Increased Plasma Levels of Pro- and Anti-inflammatory Cytokines in Patients with Febrile Seizures

Authors


Address correspondence and reprint requests to Dr. M. Helminen at Department of Pediatrics, Tampere University Hospital, P.O. Box 2000, FIN-33521, Tampere, Finland. E-mail:merja.helminen@uta.fi

Abstract

Summary:  Purpose: Pro- and antiinflammatory cytokines regulate the febrile response during infection. Febrile seizures (FSs) conversely are associated with rapid onset of high fever. Activation of the cytokine network has been shown in previous studies of FSs and cytokines. In this study, the association between cytokines and FSs was further investigated.

Methods: Interleukin-1β (IL-1β), interleukin-1 receptor antagonist (IL-1RA), interleukin-6 (IL-6), interleukin-10, and tumor necrosis factor-α plasma levels were measured with enzyme-linked immunosorbent assay in 55 children with FSs and in 20 age-matched febrile controls immediately on arrival at the hospital. Cerebrospinal fluid cytokine levels also were measured in 16 FS children.

Results: The plasma IL-1RA/IL-1β ratio (mean, 2,133 vs. 119; median, 790 vs. 105; p < 0.0001) and plasma IL-6 (mean, 41.7 pg/ml vs. 16.1 pg/ml; median, 19.6 pg/ml vs. 10.5 pg/ml; p = 0.005) were significantly higher in FS patients compared with control children. Logistic regression analysis was used to find the most significant predisposing factors for FSs. In this analysis, the high plasma IL-1RA/IL-1β ratio was the most significant factor connected to FSs (OR, 41.5; 95% CI, 4.9–352.8), but high plasma IL-6 also was significantly associated with FSs (OR, 5.3; 95% CI, 1.4–20.3).

Conclusions: Present results support the hypothesis that the cytokine network is activated and could have a role in the pathogenesis of FS.

Febrile seizures (FSs) are the most common type of convulsions in childhood (1). There is evidence that genetics is involved in the pathogenesis of FSs (1,2). In family studies, several loci associated with FSs have been found (2). In our previous study, we found increased frequency of interleukin-1β (IL-1B, single nucleotide polymorphism at the position -511) allele 2 in FS children compared with healthy controls (3).

During infection, both pro- and antiinflammatory cytokines are produced (4). Proinflammatory cytokines such as IL-1β, tumor necrosis factor-α (TNF-α), and interleukin-6 (IL-6) participate in inducing acute-phase reaction including fever. Antiinflammatory cytokines such as IL-1 receptor antagonist (IL-1RA) and IL-10 have a negative feedback effect during the febrile response (4,5). The balance between these cytokines influences the level of fever and could therefore have a role in the pathogenesis of FSs.

In previous studies of cytokines and FSs, activation of the cytokine network has been shown (6–9). In this study, we wanted to analyze further the association between cytokines and FSs.

MATERIALS AND METHODS

The study was carried out between October 1997 and March 2000 at the Tampere University Hospital and University of Tampere Medical School. The study protocol was approved by the ethical committee of the hospital. Informed consent was obtained from the parents.

Blood samples were obtained from children with FSs (n = 55) and children with febrile illness without convulsions (n = 20). Inclusion criteria for FS patients were age from 6 months to 5 years, temperature ≥38.5°C, and no other identifiable cause for the seizure. The control group was matched for age and temperature and had no convulsion during the illness and no known history of FSs. Data regarding family history for FSs, earlier FSs, duration of the seizure, and duration of fever before seizure were obtained from the parents by using a questionnaire. Family history was regarded as positive when seizure was reported in first-degree relatives. Laboratory and clinical data including C-reactive protein (CRP), leukocytes, antibiotic treatment, duration of fever before obtaining the blood and cerebrospinal fluid (CSF) samples, and fever when hospitalized were obtained from the medical records. CSF samples were obtained from 16 FS children based on the clinical judgment of the attending pediatrician. The CSF samples had cell count <5/μl and normal protein level (0.24–0.50 g/L).

For statistical analysis, infections were divided into two groups: viral and bacterial. Viral infection was defined as an infection with fever, low CRP, and no requirement for antibiotic treatment. Bacterial infections were treated with antibiotics and included both focal and septic infections.

IL-1β, IL-1RA, IL-6, IL-10, and TNF-α plasma and CSF levels of FS patients and plasma levels of control patients were measured by using an enzyme-linked immunosorbent assay according to the manufacturer's instructions (for IL-1β, IL-6, IL-10, and TNF-α, a CLB kit, Amsterdam, the Netherlands; and for IL-1RA, RD systems Quantikine kit, Minneapolis, MN, U.S.A.). The detection limits were 0.4 pg/ml for IL-1β, 46.9 pg/ml for IL-1RA, 0.6 pg/ml for IL-6, 1.2 pg/ml for IL-10, and 1.4 pg/ml for TNF-α.

Statistical analysis

The χ2 test was used to compare the clinical characteristics of FS patients and the controls. Cytokine plasma levels, fever when hospitalized, and laboratory results were compared by using the Mann–Whitney U test. The Bonferroni correction was used for p values of plasma cytokine levels except for the IL-1RA/IL-1β ratio. Binary logistic regression analysis was used to define associations between FSs, cytokine plasma levels, and clinical parameters. For this analysis, cytokine plasma levels, age, and fever at the time of hospitalization were divided into two groups by using median value as a cut-off point. Statistical calculations were performed by using SPSS for Windows, version 9.0. (SPSS Inc., Chicago, IL, U.S.A.).

RESULTS

FS patients and controls

Fifty-five FS patients and 20 control children with febrile illness without convulsion were included in the study. Fifteen (27%) FS children had a family history of FSs, and eight (15%) had a of previous FS. FS children and the controls did not differ by sex, age, type of infection, duration of fever before the blood sample, fever when hospitalized, or laboratory data. The clinical and laboratory data of FS and control children are summarized in Table 1.

Table 1.  Clinical and laboratory characteristics of febrile seizure (FS) and control children
 FS
(N = 55)
Controls
(N = 20)
p-value
  • a

     Mean ± SD

Age (mo)23.9 ± 12.422.1 ± 11.50.6
Fever when hospitalized (°C)a39.2 ± 1.039.2 ± 0.90.99
C-reactive protein (mg/l)a13.1 ± 18.535.1 ± 47.60.09
Leukocytes (×109/l)a10.8 ± 5.514.6 ± 20.80.9
Gender (M/F)25/308/120.3
Type of infection   
 Bacterial/Viral18/378/120.6
Duration of fever before the
 blood sample (<24h/>24h)
40/1512/80.3
Duration of convulsion   
 <5 min37  
 5–15 min16  
 >15 min2  

Plasma cytokine levels

The blood and CSF samples were taken immediately on arrival at the hospital. FS patients had significantly higher plasma IL-1RA levels (mean, 16,013 pg/ml vs. 4,210 pg/ml; median, 8,450 pg/ml vs. 2,860 pg/ml; p = 0.0005), higher plasma IL-1RA/IL-1β ratio (mean, 2,133 vs. 119; median, 790 vs. 105; p < 0.0001), and higher plasma IL-6 levels (mean, 41.7 pg/ml vs. 16.1 pg/ml; median, 19.6 pg/ml vs. 10.5 pg/ml; p = 0.005) than did control children. There was a trend of lower plasma IL-1β levels in FS children when compared with controls (mean, 14.4 pg/ml vs. 66.1 pg/ml; median, 10.1 pg/ml vs. 24.9 pg/ml; p = 0.1), but the difference was not statistically significant after the Bonferroni correction. There was no difference in plasma IL-10 (mean, 23.3 pg/ml vs. 48.1 pg/ml; median, 14.8 pg/ml vs. 21.4 pg/ml; NS) or plasma TNF-α levels (mean, 2.7 pg/ml vs. 21.3 pg/ml; median, 0 pg/ml vs. 0.8 pg/ml; NS) between FS and control patients. The significant cytokine plasma levels and the plasma IL-1RA/IL-1β ratio are presented in Fig. 1.

Figure 1.

Plasma levels of interleukin-6 (IL-6), interleukin-1 receptor antagonist (IL-1RA), interleukin-1β (IL-1β), and the IL-1Ra /IL-1β ratio of febrile seizure (FS) patients and control children. The median values are indicated by a line.

To find out the most significant predisposing factors for FSs, logistic regression analysis was used. We included in the univariate logistic regression analysis the plasma cytokines studied, age, sex, type of infection, fever at the time of hospitalization, and duration of fever before the blood sample. Significant associations were found between FS and high plasma IL-1RA levels (OR, 6.5; 95% CI, 1.9–22.0), FS and high plasma IL-1RA/IL-1β ratio (OR, 36.0; 95% CI, 4.5–289.9), and FS and high plasma IL-6 levels (OR, 4.2; 95% CI, 1.4–13.8). These significant variables were included in the multivariate logistic regression analysis. In this analysis, the high plasma IL-1RA/IL-1β ratio was the most significant factor connected to FSs (OR, 41.5; 95% CI, 4.9–352.8), but high plasma IL-6 levels were also significantly associated with FSs (OR, 5.3; 95% CI, 1.4–20.3).

CSF cytokines of FS children

CSF IL-6 levels were detectable in all 16 FS children studied, IL-1RA in nine of 12 FS children, IL-1β in one of 10 children, IL-10 in 10 of 16, and TNF-α was undetectable in the 15 children studied (Table 2).

Table 2.  Cerebrospinal fluid (CSF) cytokine levels (pg/ml) of febrile seizure (FS) children
 No. of FSMean ± SDMedian
Interleukin-61630.0 ± 67.29.4
Interleukin-1 receptor12296 ± 449170
Antagonist   
 Interleukin-1β100.2 ± 0.50
 Interkeukin-10169.5 ± 10.27.2
Tumor necrosis factor-α150 ± 00

DISCUSSION

In previous studies of cytokines and FSs, activation of the cytokine network has been shown, but the exact role of the cytokines in the pathogenesis of FSs is not known (6–9). Helminen and Vesikari (6) have shown increased IL-1 production of peripheral blood mononuclear cells of FS patients after stimulation with lipopolysaccharide (6). Tutuncuoglu et al. (7) reported increased plasma IL-1β and CSF TNF-α levels during the acute phase of FSs. Lahat et al. (8) reported no difference in serum and CSF IL-1β levels between FS and control children. In the present study, no significant difference was found in plasma IL-1β levels between FS and control children. Because IL-1β has a very short half-life, the levels could have been higher at the very beginning of the infection (4).

In the regulation of inflammation, balance between pro- and antiinflammatory cytokines may be more critical than a single cytokine concentration (4). In animal models, 100-fold molar excess of IL-1RA was needed to prevent the response to IL-1 (4). In this study, the plasma IL-1RA/IL-1β ratio was >700 in FS patients and >100 in controls. The production of IL-1RA is stimulated among other factors by IL-1β, and high IL-1RA levels could therefore indicate previous high IL-1β production (4).

In previous studies, CSF cytokine levels have been very low or undetectable in FS patients (7–9). In this study, elevated levels of IL-1RA, IL-6, and IL-10 were found in the CSF of FS patients. In previous studies, the authors speculated that the intrathecal presence of these cytokines could be used to differentiate FS and CNS infections (10). According to our results, detectable CSF cytokine levels of IL-6, IL-10, or IL-1RA cannot be used for this purpose. Elevated cytokine levels in plasma and detectable levels in CSF in FS patients could be a cause for seizure, or they could be produced as a consequence of seizure activity.

IL-1B allele 2 has been connected to increased in vitro production of IL-1β(11). In a previous study, we found increased frequency and carriage of IL-1B allele 2 in FS patients (3). We could not find any association between IL-1B polymorphisms and plasma IL-1β levels in FS children (unpublished data). Kanemoto et al. (12) reported that IL-1β allele 2 is more common in Japanese temporal lobe epilepsy patients with hippocampal sclerosis. Peltola et al. (13) found also increased frequency of IL-1B allele 2 in Finnish patients with localization-related epilepsy. Two other investigations failed to show this association (14,15).

The results of the present study support the hypothesis that the cytokine network is associated with FS. Whether these elevated cytokine levels have some role in the pathogenesis of FS or reflect seizure activity must be investigated further.

Acknowledgment: This work was supported by a grant from The Research Fund of Tampere University Hospital. We thank Mrs. Sinikka Repo-Koskinen for expert technical assistance and Heini Huhtala, MSc, for expert advice concerning statistical analysis of the data.

Ancillary