF from patients with tick-borne encephalitis. Acta Cytol. 42, 330–334. Kagami, S., Nakajima, H., Kumano, K., Suzuki, K., Suto, A., Imada, K., Davey, H.W., Saito, Y., Takatsu, K., Leonard, W.J., Iwamoto, I., 2000. Both Stat5a and Stat5b are required for antigen-induced eosinophil and T-cell recruitment into the tissue. Blood 95, 1370–1377. Kaiser, R., Holzmann, H., 2000. Laboratory findings in tick-borne encephalitis −correlation with clinical outcome. Infection 28, 78–84. Kaiser, R., 2002. Tick-borne encephalitis (TBE) in Germany and clinical course of the disease. Int. J. Med. Microbiol. 33, 58–61. Kothur, K., Wienholt, L., Mohammad, S.S., Tantsis, E.M., Pillai, S., Britton, P.N., Jones, C.A., Angiti, R.R., Barnes, E.H., Schlub, T., Bandodkar, S., Brilot, F., Dale, R.C., 2016. Utility of CSF cytokine/chemokines as markers of active intrathecal inflammation: comparison of demyelinating, anti-NMDAR and enteroviral encephalitis. PLoS One 11, e0161656. Kouro, T., Kikuchi, Y., Kanazawa, H., Hirokawa, K., Harada, N., Shiiba, M., Wakao, H., Takaki, S., Takatsu, K., 1996. Critical proline residues of the cytoplasmic domain of the IL-5 receptor alpha chain and its function in IL-5-mediated activation of JAK kinase and STAT5. Int. Immunol. 8, 237–245. Libraty, D.H., Nisalak, A., Endy, T.P., Suntayakorn, S., Vaughn, D.W., Innis, B.L., 2002. Clinical and immunological risk factors for severe disease in Japanese encephalitis. Trans. R. Soc. Trop. Med. Hyg. 96, 173–178. Mansfield, K.L., Johnson, N., Phipps, L.P., Stephenson, J.R., Fooks, A.R., Solomon, T., 2009. Tick-borne encephalitis virus − a review of an emerging zoonosis. J. Gen. Virol. 90, 1781–1794. Martinez-Moczygemba, M., Huston, D.P., Lei, J.T., 2007. JAK kinases control IL-5 receptor ubiquitination, degradation and internalization. J. Leukoc. Biol. 81, 1137–1148. Meads, M.B., Li, Z.W., Dalton, W.S., 2010. A novel TNF receptor-associated factor 6 binding domain mediates NF-кB signaling by the common cytokine receptor beta subunit. J. Immunol. 185, 1606–1615. Mickiené, A., Laiškonis, A., Günther, G., Vene, S., Lundkvist, A., Lindquist, L., 2002. Tickborne encephalitis in an area of high endemicity in Lithuania: disease severity and long-term prognosis. Clin. Infect. Dis. 35, 650–658. Ogata, N., Kouro, T., Yamada, A., Koike, M., Hanai, N., Ishikawa, T., Takatsu, K., 1998. JAK2 and JAK1 constitutively associate with an interleukin 5 (IL-5) receptor and βc subunit, respectively and are activated upon IL-5 stimulation. Blood 91, 2264–2271. Palus, M., Vojtíšková, J., Salát, J., Kopecký, J., Grubhoffer, L., Lipoldová, M., Demant, P., Růžek, D., 2013. Mice with different susceptibility to tick-borne encephalitis virus infection show selective neutralizing antibody response and inflammatory reaction in the central nervous system. J. Neuroinflammation 10, 77–89. Palus, M., Bílý, T., Elsterová, J., Langhansová, H., Salát, J., Vancová, M., Růžek, D., 2014. Infection and injury of human astrocytes by tick-borne encephalitis virus. J. Gen. Virol. 95, 2411–2426. Palus, M., Mancova, M., Sirmarova, J., Elsterova, J., Perner, J., Růžek, D., 2017. Tickborne encephalitis virus infects human brain microvascular endothelial cells without compromising blood-brain barrier integrity. Virology 507, 110–122. Pietikäinen, A., Maksimov, M., Kauko, T., Hurme, S., Salmi, M., Hytönen, J., 2016. Cerebrospinal fluid cytokines in Lyme neuroborreliosis. J. Neuroinflammation 13, 273–282. Růžek, D., Salát, J., Palus, M., Gritsun, T.S., Gould, E.A., Dyková, I., Skallová, A., Jelínek, J., Kopecký, J., Grubhoffer, L., 2009. CD8+ T-cells mediate immunopathology in tick-borne encephalitis. Virology 384, 1–6. Randolph, S.E., 2001. The shifting landscape of tick-borne zoonoses: tick-borne encephalitis and Lyme borreliosis in Europe. Phil. Trans. R. Soc. Lond. 356, 1045–1056. Satoh, S., Katagiri, T., Takaki, S., Kikuchi, Y., Hitoshi, Y., Yonehara, S., Tsukada, S., Kitamura, D., Watanabe, T., Witte, O.N., Takatsu, K., 1995. IL-5 receptor-mediated tyrosine phosphorylation of SH2/SH3-containing proteins and activation of Bruton's tyrosine and Janus-2 kinases. J. Exp. Med. 180, 2101–2111. Schellinger, P.D., Schmutzhard, E., Fiebach, J.B., Pfausler, B., Maier, H., Schwab, S., 2000. Poliomyelitic-like illness in central European encephalitis. Neurology 55, 299–302. Takaki, S., Kanazawa, H., Shiiba, M., Takatsu, K., 1994. A critical cytoplasmic domain of the interleukin-5 (IL-5) receptor α chain and its function in IL-5-mediated growth signal transduction. Mol. Cell. Biol. 14, 7404–7413. Takatsu, K., 2011. Interleukin-5 and IL-5 receptor in health and diseases. Proc. Japan Acad. Ser. B Phys. Biol. Sci. 87, 463–485. Toporkova, M.G., Aleshin, S.E., Ozherelkov, S.V., Nadezhdina, M.V., Stephenson, J.R., Timofeev, A.V., 2008. Serum levels of interleukin 6 in recently hospitalized tickborne encephalitis patients correlate with age, but not with disease outcome. Clin. Exp. Immunol. 152, 517–521. Wang, T., Town, T., Alexopoulou, L., Anderson, J.F., Fikrig, E., Flavell, R.A., 2004. Tolllike receptor 3 mediated West Nile virus entry into the brain causing lethal encephalitis. Nat. Med. 10, 1366–1373. Weksler, M.E., Szabo, P., 2000. The effect of age on the B-cell repertoire. J. Clin. Immunol. 20, 240–249. Zahn, S., Godillot, P., Yoshimura, A., Chaiken, I., 2000. IL-5-induced JAB–JAK2 interaction. Cytokine 12, 299–306. Zajkowska, J., Moniuszko-Malinowska, A., Pancewicz, S., Muszyńska-Mazur, A., Kondrusik, M., Grygorczuk, S., Świerzbińska-Pijanowska, R., Dunaj, J., Czupryna, P., 2011. Evaluation of CXCL10, CXCL11, CXCL12 and CXCL13 chemokines in serum and cerebrospinal fluid in patients with tick borne encephalitis (TBE). Adv. Med. Sci. 56, 311–317. Zhang, J.S., Zhao, Q.-M., Zuo, S.-Q., Jia, N., Guo, X.-F., 2012. Cytokine and chemokine responses to Japanese encephalitis live attenuated vaccine in a human population. Int. J. Infect. Dis. 16, e285–e288.

1 The specific peripheral IgM response in TBE is associated with a better outcome 2 The early specific peripheral IgG response is associated with milder clinical presentation and complete recovery. 3 The intrathecal anti-TBEV IgM response is associated with quicker resolution of the cellular CSF infiltrate. 4 IL-5 is synthesized intrathecally and may play a protective role in CNS during TBE, but its expression does not correlate with the specific humoral response. Acknowledgement The study was founded through the grant from the Medical University of Białystok (grant no N/ST/ZB/15/003/1145) References Arjona, A., Foellmer, H.G., Town, T., Leng, L., McDonald, C., Wang, T., Wong, S.J., Montgomery, R.R., Fikrig, E., Bucala, R., 2007. Abrogation of macrophage migration inhibitory factor decreases West Nile virus lethality by limiting viral neuroinvasion. J. Clin. Invest. 117, 3059–3066. Atrasheuskaya, A.V., Fredeking, T.M., Ignatyev, G.M., 2003. Changes in immune parameters and their correction in human cases of tick-borne encephalitis. Clin. Exp. Immunol. 131, 148–154. Bréhin, A.-C., Mouriès, J., Frenkiel, M.-P., Dadaglio, G., Desprès, P., Lafon, M., Couderc, T., 2008. Dynamics of immune cell recruitment during West Nile encephalitis and identification of a new CD19+B220-BST-2+ leukocyte population. J. Immunol. 180, 6760–6767. Cerar, T., Ogrinc, K., Lotric-Furlan, S., Kobal, J., Levicnik-Stezinar, S., Strle, F., RuzićSabljic, E., 2013. Diagnostic value of cytokines and chemokines in lyme neuroborreliosis. Clin. Vaccine Immunol. 20, 1578–1584. Czupryna, P., Moniuszko, A., Pancewicz, S.A., Grygorczuk, S., Kondrusik, M., Zajkowska, J., 2011. Tick-borne encephalitis in Poland in years 1993–2008 – epidemiology and clinical presentation. A retrospective study of 687 patients. Eur. J. Neurol. 18, 673–679. Diamond, M.S., Shrestha, B., Marri, A., Mahan, D., Engle, M., 2003a. B cells and antibody play critical roles in the immediate defense of disseminated infection by West Nile encephalitis virus. J. Virol. 77, 2578–2586. Diamond, M.S., Sitati, E.M., Friend, L.D., Higgs, S., Shrestha, B., Engle, M., 2003b. A critical role for induced IgM in the protection against West Nile virus infection. J. Exp. Med. 198, 1853–1862. Elsterova, J., Palus, M., Sirmarova, J., Kopecky, J., Niller, H.H., Ruzek, D., 2017. Tickborne encephalitis virus neutralization by high dose intravenous immunoglobulin. Ticks Tick Borne Dis. 8, 253–258. Günther, G., Haglund, M., Lindquist, L., Sköldenberg, B., Fosgren, M., 1996. Intrathecal production of neopterin and beta 2 microglobulin in tick-borne encephalitis (TBE) compared to meningoencephalitis of other ethiology. Scand. J. Infect. Dis. 28, 1318. Günther, G., Haglund, M., Lindquist, L., Sköldenberg, B., Forsgren, M., 1997. Intrathecal IgM, IgA and IgG antibody response in tick-borne encephalitis: long-term follow-up related to clinical course and outcome. Clin. Diagn. Virol. 8, 17–29. Gelpi, E., Preusser, M., Laggner, U., Garzuly, F., Holzmann, H., Heinz, F.X., Budka, H., 2006. Inflammatory response in human tick-borne encephalitis: analysis of postmortem brain tissue. J. Neurovirol. 12, 322–327. Gomez, I., Marx, F., Saurwein-Teissl, M., Gould, E.A., Grubeck-Loebenstein, B., 2003. Characterization of tick-borne encephalitis virus-specific human T lymphocyte responses by stimulation with structural TBEV proteins expressed in a recombinant baculovirus. Viral Immunol. 16, 407–414. Grygorczuk, S., Osada, J., Parczewski, M., Moniuszko, A., Świerzbińska, R., Kondrusik, M., Czupryna, P., Dunaj, J., Dabrowska, M., Pancewicz, S., 2016. The expression of the chemokine receptor CCR5 in tick-borne encephalitis. J. Neuroinflammation 13, 45. Grygorczuk, S., Parczewski, M., Świerzbińska, R., Czupryna, P., Moniuszko, A., Dunaj, J., Kondrusik, M., Pancewicz, S., 2017. The increased concentration of macrophage migration inhibitory factor in serum and cerebrospinal fluid of patients with tickborne encephalitis. J. Neuroinflammation 14, 126. Gustafson, R., Svenungsson, B., Fosgren, M., Gardulf, A., Granstrom, M., 1992. Twoyear survey of the incidence of Lyme borreliosis and tick-borne encephalitis in a high-risk population in Sweden. Eur. J. Clin. Microbiol. Infect. Dis. 11, 894–900. Holub, M., Klučková, Z., Beran, O., Aster, V., Lobovská, A., 2002. Lymphocyte subset numbers in cerebrospinal fluid: comparison of tick-borne encephalitis and neuroborreliosis. Acta Neurol. Scand. 106, 302–308. Holub, M., Beran, O., Lacinová, Z., Cinek, O., Chalupa, P., 2006. Interferon-gamma and cortisol levels in cerebrospinal fluid and its relationship to the etiology of aseptic meningoencephalitis. Prague Med. Rep. 107, 343–353. Holzmann, H., 2003. Diagnosis of tick-borne encephalitis. Vaccine 1 S36-S40. Horikawa, K., Kaku, H., Nakajima, H., Davey, H.W., Hennighausen, L., Iwamoto, I., Yasue, T., Kariyone, A., Takatsu, K., 2001. Essential role of Stat5 for IL-5-dependent IgH switch recombination in mouse B cells. J. Immunol. 167, 5018–5026. Huang, H.M., Lee, Y.L., Chang, T.W., 2006. JAK1 N-terminus binds to conserved Box 1 and Box 2 motifs of cytokine receptor common beta subunit but signal activation requires JAK1 C-terminus. J. Cell. Biochem. 99, 1078–1084.

16