- Email: [email protected]
Febrile seizures associated with influenza A
Brain & Development 29 (2007) 30–38 www.elsevier.com/locate/braindev
Febrile seizures associated with influenza A Keita Hara a,*, Takuya Tanabe a, Tomoki Aomatsu a, Nao Inoue a, Hirohumi Tamaki a, Nami Okamoto a, Keisuke Okasora a, Takahiro Morimoto a, Hiroshi Tamai b a
Division of Pediatrics, Hirakata City Hospital, 2-14-1 Kinyahonmachi, Hirakata, Osaka, Japan b Department of Pediatrics, Osaka Medical College, Osaka, Japan Received 29 December 2005; received in revised form 18 May 2006; accepted 29 May 2006
Abstract To clarify the clinical impact of influenza A on the development of febrile seizures (FS), consecutive FS patients brought to our hospital between October 2003 and September 2004 were prospectively surveyed. Patients infected with influenza A (influenza A patients) and those uninfected with influenza (non-influenza patients) were compared with regard to clinical characteristics of FS. Influenza infection was determined by rapid antigen test and/or serologically. Associations of influenza A with atypical findings of FS, including partial seizures, prolonged seizures, multiple seizures during the same illness, and 30-min or longer prolonged postictal impairment of consciousness (PPIC), were analyzed by multiple logistic regression. A total of 215 patients (47 influenza A and 168 non-influenza patients) were enrolled in the study. Age was significantly higher in the influenza A group (39.85 ± 22.16 months vs. 27.51 ± 17.14 months, P < 0.001). Of 42 patients aged 48 months or older, which corresponded to the 80th percentile for age, 15 (35.7%) were influenza A patients, with a significantly higher incidence of such patients than in the subgroup of patients aged 47 months or younger (32/173, 18.5%) (P = 0.015). On multiple logistic regression analysis, influenza A was independently associated with PPIC (odds ratio: 4.44, 95% confidence interval: 1.52–12.95, P = 0.006), but not with other atypical findings. The positive association of influenza A with PPIC suggests that influenza may affect state of consciousness at the same time that it induces seizures with fever. Ó 2006 Elsevier B.V. All rights reserved. Keywords: Influenza A; Febrile seizures; Postictal impaired consciousness
1. Introduction Febrile seizures (FS) develop in infants or young children aged 6 months to 5 years in association with fever, and the neurological and developmental prognosis of such patients is good, unlike that of patients with central nervous infection or acute encephalopathy. Typical FS exhibits the following clinical features: (1) Seizure suddenly occurs with generalized tonic–clonic or clonic convulsions, (2) duration is less than 15 min, and (3) there is no repetition of seizure within 24 h. Cases *
Corresponding author. Tel.: +81 072 847 2821; fax: +81 072 847 6867. E-mail address: [email protected] (K. Hara). 0387-7604/$ - see front matter Ó 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.braindev.2006.05.010
exhibiting these features are defined as having simple febrile seizures [1]. However, many patients exhibit atypical symptoms such as partial seizures, prolonged seizures, and repeated seizures within a short period. These patients are defined as having complex FS, which is clinically distinguished from simple FS because of the risk of subsequent afebrile seizure [2,3], though there have been only a few studies of factors affecting these atypical findings [4,5]. Association of influenza virus infection with FS has been reported [6–10]. Although there are only a limited number of reports of specificity of symptoms, some have mentioned a relationship between influenza A and complex FS [8–10]. On the other hand, various cases of encephalitis/encephalopathy, from mild to severe,
K. Hara et al. / Brain & Development 29 (2007) 30–38
associated with influenza infection have been reported in the Japanese pediatric population [11–14]. In the severest cases, common clinical features are seizure of sudden onset accompanied by fever, and subsequent coma rapidly evolving to death. Influenza infection is thus an important cause of death in children in Japan. In addition to encephalitis and encephalopathy, influenza viral infection associated with transient disorders of consciousness such as mutism, delirium, and visual hallucinations with or without seizures have been reported [15–17]. Involvement of influenza-specific pathology in the development of these neurological symptoms has been discussed, but only in case reports or retrospective investigations. It is possible that seizures different from simple FS or accompanying disorder of consciousness occur at high incidence during influenza infection. In this study, we prospectively surveyed FS patients in a one-year period during 2003–2004 which featured an outbreak of influenza A in a geographic region of Japan, in order to clarify the characteristics of FS associated with influenza A as well as the association of influenza A with the atypical features accompanying FS including disorders of consciousness.
31
to cause FS in older children. Based on this hypothesis, no upper limit of age was set in this study, although patients aged less than 6 months were excluded. Patients with seizures with definable causes such as acute central nervous infection, acute encephalopathy, intracranial lesions, and acute abnormal metabolism and seizures associated with fever noted in patients with a past history of afebrile seizure were excluded from the FS group. If patients had atypical feature(s) required to be differentiated from symptoms of encephalitis/encephalopathy, such as prolonged seizure, prolonged postictal impaired consciousness, or multiple seizures, CSF examination, CT scan, and/or electroencephalogram were performed as appropriate. Since clear clinical differentiation between mild encephalopathy and complex FS is difficult in some cases, patients with prolonged impairment of consciousness for 12 h or longer following seizures were definitely excluded from the FS group. FS patients with underlying symptomatic etiologies such as obvious developmental retardation, neurological abnormality, and past history of central nervous system disorder were excluded from the study. 2.3. Prevalence of influenza in the 2003–2004 season
2. Patients and methods 2.1. Study population Hirakata City Hospital is located in the Kitakawachi area, Osaka, Japan, and is the sole year-round established general hospital in Hirakata City with a department of emergency service and child neurology, serving a pediatric population of about 60,000 (under 15 years of age). Since the hospital consistently accepts most children with emergent conditions associated with seizure disorders in Hirakata in addition to FS outpatients, most FS patients in this area visit our hospital. As subjects of this study, all FS patients seen in the emergency service and the child neurology department of our hospital between October 2003 and September 2004 were selected. 2.2. Definitions of and exclusion criteria for FS All seizure events including generalized and partial seizures accompanied by fever of 38 °C or higher were defined as FS. Seizure was considered partial when it exhibited any of following features: focal, unilateral, or dominantly unilateral motor symptoms, versive eye movements, forced head turning, and arrest of motion with staring or poor responsiveness. In clinical practice, we often experience school children past the age during which FS commonly occurs who have a single benign seizure associated with fever during the course of influenza, and hence hypothesize that influenza may tend
The endemic period of influenza infection in the 2003–2004 season in the Kitakawachi area, based on surveillance at established facilities, was between the 2nd and 11th weeks of 2004. In this season, a very few cases of influenza B infection were sporadically reported, though most cases reported were influenza A infection. We therefore surveyed the clinical characteristics of patients with FS associated with influenza A infection. 2.4. Determination of influenza virus infection A rapid influenza antigen test, ESPLINEÒ Influenza A & B (Fujirebio Inc., Tokyo, Japan: sensitivity: type A, 96.8%; type B, 87.9%; specificity: type A, 97.4%; type B, 99.4%, compared with viral isolation), was performed in all patients brought to our institution for possible FS. The test was performed on nasal wash specimens obtained by a standardized protocol using the manufacturer’s instructions. In most patients encountered during the period endemic for influenza, serological analysis for influenza A and B was also performed on paired serum samples using complement fixation assays. Patients positive for influenza A on rapid diagnosis test and/or with a significant rising titer to influenza A on paired sera were regarded as infected with influenza A (influenza A patients). To exclude false-negative patients, we excluded those encountered during the period endemic for influenza who were negative for influenza A and B on rapid antigen testing but not examined serologically, and
32
K. Hara et al. / Brain & Development 29 (2007) 30–38
the remaining patients were regarded as uninfected with influenza (non-influenza patients). 2.5. Data collection for determination of clinical characteristics of FS Data were collected prospectively by only one of the authors, who completed the data collection sheets for the following common survey items in order to avoid differences in method of investigation between investigators. Age, sex, past medical history, previous FS history, age at first FS, history of seizure disorders in first-degree relatives, time elapsed from initial significant fever to seizure occurrence, maximum temperature measured in the axilla immediately after seizure occurrence, and medications administered before seizure occurrence were determined, along with the following information about seizure manifestations: initial clinical change related to seizure occurrence, behavioral and motor symptoms of limbs and trunk and their laterality in the initial phase of seizure, oculomotor symptoms, evolution of these motor manifestations, whether the patient was cyanotic, and responsiveness to the observer. All of these items were obtained for all cases by carefully interviewing the observers. When anticonvulsants were used to inhibit seizures, the name, method of administration, and dosage were recorded. The duration of seizure, time from seizure termination to regaining of consciousness, and neurological findings associated with seizure were evaluated based on interviews with observers and the findings of examination. In addition, follow-up study of the incidence of repeated seizures during the same febrile illness was conducted. If patients had multiple seizures, the survey items detailed above were obtained as for the first seizure. Only the patients with simple FS confirmed to have fully recovered from postictal unconsciousness were allowed to return home after ambulatory care, but were subjected to reexamination the following day. Other patients were hospitalized. Patients not hospitalized were directed to promptly return for reexamination if seizure recurred or alteration of consciousness was noted. 2.6. Data analysis Characteristics of clinical data related to FS were compared between the influenza A and non-influenza patients. Next, among the clinical characteristics of FS, we focused on three well-known items determining complex FS: partial seizure, seizure longer than 15 min, and multiple seizures during the same illness. In addition, we paid special attention to postictal disorder of consciousness in order to examine the influence of influenza A on disorders of consciousness accompanying seizure. Each association of influenza A with these four special items was examined.
Statistical analysis was performed using Dr. SPSS II (Chicago, Illinois). Univariate analysis was performed to identify differences between patient groups with and without each special item, using the t-test, v2 test, and Mann–Whitney non-parametric test, as appropriate. Differences between the influenza A and non-influenza patients were also examined by the same procedure. Findings of P < 0.05 were considered significant. Variables in the univariate analysis with P < 0.2 were included in multiple forward stepwise logistic regression analyses. Influenza A infection was included among the parameters of multiple logistic regression analysis regardless of P value in the univariate analysis because it was the main item of this investigation.
3. Results 3.1. Characteristics of the patients A total of 239 (97.1%) of the 246 FS patients brought to our hospital during the study period had no symptomatic etiology (Fig. 1). Fifty-one (21.3%) and 179 (74.9%) of these patients met the inclusion criteria for the influenza A and non-influenza groups, respectively. Only 1 patient had influenza B. All the survey instruments were completed by, respectively, 47 (92.2%) and 168 (93.9%) of the patients of the two groups. Overall, 215 patients (122 males and 93 females, age: 31.1 ± 19.0 months) were enrolled in this study. Five of them were re-included as new patients due to recurrence of FS in the study period; one of them had a FS episode associated with influenza, while the others had two episodes, both without influenza. Of the patients, 130 were outpatients and 85 inpatients. The clinical characteristics of FS in the 215 patients are shown in Table 1. Thirty-six patients (16.7%) exhibited the features of partial seizure. The duration of seizure was less than 5 min in 189 patients (87.9%), and prolonged seizure persisting for 15 min or longer was noted in 15 patients (7.0%). The frequency of seizures in the same febrile illness was once in 184 patients (85.6%), twice in 29, and three and four times in 1 patient each. A total of 71 patients (33.0%) exhibited one or more features of complex FS. Postictal impairment of consciousness manifested by drowsiness, lethargy, or coma was prolonged for more than 30 min in 28 patients (13.0%); we termed this condition prolonged postictal impairment of consciousness (PPIC). Fifty-seven patients including 25 patients with PPIC had one or more of brain CT scanning, electroencephalogram, and CSF examination, to be differentiated from having other intracranial pathology than FS, and all of them were confirmed that they did not have.
K. Hara et al. / Brain & Development 29 (2007) 30–38
33
FS patients brought to our institute during the study period N=246
FS patients without symptomatic etiology N=239
excluded because of symptomatic etiology N=7 6 · mental retardation · past history of encephalopathy 1
influenza A patients N=51
excluded because of incompletion of survey instruments N=4
influenza A patients N=47
suspect of non-influenza patients N=187
excluded because of influenza B N=1
non-influenza patients N=179
excluded because of the possibility of false negative N=8
non-influenza patients N=168
excluded because of incompletion of survey instruments N=11
Fig. 1. Diagnostic flow chart for influenza A and non-influenza patients.
3.2. Comparison of clinical features between the influenza A and non-influenza patients Age (39.85 ± 22.16 months vs. 27.51 ± 17.14 months, P < 0.001) and age at first FS (28.26 ± 18.46 months vs. 21.98 ± 12.86 months, P < 0.001) were significantly higher in the influenza A group, but no significant differences were found between the groups in incidence of past history of FS or of history of seizure in first-degree relatives (Table 2). PPIC tended to occur more often in the influenza A patients than in the non-influenza patients (10/47 [21.3%] vs. 18/168 [10.7%], P = 0.057). No significant differences were found between the influenza A and non-influenza groups in partial seizures, prolonged seizures, or multiple seizures during the same illness. Fig. 2 shows the numbers of patients in 6-month age subgroups in the influenza A and non-influenza groups. A large age peak was present at 12–17 months, and the number of patients in the non-influenza patients decreased at the age of 24 months. In contrast, in the influenza A group, a peak was noted at 24–29 months, but a number of patients were present in the older age groups, and the number of patients 72 months of age or older was higher than in the non-influenza group. The cumulative ratio of influenza A patients to total
number of patients increased with age. Forty-two patients aged 48 months or older, which corresponded to the 80th percentile for age, were considered older FS patients, and the proportion of influenza A patients among the older FS patients was compared with that for the remaining FS patients (Fig. 3). Fifteen (35.7%) of the 42 older FS patients were infected with influenza A, a proportion significantly higher than that for patients aged 47 months or younger (32/173, 18.5%) (P = 0.015). No significant differences were found in the incidences of either family history of seizure (10/32 [31.3%] vs. 7/15 [46.7%], P = 0.305) or past history of FS (14/32 [43.8%] vs. 8/15 [53.3%], P = 0.540) between the two age subgroups in the influenza A group. There were only 3 patients (20.0%) with neither previous FS history nor family history of seizures in the older subgroup of influenza A patients. 3.3. Association of influenza A with atypical features of FS On univariate analyses of FS patients with and without each atypical feature, partial seizure was associated with age (41.00 ± 23.41 months vs. 28.03 ± 17.26 months, P = 0.00015), past history of FS (21/36
34
K. Hara et al. / Brain & Development 29 (2007) 30–38
Table 1 Clinical characteristics of 215 patients N = 215 Age (months; mean ± SD) Male Positive history of seizures in first-degree relatives Times of previous FS
30.20 ± 19.00 122 (56.7%) 94 (43.7%) None (129) One time (53) Two times (17) Three times (6) Four times (6) More than four times (4) 21.98 ± 12.90 8 [2.5, 20.2] 39.48 ± 0.62 79 (36.7%) 13 (6.0%) Generalized convulsion without partial feature (179) Partial seizure (36) Within 5 min (189) From more than 5 to less than 15 min (11) From 15 to less than 30 min (5) 30 min or more (10) 1 (184) 2 (29) 3 (1) 4 (1) Less than 30 min (187) From 30 to less than 60 min (7) From 60 to less than 120 min (5) 120 min or more (16)
Age at the first FS (months; mean ± SD) Time from fever to seizure occurrence (hours; median [25%, 75%]) Body temperature at seizure occurrence (°C; mean ± SD) Oral antihistamines before seizure occurrence Intravenous diazepam Seizure manifestations Seizure durations
Number of seizures during same febrile illness
Postictal impaired consciousness (Drowsiness, Lethargy, Coma)
FS indicates febrile seizures; SD, standard deviation; min, minute. Table 2 Clinical characteristics of FS episodes in the influenza A and non-influenza patients
Age (months; mean ± SD) Male (%) Positive history of seizures in first-degree relatives (%) Positive previous FS history (%) Age at first FS (months; mean ± SD) Time from fever to seizure occurrence (hours; median [25%, 75%]) Body temperature at seizure occurrence (°C; mean ± SD) Partial seizure (%) Prolonged seizure (%) Multiple seizures during the same febrile illness (%) PPIC (%)
Influenza A patients (N = 47)
Non-influenza patients (N = 168)
P Value
39.85 ± 22.16 29 (61.7) 17 (36.2) 22 (46.8) 28.26 ± 18.46 9 [5.5, 18] 39.60 ± 0.55 11 (23.4) 2 (4.3) 6 (12.8) 10 (21.3)
27.51 ± 17.14 93 (55.4) 77 (45.8) 64 (38.1) 21.98 ± 12.86 7 [2, 19] 39.51 ± 0.63 25 (14.9) 13 (7.7) 26 (15.5) 18 (10.7)
<0.001 0.438 0.238 0.281 <0.001 0.272 0.383 0.167 0.407 0.644 0.057
FS indicates febrile seizures; SD, standard deviation; PPIC, prolonged postictal impaired consciousness.
[58.3%] vs. 65/179 [36.3%], P = 0.014), and body temperature at seizure occurrence (39.29 ± 0.61 °C vs. 39.57 ± 0.61 °C, P = 0.032). Prolonged seizure was associated with partial seizure (7/15 [46.7%] vs. 29/200 [14.5%], P = 0.001). The occurrence of multiple seizures during the same febrile illness was associated with age (22.91 ± 17.80 months vs. 31.48 ± 18.96 months, P = 0.018). None of these three atypical features determining complex FS was associated with influenza A. PPIC was associated with body temperature at seizure occurrence (39.80 ± 0.65 °C vs. 39.46 ± 0.60 °C, P = 0.011), intravenous administration of diazepam
for sedation of seizure (9/28 [32.1%] vs. 4/187 [2.1%], odds ratio [OR]: 21.67; P < 0.0001), and prolonged seizure (10/28 [35.7%] vs. 5/187 [2.7%], OR: 20.22; P < 0.0001). PPIC was also associated with influenza A, though not with a high level of significance (10/28 [35.7%] vs. 37/187 [19.8%], OR: 2.25; P = 0.057). On multivariate analyses, categorization was performed using dummy variables with regard to age (<19 months, n = 72; 19 months 6, <33 months, n = 70; 33 months 6, n = 72), time elapsed from fever to seizure occurrence (<5 h, n = 73; 5 h 6, <15 h, n = 75; 15 h 6, n = 67), and body temperature at seizure
K. Hara et al. / Brain & Development 29 (2007) 30–38
Fig. 2. Numbers of patients in 6-month age groups. Number of noninfluenza patients peaked at 12–18 months of age, with a distinct decrease with aging, whereas influenza A patients were constant in number across age groups after a small peak at 24–30 months. The cumulative percentage of influenza A patients was only 10% in the patient group aged under 24 months, but when patients aged over 24 months were included, a trend toward increase was noted with the addition of higher-aged groups, and cumulative percentage finally reached 21.9% for the group of all patients.
occurrence (<39.0 °C, n = 32; 39.0 °C 6, <39.5 °C, n = 70; 39.5 °C 6, <40.0 °C, n = 51; 40.0 °C 6, n = 62), and incorporated in the logistic model. In addition to prolonged seizure (OR: 25.36; 95% CI: 4.23–151.89; P < 0.001) and intravenous diazepam (OR: 8.32; 95% CI: 1.53–45.31; P = 0.014), influenza A (OR: 4.44; 95% confidence interval [CI]: 1.52–12.95; P = 0.006) was independently associated with high risk of PPIC (Table 3). On the other hand, it was not associated with partial seizures, prolonged seizures, or multiple seizures. Age of 33 months or older was significantly associated with a high risk of partial seizure, compared to age below 19 months (OR: 3.82; 95% CI: 1.47–9.92; P = 0.006). The risk of partial seizure decreased with increase in body temperature at seizure occurrence, and body temperature between 39.5 °C and 40.0 °C was significantly associated with a low risk of partial seizure, compared to body temperature lower than 39.0 °C (OR: 0.22, 95% CI: 0.06–0.85, P = 0.027). Partial seizure was strongly associated with a risk of prolonged seizure (OR: 5.16, 95% CI: 1.73– 15.32, P = 0.003). Age below 19 months was significantly associated with a high risk of multiple seizures, compared to age of 33 months or older (OR: 5.06, 95% CI: 1.78–14.36, P = 0.002).
4. Discussion In our prospective one-year study, patients with influenza A comprised 21.3 percent of all FS patients,
35
a proportion close to that reported by Chiu et al. [8], 21.7%, in 1998. On comparison of the influenza A and non-influenza groups, age was significantly higher in the former, while there were no significant differences in past history of FS or family history of seizure disorder between the two groups. The proportion of influenza A patients was significantly higher in the older age subgroup of FS patients than in the remaining FS patients, and most of the older patients with influenza A had a family history of seizures or past FS history. Our findings suggest that influenza A may be a major cause of FS, especially in older children, but that the pathogenesis of FS in older children may mainly involve the susceptibility to seizures they originally have. In the present study, no association was found between influenza A and partial seizures. According to the seizure classification using the International League Against Epilepsy criteria [18], partial seizures are those in which the first clinical and electroencephalographic changes indicate initial activation of a system of neurons limited to part of one cerebral hemisphere. The presence of underlying disorders such as congenital focal cortical lesions or acquired impairment of cerebral parenchyma in consequence of various insults is assumed to be a cause of partial seizures. There has, in fact, been a report of association of prior neurological abnormality with partial seizure [19]. Berg et al. [4] reported that 16.1% of 428 patients with first FS exhibited partial seizure in a prospective cohort study, an incidence similar to that in our patients (16.7%), though no association was noted between neurological abnormality and partial seizure in their study. Since we excluded patients with past history of central nervous system disorder, neurological abnormality, or mental retardation, patients with an underlying symptomatic etiology were excluded. These findings indicate that many cases of partial seizure are unrelated neurological abnormality. Susceptibility to FS is strongly influenced by genetic factors [20]. In a prospective family study, 10% of siblings and 10% of offspring of FS probands also had FS [21]. There are many cases of idiopathic partial epilepsy associated with genetic predisposition at a higher age than that of peak incidence of FS, and subgroups with a high incidence of past FS have been reported [22–24]. Since 58.3% of our patients with partial seizures had a family history of seizure, there may be a genetic predisposition to development of this type of seizure. The incidence of multiple seizures was 12.8% in patients with FS associated with influenza A, and not significantly different from that in non-influenza patients, nor was an association found between influenza A and multiple seizures on multivariate analysis. In the study reported by Chiu et al. [8], on the other hand, the incidences of multiple seizures in FS inpatients with influenza A were 22.2% in 1997 and 31.5% in 1998, suggesting association of influenza A with multiple seizures,
36
K. Hara et al. / Brain & Development 29 (2007) 30–38
Age 47 months (N = 173)
Age
48 months (N = 42) Influenza A Non-influenza
N=15 (35.7%)
N=32 (18.5%)
· Family history of seizures (+) · Previous FS history (+) · Family history of seizures (-)
· Family history of seizures (+) 10 (31.3%) · Previous FS history (+) 14 (43.8%) · Family history of seizures (-) 15 (46.9%) and previous FS history (-)
and previous FS history (-)
7 (46.7%) 8 (53.3%) 3 (20.0%)
P = 0.015 Fig. 3. Older FS patients, aged 48 months or older (N = 42), were compared with the other FS patients aged 47 months or younger (N = 173) for proportion of influenza A patients. The proportion in the older age group (15/42, 35.7%) was significantly higher than that in the remaining patients (32/173, 18.5%) (P = 0.015), but the influenza A patients in the older age group exhibited high incidence of family history of seizure and past history of FS, as did the remaining patients. Note that only 3 patients had neither previous FS history nor family history of seizures among the older-aged influenza A patients.
though the statistical significance of these findings was not presented. Selection of patients may have been biased in their study, since they retrospectively collected only inpatients, which may have resulted in a higher incidence of multiple seizures than in our study. Low age was independently associated with multiple seizures in our study. Suga et al. [25] clinically investigated FS associated with primary HHV-6 infection, and found that occurrence at a low age and a high incidence of clustering seizures were characteristics of this condition. It is possible that multiple seizures are an infective-pathogen-dependent characteristic of infectious diseases that occur at high incidence in young infants. Only a few studies have mentioned postictal impairment of consciousness in FS. Yamamoto [26] emphasized the characteristics of postictal impairment of consciousness with convulsive manifestations in FS patients, based on corresponding electroencephalographic findings of non-ictal patterns. In a study of 213 consecutive cases of FS, Okumura et al. [27] found the duration of postictal unconsciousness to be less than 30 min in 93% of patients, and found associations of non-generalized seizure, seizures P5 min in duration,
and use of intravenous diazepam with prolonged unconsciousness. In our one-year study, postictal impairment of consciousness manifested by drowsiness, lethargy, or coma persisted for 30 min or longer in 13% of patients. It is reasonable for prolonged seizure or use of intravenous diazepam to be associated with this relatively rare condition, considering their relatively direct effects on the central nervous system. However, ours is the first report in which strong association of influenza A with PPIC is described. Influenza infection often progresses to encephalitis/encephalopathy, particularly in the pediatric population [11–14], and influenza-specific involvement of the central nervous system is suspected in such patients. Although our patients with PPIC were distinctly different from those with encephalitis or encephalopathy in the benignancy of their clinical course, the findings of our study also suggest the neurovirulence of influenza virus. There have recently been reports suggesting the involvement of various inflammatory cytokines, such as interleukin six, in the pathology of influenza-associated FS and encephalitis/encephalopathy [28,29]. On the other hand, some clinical and animal studies have demonstrated relationships between
K. Hara et al. / Brain & Development 29 (2007) 30–38
37
Table 3 Logistic regression analysis: independent risk factors for atypical features of FS; partial seizure, prolonged seizure, multiple seizures during the same febrile illness, and PPIC Atypical features
Risk factors
Partial seizure
Age (months) <19 19 6, <33 33 6 Body temperature at seizure occurrence (°C) <39.0 39.0 6, <39.5 39.5 6, <40.0 40.0 6
1.00 1.26 3.82
0.43–3.71 1.47–9.92
0.669 0.006
1.00 0.85 0.22 0.40
0.31–2.36 0.06–0.85 0.13–1.25
0.757 0.027 0.115
Prolonged seizure
Partial seizure
5.16
1.73–15.32
0.003
Multiple seizures
Age (months) 33 6 19 6, <33 <19
1.00 1.45 5.06
0.45–4.93 1.78–14.36
0.515 0.002
4.44 25.36 8.32
1.52–12.95 4.23–151.89 1.53–45.31
PPIC
Influenza A Prolonged seizure Intravenous diazepam
OR
95% CI
P
0.006 <0.001 0.014
FS indicates febrile seizures; OR, odds ratio; Cl, confidence interval; PPIC, prolonged postictal impaired consciousness.
inflammatory cytokines and sleep regulation [30,31]. Unbalanced endogenous cytokine levels may play roles in the pathogenesis of persistent postictal impairment of consciousness, particularly in individuals with influenza virus infection. That many cases of influenza-associated encephalitis/encephalopathy have been reported from Japan compared with other countries indicates the susceptibility of Japanese children to this specific pathological condition. By the same token, PPIC associated with influenza may be a specifically Japanese problem. These speculations need to be confirmed in further studies. In conclusion, this study provides new information on the independent association of influenza A infection with the development of disorders of consciousness subsequent to febrile seizure.
References [1] Camfield P, Camfield C, Kurlemann G. Febrile seizures. Epileptic syndromes in infancy, childhood, and adolescence. 3rd ed. London: John Libbey & Co Ltd; 2002, p. 145–52. [2] Tsai ML, Hung KL. Risk factors for subsequent epilepsy after febrile convulsions. J Formos Med Assoc 1995;94:327–31. [3] Sapir D, Leitner Y, Harel S, Kraumer U. Unprovoked seizures after complex febrile convulsions. Brain Dev 2000;22:484–6. [4] Berg AT, Shinnar S. Complex febrile seizures. Epilepsia 1996;37:126–33. [5] Maytal J, Shinnar S. Febrile status epilepticus. Pediatrics 1990;86:611–6. [6] Brocklebank JT, Court SD, McQuillin J, Gardner PS. InfluenzaA infection in children. Lancet 1972;2:497–500.
[7] Price DA, Postlethwaite RJ, Longson M. Influenzavirus A2 infections presenting with febril convulsions and gastrointestinal symptoms in young children. Clin Pediatr 1976;15:361–7. [8] Chiu SS, Tse CY, Lau YL, Peiris M. Influenza A infection is an important cause of febrile seizures. Pediatrics 2001;108:E63. [9] van Zeijl JH, Mullaart RA, Borm GF, Galama JM. Recurrence of febrile seizures in the respiratory season is associated with influenza A. J Pediatr 2004;145:800–5. [10] Stricker T, Sennhauser FH. Complex febrile seizures associated with influenza A. Pediatr Infect Dis J 2004;23(5):480. [11] Yoshikawa H, Yamazaki S, Watanabe T, Abe T. Study of influenza-associated encephalitis/encephalopathy in children during the 1997 to 2001 influenza seasons. J Child Neurol 2001;16:885–90. [12] Morishima T, Togashi T, Yokota S, Okuno Y, Miyazaki C, Tashiro M, et al. Encephalitis and encephalopathy associated with an influenza epidemic in Japan. Clin Infect Dis 2002;35:512–7. [13] Fujimoto S, Kobayashi M, Uemura O, Iwasa M, Ando T, Katoh T, et al. PCR on cerebrospinal fluid to show influenza-associated acute encephalopathy or encephalitis. Lancet 1998;352:873–5. [14] Takahashi M, Yamada T, Nakashita Y, Saikusa H, Deguchi M, Kida H, et al. Influenza virus-induced encephalopathy: clinicopathologic study of an autopsied case. Pediatr Int 2000;42:204–14. [15] Maricich SM, Neul JL, Lotze TE, Cazacu AC, Uyeki TM, Demmler GJ, et al. Neurologic complications associated with influenza A in children during the 2003-2004 influenza season in Houston, Texas. Pediatrics 2004;114:e626–33. [16] Huang YC, Lin TY, Wu SL, Tsao KC. Influenza A-associated central nervous system dysfunction in children presenting as transient visual hallucination. Pediatr Infect Dis J 2003;22:366–8. [17] Okumura A, Nakano T, Fukumoto Y, Higuchi K, Kamiya H, Watanabe K, et al. Delirious behavior in children with influenza: its clinical features and EEG findings. Brain Dev 2005;27:271–4. [18] International League Against Epilepsy. Proposal for revised clinical and electroencephalographic classification of epileptic seizures. Epilepsia 1981;22:489–501.
38
K. Hara et al. / Brain & Development 29 (2007) 30–38
[19] Shinnar S, Berg AT, Moshe SL, Petix M, Maytal J, Kang H, et al. Risk of seizure recurrence following a first unprovoked seizure in childhood: a prospective study. Pediatrics 1990;85:1076–85. [20] Kugler SL, Johnson WG. Genetics of the febrile seizure susceptibility trait. Brain Dev 1998;20:265–74. [21] Doose H, Maurer A. Seizure risk in offspring of individuals with a history of febrile convulsions. Eur J Pediatr 1997;156:476–81. [22] Casetta I, Monetti VC, Malagu S, Paolino E, Govoni V, Fainardi E, et al. Risk factors for cryptogenic and idiopathic partial epilepsy: a community-based case-control study in Copparo, Italy. Neuroepidemiology 2002;21:251–4. [23] Ward N, Evanson J, Cockerell OC. Idiopathic familial temporal lobe epilepsy with febrile convulsions. Seizure 2002;11:16–9. [24] Doose H, Petersen B, Neubauer BA. Occipital sharp waves in idiopathic partial epilepsies–clinical and genetic aspects. Epilepsy Res 2002;48:121–30. [25] Suga S, Suzuki K, Ihira M, Yoshikawa T, Kajita Y, Ozaki T, et al. Clinical characteristics of febrile convulsions during primary HHV-6 infection. Arch Dis Child 2000;82:62–6.
[26] Yamamoto N. Prolonged nonepileptic twilight state with convulsive manifestations after febrile convulsions: a clinical and electroencephalographic study. Epilepsia 1996;37:31–5. [27] Okumura A, Uemura N, Suzuki M, Itomi K, Watanabe K. Unconsciousness and delirious behavior in children with febrile seizures. Pediatr Neurol 2004;30:316–9. [28] Masuyama T, Matsuo M, Ichimaru T, Ishii K, Tsuchiya K, Hamasaki Y. Possible contribution of interferon-a to febrile seizures in influenza. Pediatr Neurol 2002;27:289–92. [29] Aiba H, Mochizuki M, Kimura M, Hojo H. Predictive value of serum interleukin-6 level in influenza virus-associated encephalopathy. Neurology 2001;57:295–9. [30] Vgontzas AN, Papanicolaou DA, Bixler EO, Kales A, Tyson K, Chrousos GP. Elevation of plasma cytokines in disorders of excessive daytime sleepiness: role of sleep disturbance and obesity. J Clin Endocrinol Metab 1997;82:1313–6. [31] Hogan D, Morrow JD, Smith EM, Opp MR. Interleukin-6 alters sleep of rats. J Neuroimmunol 2003;137:59–66.
Febrile seizures associated with influenza A Keita Hara a,*, Takuya Tanabe a, Tomoki Aomatsu a, Nao Inoue a, Hirohumi Tamaki a, Nami Okamoto a, Keisuke Okasora a, Takahiro Morimoto a, Hiroshi Tamai b a
Division of Pediatrics, Hirakata City Hospital, 2-14-1 Kinyahonmachi, Hirakata, Osaka, Japan b Department of Pediatrics, Osaka Medical College, Osaka, Japan Received 29 December 2005; received in revised form 18 May 2006; accepted 29 May 2006
Abstract To clarify the clinical impact of influenza A on the development of febrile seizures (FS), consecutive FS patients brought to our hospital between October 2003 and September 2004 were prospectively surveyed. Patients infected with influenza A (influenza A patients) and those uninfected with influenza (non-influenza patients) were compared with regard to clinical characteristics of FS. Influenza infection was determined by rapid antigen test and/or serologically. Associations of influenza A with atypical findings of FS, including partial seizures, prolonged seizures, multiple seizures during the same illness, and 30-min or longer prolonged postictal impairment of consciousness (PPIC), were analyzed by multiple logistic regression. A total of 215 patients (47 influenza A and 168 non-influenza patients) were enrolled in the study. Age was significantly higher in the influenza A group (39.85 ± 22.16 months vs. 27.51 ± 17.14 months, P < 0.001). Of 42 patients aged 48 months or older, which corresponded to the 80th percentile for age, 15 (35.7%) were influenza A patients, with a significantly higher incidence of such patients than in the subgroup of patients aged 47 months or younger (32/173, 18.5%) (P = 0.015). On multiple logistic regression analysis, influenza A was independently associated with PPIC (odds ratio: 4.44, 95% confidence interval: 1.52–12.95, P = 0.006), but not with other atypical findings. The positive association of influenza A with PPIC suggests that influenza may affect state of consciousness at the same time that it induces seizures with fever. Ó 2006 Elsevier B.V. All rights reserved. Keywords: Influenza A; Febrile seizures; Postictal impaired consciousness
1. Introduction Febrile seizures (FS) develop in infants or young children aged 6 months to 5 years in association with fever, and the neurological and developmental prognosis of such patients is good, unlike that of patients with central nervous infection or acute encephalopathy. Typical FS exhibits the following clinical features: (1) Seizure suddenly occurs with generalized tonic–clonic or clonic convulsions, (2) duration is less than 15 min, and (3) there is no repetition of seizure within 24 h. Cases *
Corresponding author. Tel.: +81 072 847 2821; fax: +81 072 847 6867. E-mail address: [email protected] (K. Hara). 0387-7604/$ - see front matter Ó 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.braindev.2006.05.010
exhibiting these features are defined as having simple febrile seizures [1]. However, many patients exhibit atypical symptoms such as partial seizures, prolonged seizures, and repeated seizures within a short period. These patients are defined as having complex FS, which is clinically distinguished from simple FS because of the risk of subsequent afebrile seizure [2,3], though there have been only a few studies of factors affecting these atypical findings [4,5]. Association of influenza virus infection with FS has been reported [6–10]. Although there are only a limited number of reports of specificity of symptoms, some have mentioned a relationship between influenza A and complex FS [8–10]. On the other hand, various cases of encephalitis/encephalopathy, from mild to severe,
K. Hara et al. / Brain & Development 29 (2007) 30–38
associated with influenza infection have been reported in the Japanese pediatric population [11–14]. In the severest cases, common clinical features are seizure of sudden onset accompanied by fever, and subsequent coma rapidly evolving to death. Influenza infection is thus an important cause of death in children in Japan. In addition to encephalitis and encephalopathy, influenza viral infection associated with transient disorders of consciousness such as mutism, delirium, and visual hallucinations with or without seizures have been reported [15–17]. Involvement of influenza-specific pathology in the development of these neurological symptoms has been discussed, but only in case reports or retrospective investigations. It is possible that seizures different from simple FS or accompanying disorder of consciousness occur at high incidence during influenza infection. In this study, we prospectively surveyed FS patients in a one-year period during 2003–2004 which featured an outbreak of influenza A in a geographic region of Japan, in order to clarify the characteristics of FS associated with influenza A as well as the association of influenza A with the atypical features accompanying FS including disorders of consciousness.
31
to cause FS in older children. Based on this hypothesis, no upper limit of age was set in this study, although patients aged less than 6 months were excluded. Patients with seizures with definable causes such as acute central nervous infection, acute encephalopathy, intracranial lesions, and acute abnormal metabolism and seizures associated with fever noted in patients with a past history of afebrile seizure were excluded from the FS group. If patients had atypical feature(s) required to be differentiated from symptoms of encephalitis/encephalopathy, such as prolonged seizure, prolonged postictal impaired consciousness, or multiple seizures, CSF examination, CT scan, and/or electroencephalogram were performed as appropriate. Since clear clinical differentiation between mild encephalopathy and complex FS is difficult in some cases, patients with prolonged impairment of consciousness for 12 h or longer following seizures were definitely excluded from the FS group. FS patients with underlying symptomatic etiologies such as obvious developmental retardation, neurological abnormality, and past history of central nervous system disorder were excluded from the study. 2.3. Prevalence of influenza in the 2003–2004 season
2. Patients and methods 2.1. Study population Hirakata City Hospital is located in the Kitakawachi area, Osaka, Japan, and is the sole year-round established general hospital in Hirakata City with a department of emergency service and child neurology, serving a pediatric population of about 60,000 (under 15 years of age). Since the hospital consistently accepts most children with emergent conditions associated with seizure disorders in Hirakata in addition to FS outpatients, most FS patients in this area visit our hospital. As subjects of this study, all FS patients seen in the emergency service and the child neurology department of our hospital between October 2003 and September 2004 were selected. 2.2. Definitions of and exclusion criteria for FS All seizure events including generalized and partial seizures accompanied by fever of 38 °C or higher were defined as FS. Seizure was considered partial when it exhibited any of following features: focal, unilateral, or dominantly unilateral motor symptoms, versive eye movements, forced head turning, and arrest of motion with staring or poor responsiveness. In clinical practice, we often experience school children past the age during which FS commonly occurs who have a single benign seizure associated with fever during the course of influenza, and hence hypothesize that influenza may tend
The endemic period of influenza infection in the 2003–2004 season in the Kitakawachi area, based on surveillance at established facilities, was between the 2nd and 11th weeks of 2004. In this season, a very few cases of influenza B infection were sporadically reported, though most cases reported were influenza A infection. We therefore surveyed the clinical characteristics of patients with FS associated with influenza A infection. 2.4. Determination of influenza virus infection A rapid influenza antigen test, ESPLINEÒ Influenza A & B (Fujirebio Inc., Tokyo, Japan: sensitivity: type A, 96.8%; type B, 87.9%; specificity: type A, 97.4%; type B, 99.4%, compared with viral isolation), was performed in all patients brought to our institution for possible FS. The test was performed on nasal wash specimens obtained by a standardized protocol using the manufacturer’s instructions. In most patients encountered during the period endemic for influenza, serological analysis for influenza A and B was also performed on paired serum samples using complement fixation assays. Patients positive for influenza A on rapid diagnosis test and/or with a significant rising titer to influenza A on paired sera were regarded as infected with influenza A (influenza A patients). To exclude false-negative patients, we excluded those encountered during the period endemic for influenza who were negative for influenza A and B on rapid antigen testing but not examined serologically, and
32
K. Hara et al. / Brain & Development 29 (2007) 30–38
the remaining patients were regarded as uninfected with influenza (non-influenza patients). 2.5. Data collection for determination of clinical characteristics of FS Data were collected prospectively by only one of the authors, who completed the data collection sheets for the following common survey items in order to avoid differences in method of investigation between investigators. Age, sex, past medical history, previous FS history, age at first FS, history of seizure disorders in first-degree relatives, time elapsed from initial significant fever to seizure occurrence, maximum temperature measured in the axilla immediately after seizure occurrence, and medications administered before seizure occurrence were determined, along with the following information about seizure manifestations: initial clinical change related to seizure occurrence, behavioral and motor symptoms of limbs and trunk and their laterality in the initial phase of seizure, oculomotor symptoms, evolution of these motor manifestations, whether the patient was cyanotic, and responsiveness to the observer. All of these items were obtained for all cases by carefully interviewing the observers. When anticonvulsants were used to inhibit seizures, the name, method of administration, and dosage were recorded. The duration of seizure, time from seizure termination to regaining of consciousness, and neurological findings associated with seizure were evaluated based on interviews with observers and the findings of examination. In addition, follow-up study of the incidence of repeated seizures during the same febrile illness was conducted. If patients had multiple seizures, the survey items detailed above were obtained as for the first seizure. Only the patients with simple FS confirmed to have fully recovered from postictal unconsciousness were allowed to return home after ambulatory care, but were subjected to reexamination the following day. Other patients were hospitalized. Patients not hospitalized were directed to promptly return for reexamination if seizure recurred or alteration of consciousness was noted. 2.6. Data analysis Characteristics of clinical data related to FS were compared between the influenza A and non-influenza patients. Next, among the clinical characteristics of FS, we focused on three well-known items determining complex FS: partial seizure, seizure longer than 15 min, and multiple seizures during the same illness. In addition, we paid special attention to postictal disorder of consciousness in order to examine the influence of influenza A on disorders of consciousness accompanying seizure. Each association of influenza A with these four special items was examined.
Statistical analysis was performed using Dr. SPSS II (Chicago, Illinois). Univariate analysis was performed to identify differences between patient groups with and without each special item, using the t-test, v2 test, and Mann–Whitney non-parametric test, as appropriate. Differences between the influenza A and non-influenza patients were also examined by the same procedure. Findings of P < 0.05 were considered significant. Variables in the univariate analysis with P < 0.2 were included in multiple forward stepwise logistic regression analyses. Influenza A infection was included among the parameters of multiple logistic regression analysis regardless of P value in the univariate analysis because it was the main item of this investigation.
3. Results 3.1. Characteristics of the patients A total of 239 (97.1%) of the 246 FS patients brought to our hospital during the study period had no symptomatic etiology (Fig. 1). Fifty-one (21.3%) and 179 (74.9%) of these patients met the inclusion criteria for the influenza A and non-influenza groups, respectively. Only 1 patient had influenza B. All the survey instruments were completed by, respectively, 47 (92.2%) and 168 (93.9%) of the patients of the two groups. Overall, 215 patients (122 males and 93 females, age: 31.1 ± 19.0 months) were enrolled in this study. Five of them were re-included as new patients due to recurrence of FS in the study period; one of them had a FS episode associated with influenza, while the others had two episodes, both without influenza. Of the patients, 130 were outpatients and 85 inpatients. The clinical characteristics of FS in the 215 patients are shown in Table 1. Thirty-six patients (16.7%) exhibited the features of partial seizure. The duration of seizure was less than 5 min in 189 patients (87.9%), and prolonged seizure persisting for 15 min or longer was noted in 15 patients (7.0%). The frequency of seizures in the same febrile illness was once in 184 patients (85.6%), twice in 29, and three and four times in 1 patient each. A total of 71 patients (33.0%) exhibited one or more features of complex FS. Postictal impairment of consciousness manifested by drowsiness, lethargy, or coma was prolonged for more than 30 min in 28 patients (13.0%); we termed this condition prolonged postictal impairment of consciousness (PPIC). Fifty-seven patients including 25 patients with PPIC had one or more of brain CT scanning, electroencephalogram, and CSF examination, to be differentiated from having other intracranial pathology than FS, and all of them were confirmed that they did not have.
K. Hara et al. / Brain & Development 29 (2007) 30–38
33
FS patients brought to our institute during the study period N=246
FS patients without symptomatic etiology N=239
excluded because of symptomatic etiology N=7 6 · mental retardation · past history of encephalopathy 1
influenza A patients N=51
excluded because of incompletion of survey instruments N=4
influenza A patients N=47
suspect of non-influenza patients N=187
excluded because of influenza B N=1
non-influenza patients N=179
excluded because of the possibility of false negative N=8
non-influenza patients N=168
excluded because of incompletion of survey instruments N=11
Fig. 1. Diagnostic flow chart for influenza A and non-influenza patients.
3.2. Comparison of clinical features between the influenza A and non-influenza patients Age (39.85 ± 22.16 months vs. 27.51 ± 17.14 months, P < 0.001) and age at first FS (28.26 ± 18.46 months vs. 21.98 ± 12.86 months, P < 0.001) were significantly higher in the influenza A group, but no significant differences were found between the groups in incidence of past history of FS or of history of seizure in first-degree relatives (Table 2). PPIC tended to occur more often in the influenza A patients than in the non-influenza patients (10/47 [21.3%] vs. 18/168 [10.7%], P = 0.057). No significant differences were found between the influenza A and non-influenza groups in partial seizures, prolonged seizures, or multiple seizures during the same illness. Fig. 2 shows the numbers of patients in 6-month age subgroups in the influenza A and non-influenza groups. A large age peak was present at 12–17 months, and the number of patients in the non-influenza patients decreased at the age of 24 months. In contrast, in the influenza A group, a peak was noted at 24–29 months, but a number of patients were present in the older age groups, and the number of patients 72 months of age or older was higher than in the non-influenza group. The cumulative ratio of influenza A patients to total
number of patients increased with age. Forty-two patients aged 48 months or older, which corresponded to the 80th percentile for age, were considered older FS patients, and the proportion of influenza A patients among the older FS patients was compared with that for the remaining FS patients (Fig. 3). Fifteen (35.7%) of the 42 older FS patients were infected with influenza A, a proportion significantly higher than that for patients aged 47 months or younger (32/173, 18.5%) (P = 0.015). No significant differences were found in the incidences of either family history of seizure (10/32 [31.3%] vs. 7/15 [46.7%], P = 0.305) or past history of FS (14/32 [43.8%] vs. 8/15 [53.3%], P = 0.540) between the two age subgroups in the influenza A group. There were only 3 patients (20.0%) with neither previous FS history nor family history of seizures in the older subgroup of influenza A patients. 3.3. Association of influenza A with atypical features of FS On univariate analyses of FS patients with and without each atypical feature, partial seizure was associated with age (41.00 ± 23.41 months vs. 28.03 ± 17.26 months, P = 0.00015), past history of FS (21/36
34
K. Hara et al. / Brain & Development 29 (2007) 30–38
Table 1 Clinical characteristics of 215 patients N = 215 Age (months; mean ± SD) Male Positive history of seizures in first-degree relatives Times of previous FS
30.20 ± 19.00 122 (56.7%) 94 (43.7%) None (129) One time (53) Two times (17) Three times (6) Four times (6) More than four times (4) 21.98 ± 12.90 8 [2.5, 20.2] 39.48 ± 0.62 79 (36.7%) 13 (6.0%) Generalized convulsion without partial feature (179) Partial seizure (36) Within 5 min (189) From more than 5 to less than 15 min (11) From 15 to less than 30 min (5) 30 min or more (10) 1 (184) 2 (29) 3 (1) 4 (1) Less than 30 min (187) From 30 to less than 60 min (7) From 60 to less than 120 min (5) 120 min or more (16)
Age at the first FS (months; mean ± SD) Time from fever to seizure occurrence (hours; median [25%, 75%]) Body temperature at seizure occurrence (°C; mean ± SD) Oral antihistamines before seizure occurrence Intravenous diazepam Seizure manifestations Seizure durations
Number of seizures during same febrile illness
Postictal impaired consciousness (Drowsiness, Lethargy, Coma)
FS indicates febrile seizures; SD, standard deviation; min, minute. Table 2 Clinical characteristics of FS episodes in the influenza A and non-influenza patients
Age (months; mean ± SD) Male (%) Positive history of seizures in first-degree relatives (%) Positive previous FS history (%) Age at first FS (months; mean ± SD) Time from fever to seizure occurrence (hours; median [25%, 75%]) Body temperature at seizure occurrence (°C; mean ± SD) Partial seizure (%) Prolonged seizure (%) Multiple seizures during the same febrile illness (%) PPIC (%)
Influenza A patients (N = 47)
Non-influenza patients (N = 168)
P Value
39.85 ± 22.16 29 (61.7) 17 (36.2) 22 (46.8) 28.26 ± 18.46 9 [5.5, 18] 39.60 ± 0.55 11 (23.4) 2 (4.3) 6 (12.8) 10 (21.3)
27.51 ± 17.14 93 (55.4) 77 (45.8) 64 (38.1) 21.98 ± 12.86 7 [2, 19] 39.51 ± 0.63 25 (14.9) 13 (7.7) 26 (15.5) 18 (10.7)
<0.001 0.438 0.238 0.281 <0.001 0.272 0.383 0.167 0.407 0.644 0.057
FS indicates febrile seizures; SD, standard deviation; PPIC, prolonged postictal impaired consciousness.
[58.3%] vs. 65/179 [36.3%], P = 0.014), and body temperature at seizure occurrence (39.29 ± 0.61 °C vs. 39.57 ± 0.61 °C, P = 0.032). Prolonged seizure was associated with partial seizure (7/15 [46.7%] vs. 29/200 [14.5%], P = 0.001). The occurrence of multiple seizures during the same febrile illness was associated with age (22.91 ± 17.80 months vs. 31.48 ± 18.96 months, P = 0.018). None of these three atypical features determining complex FS was associated with influenza A. PPIC was associated with body temperature at seizure occurrence (39.80 ± 0.65 °C vs. 39.46 ± 0.60 °C, P = 0.011), intravenous administration of diazepam
for sedation of seizure (9/28 [32.1%] vs. 4/187 [2.1%], odds ratio [OR]: 21.67; P < 0.0001), and prolonged seizure (10/28 [35.7%] vs. 5/187 [2.7%], OR: 20.22; P < 0.0001). PPIC was also associated with influenza A, though not with a high level of significance (10/28 [35.7%] vs. 37/187 [19.8%], OR: 2.25; P = 0.057). On multivariate analyses, categorization was performed using dummy variables with regard to age (<19 months, n = 72; 19 months 6, <33 months, n = 70; 33 months 6, n = 72), time elapsed from fever to seizure occurrence (<5 h, n = 73; 5 h 6, <15 h, n = 75; 15 h 6, n = 67), and body temperature at seizure
K. Hara et al. / Brain & Development 29 (2007) 30–38
Fig. 2. Numbers of patients in 6-month age groups. Number of noninfluenza patients peaked at 12–18 months of age, with a distinct decrease with aging, whereas influenza A patients were constant in number across age groups after a small peak at 24–30 months. The cumulative percentage of influenza A patients was only 10% in the patient group aged under 24 months, but when patients aged over 24 months were included, a trend toward increase was noted with the addition of higher-aged groups, and cumulative percentage finally reached 21.9% for the group of all patients.
occurrence (<39.0 °C, n = 32; 39.0 °C 6, <39.5 °C, n = 70; 39.5 °C 6, <40.0 °C, n = 51; 40.0 °C 6, n = 62), and incorporated in the logistic model. In addition to prolonged seizure (OR: 25.36; 95% CI: 4.23–151.89; P < 0.001) and intravenous diazepam (OR: 8.32; 95% CI: 1.53–45.31; P = 0.014), influenza A (OR: 4.44; 95% confidence interval [CI]: 1.52–12.95; P = 0.006) was independently associated with high risk of PPIC (Table 3). On the other hand, it was not associated with partial seizures, prolonged seizures, or multiple seizures. Age of 33 months or older was significantly associated with a high risk of partial seizure, compared to age below 19 months (OR: 3.82; 95% CI: 1.47–9.92; P = 0.006). The risk of partial seizure decreased with increase in body temperature at seizure occurrence, and body temperature between 39.5 °C and 40.0 °C was significantly associated with a low risk of partial seizure, compared to body temperature lower than 39.0 °C (OR: 0.22, 95% CI: 0.06–0.85, P = 0.027). Partial seizure was strongly associated with a risk of prolonged seizure (OR: 5.16, 95% CI: 1.73– 15.32, P = 0.003). Age below 19 months was significantly associated with a high risk of multiple seizures, compared to age of 33 months or older (OR: 5.06, 95% CI: 1.78–14.36, P = 0.002).
4. Discussion In our prospective one-year study, patients with influenza A comprised 21.3 percent of all FS patients,
35
a proportion close to that reported by Chiu et al. [8], 21.7%, in 1998. On comparison of the influenza A and non-influenza groups, age was significantly higher in the former, while there were no significant differences in past history of FS or family history of seizure disorder between the two groups. The proportion of influenza A patients was significantly higher in the older age subgroup of FS patients than in the remaining FS patients, and most of the older patients with influenza A had a family history of seizures or past FS history. Our findings suggest that influenza A may be a major cause of FS, especially in older children, but that the pathogenesis of FS in older children may mainly involve the susceptibility to seizures they originally have. In the present study, no association was found between influenza A and partial seizures. According to the seizure classification using the International League Against Epilepsy criteria [18], partial seizures are those in which the first clinical and electroencephalographic changes indicate initial activation of a system of neurons limited to part of one cerebral hemisphere. The presence of underlying disorders such as congenital focal cortical lesions or acquired impairment of cerebral parenchyma in consequence of various insults is assumed to be a cause of partial seizures. There has, in fact, been a report of association of prior neurological abnormality with partial seizure [19]. Berg et al. [4] reported that 16.1% of 428 patients with first FS exhibited partial seizure in a prospective cohort study, an incidence similar to that in our patients (16.7%), though no association was noted between neurological abnormality and partial seizure in their study. Since we excluded patients with past history of central nervous system disorder, neurological abnormality, or mental retardation, patients with an underlying symptomatic etiology were excluded. These findings indicate that many cases of partial seizure are unrelated neurological abnormality. Susceptibility to FS is strongly influenced by genetic factors [20]. In a prospective family study, 10% of siblings and 10% of offspring of FS probands also had FS [21]. There are many cases of idiopathic partial epilepsy associated with genetic predisposition at a higher age than that of peak incidence of FS, and subgroups with a high incidence of past FS have been reported [22–24]. Since 58.3% of our patients with partial seizures had a family history of seizure, there may be a genetic predisposition to development of this type of seizure. The incidence of multiple seizures was 12.8% in patients with FS associated with influenza A, and not significantly different from that in non-influenza patients, nor was an association found between influenza A and multiple seizures on multivariate analysis. In the study reported by Chiu et al. [8], on the other hand, the incidences of multiple seizures in FS inpatients with influenza A were 22.2% in 1997 and 31.5% in 1998, suggesting association of influenza A with multiple seizures,
36
K. Hara et al. / Brain & Development 29 (2007) 30–38
Age 47 months (N = 173)
Age
48 months (N = 42) Influenza A Non-influenza
N=15 (35.7%)
N=32 (18.5%)
· Family history of seizures (+) · Previous FS history (+) · Family history of seizures (-)
· Family history of seizures (+) 10 (31.3%) · Previous FS history (+) 14 (43.8%) · Family history of seizures (-) 15 (46.9%) and previous FS history (-)
and previous FS history (-)
7 (46.7%) 8 (53.3%) 3 (20.0%)
P = 0.015 Fig. 3. Older FS patients, aged 48 months or older (N = 42), were compared with the other FS patients aged 47 months or younger (N = 173) for proportion of influenza A patients. The proportion in the older age group (15/42, 35.7%) was significantly higher than that in the remaining patients (32/173, 18.5%) (P = 0.015), but the influenza A patients in the older age group exhibited high incidence of family history of seizure and past history of FS, as did the remaining patients. Note that only 3 patients had neither previous FS history nor family history of seizures among the older-aged influenza A patients.
though the statistical significance of these findings was not presented. Selection of patients may have been biased in their study, since they retrospectively collected only inpatients, which may have resulted in a higher incidence of multiple seizures than in our study. Low age was independently associated with multiple seizures in our study. Suga et al. [25] clinically investigated FS associated with primary HHV-6 infection, and found that occurrence at a low age and a high incidence of clustering seizures were characteristics of this condition. It is possible that multiple seizures are an infective-pathogen-dependent characteristic of infectious diseases that occur at high incidence in young infants. Only a few studies have mentioned postictal impairment of consciousness in FS. Yamamoto [26] emphasized the characteristics of postictal impairment of consciousness with convulsive manifestations in FS patients, based on corresponding electroencephalographic findings of non-ictal patterns. In a study of 213 consecutive cases of FS, Okumura et al. [27] found the duration of postictal unconsciousness to be less than 30 min in 93% of patients, and found associations of non-generalized seizure, seizures P5 min in duration,
and use of intravenous diazepam with prolonged unconsciousness. In our one-year study, postictal impairment of consciousness manifested by drowsiness, lethargy, or coma persisted for 30 min or longer in 13% of patients. It is reasonable for prolonged seizure or use of intravenous diazepam to be associated with this relatively rare condition, considering their relatively direct effects on the central nervous system. However, ours is the first report in which strong association of influenza A with PPIC is described. Influenza infection often progresses to encephalitis/encephalopathy, particularly in the pediatric population [11–14], and influenza-specific involvement of the central nervous system is suspected in such patients. Although our patients with PPIC were distinctly different from those with encephalitis or encephalopathy in the benignancy of their clinical course, the findings of our study also suggest the neurovirulence of influenza virus. There have recently been reports suggesting the involvement of various inflammatory cytokines, such as interleukin six, in the pathology of influenza-associated FS and encephalitis/encephalopathy [28,29]. On the other hand, some clinical and animal studies have demonstrated relationships between
K. Hara et al. / Brain & Development 29 (2007) 30–38
37
Table 3 Logistic regression analysis: independent risk factors for atypical features of FS; partial seizure, prolonged seizure, multiple seizures during the same febrile illness, and PPIC Atypical features
Risk factors
Partial seizure
Age (months) <19 19 6, <33 33 6 Body temperature at seizure occurrence (°C) <39.0 39.0 6, <39.5 39.5 6, <40.0 40.0 6
1.00 1.26 3.82
0.43–3.71 1.47–9.92
0.669 0.006
1.00 0.85 0.22 0.40
0.31–2.36 0.06–0.85 0.13–1.25
0.757 0.027 0.115
Prolonged seizure
Partial seizure
5.16
1.73–15.32
0.003
Multiple seizures
Age (months) 33 6 19 6, <33 <19
1.00 1.45 5.06
0.45–4.93 1.78–14.36
0.515 0.002
4.44 25.36 8.32
1.52–12.95 4.23–151.89 1.53–45.31
PPIC
Influenza A Prolonged seizure Intravenous diazepam
OR
95% CI
P
0.006 <0.001 0.014
FS indicates febrile seizures; OR, odds ratio; Cl, confidence interval; PPIC, prolonged postictal impaired consciousness.
inflammatory cytokines and sleep regulation [30,31]. Unbalanced endogenous cytokine levels may play roles in the pathogenesis of persistent postictal impairment of consciousness, particularly in individuals with influenza virus infection. That many cases of influenza-associated encephalitis/encephalopathy have been reported from Japan compared with other countries indicates the susceptibility of Japanese children to this specific pathological condition. By the same token, PPIC associated with influenza may be a specifically Japanese problem. These speculations need to be confirmed in further studies. In conclusion, this study provides new information on the independent association of influenza A infection with the development of disorders of consciousness subsequent to febrile seizure.
References [1] Camfield P, Camfield C, Kurlemann G. Febrile seizures. Epileptic syndromes in infancy, childhood, and adolescence. 3rd ed. London: John Libbey & Co Ltd; 2002, p. 145–52. [2] Tsai ML, Hung KL. Risk factors for subsequent epilepsy after febrile convulsions. J Formos Med Assoc 1995;94:327–31. [3] Sapir D, Leitner Y, Harel S, Kraumer U. Unprovoked seizures after complex febrile convulsions. Brain Dev 2000;22:484–6. [4] Berg AT, Shinnar S. Complex febrile seizures. Epilepsia 1996;37:126–33. [5] Maytal J, Shinnar S. Febrile status epilepticus. Pediatrics 1990;86:611–6. [6] Brocklebank JT, Court SD, McQuillin J, Gardner PS. InfluenzaA infection in children. Lancet 1972;2:497–500.
[7] Price DA, Postlethwaite RJ, Longson M. Influenzavirus A2 infections presenting with febril convulsions and gastrointestinal symptoms in young children. Clin Pediatr 1976;15:361–7. [8] Chiu SS, Tse CY, Lau YL, Peiris M. Influenza A infection is an important cause of febrile seizures. Pediatrics 2001;108:E63. [9] van Zeijl JH, Mullaart RA, Borm GF, Galama JM. Recurrence of febrile seizures in the respiratory season is associated with influenza A. J Pediatr 2004;145:800–5. [10] Stricker T, Sennhauser FH. Complex febrile seizures associated with influenza A. Pediatr Infect Dis J 2004;23(5):480. [11] Yoshikawa H, Yamazaki S, Watanabe T, Abe T. Study of influenza-associated encephalitis/encephalopathy in children during the 1997 to 2001 influenza seasons. J Child Neurol 2001;16:885–90. [12] Morishima T, Togashi T, Yokota S, Okuno Y, Miyazaki C, Tashiro M, et al. Encephalitis and encephalopathy associated with an influenza epidemic in Japan. Clin Infect Dis 2002;35:512–7. [13] Fujimoto S, Kobayashi M, Uemura O, Iwasa M, Ando T, Katoh T, et al. PCR on cerebrospinal fluid to show influenza-associated acute encephalopathy or encephalitis. Lancet 1998;352:873–5. [14] Takahashi M, Yamada T, Nakashita Y, Saikusa H, Deguchi M, Kida H, et al. Influenza virus-induced encephalopathy: clinicopathologic study of an autopsied case. Pediatr Int 2000;42:204–14. [15] Maricich SM, Neul JL, Lotze TE, Cazacu AC, Uyeki TM, Demmler GJ, et al. Neurologic complications associated with influenza A in children during the 2003-2004 influenza season in Houston, Texas. Pediatrics 2004;114:e626–33. [16] Huang YC, Lin TY, Wu SL, Tsao KC. Influenza A-associated central nervous system dysfunction in children presenting as transient visual hallucination. Pediatr Infect Dis J 2003;22:366–8. [17] Okumura A, Nakano T, Fukumoto Y, Higuchi K, Kamiya H, Watanabe K, et al. Delirious behavior in children with influenza: its clinical features and EEG findings. Brain Dev 2005;27:271–4. [18] International League Against Epilepsy. Proposal for revised clinical and electroencephalographic classification of epileptic seizures. Epilepsia 1981;22:489–501.
38
K. Hara et al. / Brain & Development 29 (2007) 30–38
[19] Shinnar S, Berg AT, Moshe SL, Petix M, Maytal J, Kang H, et al. Risk of seizure recurrence following a first unprovoked seizure in childhood: a prospective study. Pediatrics 1990;85:1076–85. [20] Kugler SL, Johnson WG. Genetics of the febrile seizure susceptibility trait. Brain Dev 1998;20:265–74. [21] Doose H, Maurer A. Seizure risk in offspring of individuals with a history of febrile convulsions. Eur J Pediatr 1997;156:476–81. [22] Casetta I, Monetti VC, Malagu S, Paolino E, Govoni V, Fainardi E, et al. Risk factors for cryptogenic and idiopathic partial epilepsy: a community-based case-control study in Copparo, Italy. Neuroepidemiology 2002;21:251–4. [23] Ward N, Evanson J, Cockerell OC. Idiopathic familial temporal lobe epilepsy with febrile convulsions. Seizure 2002;11:16–9. [24] Doose H, Petersen B, Neubauer BA. Occipital sharp waves in idiopathic partial epilepsies–clinical and genetic aspects. Epilepsy Res 2002;48:121–30. [25] Suga S, Suzuki K, Ihira M, Yoshikawa T, Kajita Y, Ozaki T, et al. Clinical characteristics of febrile convulsions during primary HHV-6 infection. Arch Dis Child 2000;82:62–6.
[26] Yamamoto N. Prolonged nonepileptic twilight state with convulsive manifestations after febrile convulsions: a clinical and electroencephalographic study. Epilepsia 1996;37:31–5. [27] Okumura A, Uemura N, Suzuki M, Itomi K, Watanabe K. Unconsciousness and delirious behavior in children with febrile seizures. Pediatr Neurol 2004;30:316–9. [28] Masuyama T, Matsuo M, Ichimaru T, Ishii K, Tsuchiya K, Hamasaki Y. Possible contribution of interferon-a to febrile seizures in influenza. Pediatr Neurol 2002;27:289–92. [29] Aiba H, Mochizuki M, Kimura M, Hojo H. Predictive value of serum interleukin-6 level in influenza virus-associated encephalopathy. Neurology 2001;57:295–9. [30] Vgontzas AN, Papanicolaou DA, Bixler EO, Kales A, Tyson K, Chrousos GP. Elevation of plasma cytokines in disorders of excessive daytime sleepiness: role of sleep disturbance and obesity. J Clin Endocrinol Metab 1997;82:1313–6. [31] Hogan D, Morrow JD, Smith EM, Opp MR. Interleukin-6 alters sleep of rats. J Neuroimmunol 2003;137:59–66.