FULL-LENGTH ORIGINAL RESEARCH SUMMARY. * 1 Ting Zhang, 1 Jiannan Ma, Xiaoming Gan, and * #**Nong Xiao

Transcription

1 FULL-LENGTH ORIGINAL RESEARCH Are afebrile seizures associated with minor infections a single seizure category? A hospital-based prospective cohort study on outcomes of first afebrile seizure in early childhood * 1 Ting Zhang, 1 Jiannan Ma, Xiaoming Gan, and * #**Nong Xiao SUMMARY Ting Zhang is a graduate student pursuing a MM degree of Pediatrics at the Chongqing Medical University, Chongqing, China. Jiannan Ma, M.M.isan attending pediatric neurologist in the Neurology Department of Children s Hospital of Chongqing Medical University, Chongqing, China. Objective: To explore if afebrile seizures associated with minor infections are a single category of seizure, or a set of different kinds of seizures. Methods: We conducted this prospective cohort study on three kinds of first afebrile seizure: first afebrile seizure associated with gastrointestinal infection (AS- GI), first afebrile seizure associated with nongastrointestinal infection (AS-nGI), and first unprovoked seizure (US). The Kaplan-Meier estimate risks of recurrent seizures were analyzed and compared pairwise. The characteristics of recurrent seizures were also compared pairwise. Results: The Kaplan-Meier estimate risks of recurrent seizure at 2 years of the AS-GI, AS-nGI, and US groups were 6.9%, 23.7%, and 37.8%, respectively. The pairwise differences were significant between the AS-GI and US groups (p < 0.001) and between the AS-GI and AS-nGI groups (p = 0.001), but not significant between the US and AS-nGI groups (p = 0.066). Among unprovoked subsequent seizures in patients with recurrence, the pairwise differences were significant between the AS-GI and US groups (p < 0.001) and between the AS-GI and AS-nGI groups (p = 0.005), but not significant between the US and AS-nGI groups (p = 0.417). Significance: Afebrile seizures associated with minor infections are indeed of two distinguishable kinds: AS-GI, if free of risk factors such as a family history of epilepsy, had a better prognosis and should be categorized as an acute symptomatic seizure, whereas patients with first AS-nGI, like patients with first US, may have recurrent unprovoked seizures, which suggests this category s essential difference from AS-GI. KEY WORDS: Afebrile seizures, First seizure, Recurrence, Gastroenteritis, Epilepsy, Child. Accepted April 7, 2014; Early View publication May 23, *Departments of Rehabilitation, Neurology, Internal Medicine, Children s Hospital of Chongqing Medical University, Chongqing, China; Children Nutrition Research Center, Children s Hospital of Chongqing Medical University, Chongqing, China; Children s Hospital of Chongqing Medical University, Ministry of Education Key Laboratory of Developmental Diseases in Childhood, Ministry of Education Key Laboratory of Child Development and Disorders, Key Laboratory of Pediatrics in Chongqing, Chongqing International Science and Technology Cooperation Center for Child Development and Disorders, Chongqing, China; #Biological Engineering Institute of Chongqing University, Chongqing, China; and **Department of Pediatrics, University- Town Hospital of Chongqing Medical University, Chongqing, China 1 These authors contributed equally to this work and are co-first authors. Address correspondence to Nong Xiao, Department of Rehabilitation, Children s Hospital of Chongqing Medical University, No. 136 Zhongshan 2nd Road, Yuzhong District, Chongqing , China. xiaonongwl@163.com Wiley Periodicals, Inc International League Against Epilepsy 1001

2 1002 T. Zhang et al. Within the past 10 years, several studies have suggested a single category of seizure in addition to unprovoked seizures and febrile seizures: afebrile seizures associated with minor infections 1 or nonfebrile illness seizures. 2,3 The recurrence risks among these three were different. 1,3 The minor infections include gastrointestinal infection, respiratory infection (usually of the upper respiratory tract), and others of uncertain origin. For gastrointestinal infection, there is a concept of benign convulsion with mild gastroenteritis (CwG). The word benign indicates a good prognosis: a very low chance of having unprovoked seizures later in life. 4 6 In addition, CwG seems to be a new entity, which was more likely to be categorized as situation-related seizures than epilepsy, although it is not described in the classification of the International League Against Epilepsy (ILAE). 4,5,7 However, for afebrile seizure induced by nongastrointestinal infection, the risk of recurrence of unprovoked seizure is much higher than that for afebrile seizure induced by gastrointestinal infection, 1 whereas another study 3 shows that only this kind of first seizure and first unprovoked seizure would have subsequent unprovoked seizures. This evidence suggests that seizures caused by gastrointestinal infection and nongastrointestinal infection are essentially different: the former is probably in the category of situation-related seizures, whereas the latter may have a likeness to unprovoked seizure. Therefore, we designed this prospective cohort study on three kinds of first afebrile seizures: first afebrile seizure associated with gastrointestinal infection (AS-GI), first afebrile seizure associated with nongastrointestinal infection (AS-nGI), and first unprovoked seizure (US); and we studied their recurrent seizures, aiming to find out if afebrile seizures associated with minor infections are a single category of seizure, or a set of different kinds of seizures. Methods Study design and setting This prospective cohort study was performed at the Children s Hospital of Chongqing Medical University, one of the largest children s hospitals in China, providing primary to tertiary medical care to children mainly in the southwest of China, with >1 million annual outpatient visits and >50,000 annual in-patient visits. Evaluation Between January 2008 and June 2011, hospitalized patients with first afebrile seizure and consistent with the inclusion and exclusion criteria below were enrolled. The enrollment was conducted by pediatric neurologists. On admission, the medical history was taken and physical and neurologic examinations were conducted. The history, including details of first seizure, the temperature, and symptoms of infection before admission, was provided by caregivers and medical records by physicians in community hospitals. Family history was collected by asking caregivers and carefully double-checked because for most Chinese people, epilepsy still carries a stigma. Three-hour video electroencephalography (EEG) studies were performed while the patient was awake and asleep within a week of the seizures and were evaluated by the attending pediatric neurologists. EEG electrodes were placed according to the international system. Seizure types were classified according to ILAE terminology 8 based on the clinical manifestation from the history, physician s observations during hospitalization, and video-eeg findings. Patients with clustered seizures were administered phenobarbital, midazolam, and chloral hydrate alone or in combination, but none of our patients were treated with longterm anticonvulsants. If the parents or legal guardians agreed to participate and signed the informed consent form, our trained study personnel interviewed them to collect further information. Patients with unprovoked first seizure were assigned to the US group. Patients with afebrile seizures associated with minor infections were classified into two groups according to the site of infection based on clinical diagnosis: the AS-GI group, with symptoms of diarrhea with or without vomiting; and the AS-nGI group, with symptoms of sneezing, rhinorrhea, cough and/or rash. Patients in these two groups might also have the symptom of fever, but not on the day of the seizure. Though etiologic diagnosis was not necessary, all patients with suspicious gastrointestinal infection underwent a stool rotavirus antigen test. Stool samples were tested by SD Bioline Rotavirus Rapid Test (Standard Diagnostics, Inc., Yongin-si, Korea). Follow-up After discharge, patients were followed every 3 months by telephone. For those patients who were admitted to our hospital again, follow-up interviews were also conducted. Further information included whether the patient had had a recurrence of seizure, and if so, all the details about the seizure including the time, the temperature, and the accompany symptoms and whether the patient was treated with anticonvulsants. If a recurrent seizure occurred, the patient s follow-up ended. A follow-up was valid only if it was longer than 2 years or the patient had a recurrence of seizure. Because the last patient was enrolled at June 2011 and the follow-up ended at June 2013, the underlying shortest observation period was 2 years after the patients first seizure. We set the shortest follow-up time as 2 years considering that the 2-year recurrence risk is a useful indicator, because around 80 90% of recurrences are likely to happen within 2 years. 9 Participants The inclusion criteria were (1) first seizure <1 week before visiting our hospital; (2) age 1 month to 3 years; (3) afebrile seizure; and (4) meeting the definition below of

3 1003 Categorization of First Afebrile Seizure unprovoked seizure or afebrile seizure associated with minor infections presented. The exclusion criteria were (1) epileptic spasms; (2) acute symptomatic seizures caused by proximate insults; (3) presence of neurologic deficits (i.e., pre-, peri-, or postnatal brain damage and/or abnormal psychomotor development); and (4) taking long-term anticonvulsants. Because CwG occurs in healthy infants and young children, 4,10 we set the inclusion criteria of age from 1 month to 3 years to coincide with the peak age of incidence of CwG. Patients having epileptic spasms were excluded because of their tendency toward recurrence. To exclude proximate insults, we routinely conducted blood glucose, electrolyte panels (particularly sodium and calcium), lumbar puncture, EEG, and computerized/magnetic resonance imaging (CT/MRI), and conducted toxicology screening when necessary. Especially for patients with diarrhea and vomiting, we carefully excluded moderate to severe dehydration and electrolyte imbalance by clinical manifestations and electrolyte panels. Definitions Afebrile seizure is defined as seizures in which fever (body temperature 38 C) is not present on the same day. 2 An afebrile seizure associated with minor infections is an afebrile seizure that occurs in association with an acute infection (during the week before or 3 days after the seizure) that does not usually affect the brain. There may be fever at some time during the illness, but the child is afebrile on the day of the seizure. 2 A cluster of seizures during the period of infection was regarded as a first seizure. An unprovoked seizure is a seizure that occurs in the absence of any fever or infectious illness and acute (usually within a week) systemic or central nervous system insult. 1 For patients having an unprovoked seizure, a cluster of seizures or status seizures within 24 h was regarded as a first seizure Status epilepticus was defined as a seizure lasting >30 min or as a series of seizures without regaining consciousness between them lasting >30 min. 12 A family history of seizures is defined as a history of any kind of seizures in first- to third-degree relatives. Analysis Calculations were performed by means of SPSS statistical software for Windows, version 17.0 (Chicago, IL, U.S.A.). The level of statistical significance was established at p < The recurrence risk after first seizure was calculated by Kaplan-Meier survival estimates and one minus survival plots. The log-rank test was used to compare survival estimates risks pairwise in the Kaplan-Meier model. In addition, we conducted comparisons between groups using the chi-square test or Fisher s exact test for qualitative data, and the Kruskal-Wallis test or Mann-Whitney test for quantitative data. Ethics The study was approved by the ethics committee of the Children s Hospital of Chongqing Medical University. Informed consent was obtained from the parents or legal guardian of each child. Results Our inclusion and exclusion criteria were met by 302 children; 15 children s parents or legal guardians refused to participate, so we enrolled 287 patients at the initial evaluation. Of these patients, 31 did not have valid followup, and one took traditional Chinese antiepileptic herbs. Thus, 255 patients were included in our study. Among them, 101 (39.6%) were in the AS-GI group, 80 (31.4%) were in the AS-nGI group, and 74 (29.0%) were in the US group (Fig. S1). Demographic characteristics and clinical features The demographic characteristics and clinical features of patients in the three groups are summarized in Table 1. Among the subjects of our study, there were 114 boys (44.7%) and 141 girls (55.3%), and the age of first seizure ranged from 1 month to 3 years (median 14 months). The age distributions of the three groups were similar (data not shown). There was no statistically significant difference for any of the items of gender, age, or family history of seizures) among the three groups or between any two groups. As for the clinical features of patients among the three groups, only seizure frequencies showed a statistically significant difference (p = 0.008), other than seizure duration, having multiple seizures or having status epilepticus. Focal primary seizures were confirmed in 15 patients (5.9%) by observations of focal features such as hemiconvulsion, lateral gaze, and head deviation with or without ictal EEG recordings. Among the remaining majority who had an apparently generalized seizure, eight were revealed by ictal EEG recordings to have had focal seizures with secondary generalization, whereas the rest might have had generalized seizures or focal seizures with secondary generalization. Three-hour video-eeg examinations were performed in 243 patients: 72 in the US group, 75 in the AS-nGI group, and 96 in the AS-GI group. Abnormal interictal EEG findings (including background abnormalities and/or epileptic discharges) were not significantly different among the three groups. The ictal EEG was recorded in 11 patients (Table 2): 6 were in the US group, 2 in the AS-nGI group, and 3 in the AS-GI group. All of them revealed focal seizures with or without secondary generalization. Among the 101 patients with gastrointestinal infection, 30 patients (29.7%) tested positive for rotavirus antibody in stool, suggesting rotavirus infection.

4 1004 T. Zhang et al. Table 1. Demographic characteristic and clinical feature of 255 patients by group AS-GI n = 101 AS-nGI n = 80 US n = 74 Male (%) 41 (40.6) 32 (40.0) 41 (55.4) >0.05 a Median age (months) N/A Family history of seizures (%) 13 (12.9) 14 (17.5) 17 (23.0) >0.05 a Interval between onset of infections and that of the first seizure (%) 1 5 days 97 (96.0) 68 (85) N/A N/A 6 7 days 0 (0) 11 (13.8) N/A N/A Seizure preceded infection within 24 h 4 (4.0) 1 (1.2) N/A N/A Maximum duration of seizures, median (range) 2 ( ) 2.25 ( ) 1.5 (0.1 60) >0.05 b 1 5 min (%) 92 (91.1) 56 (70) 62 (83.8) N/A 5 30 min (%) 8 (7.9) 20 (25) 8 (10.8) N/A Status epilepticus 30 min (%) 1 (1) 4 (5.0) 4 (5.4) >0.05 c Frequency of seizures (within 24 h), median (range) 2.0 (1 10) 2.0 (1 13) 2.5 (1 20) b Multiple seizures (%) 64 (63.4) 58 (72.5) 59 (79.7) >0.05 a Focal primary seizures (%) 3 (3.0) 6 (7.5) 6 (8.1) N/A Ictal EEG abnormalities N/A Interictal EEG abnormalities (%) d 47 (49.0) 33 (44.0) 34 (47.2) >0.05 a Background abnormalities (%) 30 (31.3) 21 (28.0) 24 (33.3) >0.05 a Epileptic discharge (%) 25 (26.0) 20 (26.7) 22 (30.6) >0.05 a AS-GI, afebrile seizure associated with gastrointestinal infection; AS-nGI, afebrile seizure associated with nongastrointestinal infection; US, unprovoked seizure. a The chi-square test. b The Kruskal-Wallis test. c The Fisher s exact test. d Two hundred forty-three of 255 patients underwent EEG examination: 72 in the US group, 75 in the AS-nGI group, and 96 in the AS-GI group. p-value Table 2. Ictal EEG and seizure semiology in 11 patients with first afebrile seizure No. Origin of ictal discharge Seizure evolution Clinical manifestation Group 1 Left frontal/occipital Focal?Generalized Bilateral blinks AS-nGI 2 Left occipital Focal?Generalized Seizure induced by crying. Right lateral gaze?gtcs AS-GI 3 Right frontal Focal?Generalized GTCS AS-GI 4 Right frontal Focal Fixed gaze and head deviation to left?clonic movements US of the right leg 5 Right medial temporal Focal?Generalized Fixed gaze and head deviation to left?gtcs AS-nGI 6 Left occipital Focal?Generalized Right lateral gaze?gtcs AS-GI 7 Right frontal Focal?Generalized Oculocephalic deviation to left?clonic movements of legs US 8 Left occipital Focal Right lateral gaze?clonic movements of the right extremities US 9 Right frontal polar Focal Clonic movements of right corner of mouth, right corner US of right eye and right finger 10 Left medial-posterior temporal Focal?Generalized Right lateral gaze?gtcs US 11 Left frontal polar Focal?Generalized Oculocephalic deviation to right?staring US AS-GI, afebrile seizure associated with gastrointestinal infection; AS-nGI, afebrile seizure associated with nongastrointestinal infection; GTCS, generalized tonic clonic seizure; US, unprovoked seizure. Risk of recurrent seizures after three kinds of first afebrile seizures All of the patients who did not have a recurrence were followed for >2 years, and 36.5% of 255 patients were followed for >3 years. The longest follow-up term was 62 months. The Kaplan-Meier estimate risks of recurrent seizure at 2 years in the AS-GI, AS-nGI, and US groups were 6.9%, 23.7%, and 37.8%, respectively (Table 3). The Kaplan-Meier one minus survival plot is shown in Figure 1. The pairwise comparisons were significant between the AS-GI and US groups (p < 0.001) and between AS-GI and AS-nGI groups (p = 0.001), but not significant between the US and AS-nGI groups (p = 0.066). For the US group, Kaplan-Meier estimated risks of recurrent seizure were 24.3% at 3 months, 29.7% at 6 months, 35.1% at 1 year, and 37.8% at 2 years. Within this group, seizure recurred in 18 patients (64.3%) in the first 3 months, 22 (78.6%) within 6 months, and 26 (92.9%) within 1 year.

5 1005 Categorization of First Afebrile Seizure Table 3. The recurrence of seizure at 2 years among three groups Group n Recurrent seizure (%) No recurrent seizure (%) Comparison with AS-GI Comparison with AS-nGI AS-GI (6.9) 94 (93.1) N/A v 2 = , p = AS-nGI (23.8) 61 (76.3) v 2 = , p = N/A US (37.8) 46 (62.2) v 2 = , p < v 2 = 3.373, p = AS-nGI, afebrile seizure associated with nongastrointestinal infection; AS-GI, afebrile seizure associated with gastrointestinal infection; US, unprovoked seizure. For the US and AS-nGI groups, all of the recurrent provoked seizures were provoked by nongastrointestinal infection, whereas for the AS-GI group, all of the recurrent provoked seizures were provoked by gastrointestinal infection. Figure 1. The Kaplan-Meier one minus survival plot of seizure recurrence after first afebrile seizure, by study group. AS-GI, afebrile seizure associated with gastrointestinal infection; AS-nGI, afebrile seizure associated with nongastrointestinal infection; US, unprovoked seizure. Epilepsia ILAE For the AS-nGI group, Kaplan-Meier estimated risks of recurrent seizure were 17.5% at 3 months, 20% at 6 months, and 23.8% at 1 and 2 years. Within this group, seizure recurred in 14 patients (73.7%) in the first 3 months, 16 (84.2%) within 6 months, and 19 (100%) within 1 year. For the AS-GI group, Kaplan-Meier estimated risks of recurrent seizure were 2.0% at 3 months, 6.0% at 6 months, and 6.9% at 1 year. All of the recurrences (in seven patients) occurred within 8 months. Among them, two occurred in the first 3 months and six within 6 months. The condition coincident with recurrent seizures The condition coincident with recurrent seizure after each kind of first seizure is shown in Table 4. Among unprovoked subsequent seizures in patients with recurrence, the pairwise comparisons were significant between the AS-GI and US groups (p < 0.001) and between AS-GI and AS-nGI groups (p = 0.005), but not significant between the US and AS-nGI groups (p = 0.417). Discussion The three studies identifying afebrile seizures associated with minor infections or nonfebrile illness seizures as distinct from febrile seizures and unprovoked afebrile seizures 1 3 do not more precisely define the minor infections as gastrointestinal. Our prospective cohort study investigated not only the recurrence risk following a first afebrile seizure, but also the conditions associated with the recurrent seizure. We found that both characteristics differed between first AS-GI and first AS-nGI. We thus revealed AS-GI to be a unique entity with a better prognosis, whereas AS-nGI shared similarities with US. The rate of recurrent seizure for patients with first AS-GI in our study was 6.9%, which was lowest among the three groups and statistically significantly lower than that of patients with first US (37.8%) and first AS-nGI (23.8%). However, there was no statistically significant difference in rate of recurrent seizure between the first US group and the first AS-nGI group. In the prospective study on young children with first seizure by Martin et al., 3 seizure recurrence was highest among children with first unprovoked seizures, followed by children with febrile or nonfebrile-illness first seizures without acute gastrointestinal symptoms, and the seizure recurrence was lowest among children who had acute gastrointestinal symptoms at the time of their first seizure. The large-scale retrospective cohort study on children with first seizure in Singapore by Lee and Ong 1 also revealed that children with afebrile seizures associated with minor infections had a much lower risk of subsequent unprovoked afebrile seizures than patients with first unprovoked seizure, and among the former group, patients with gastroenteritis had a much lower risk than patients with respiratory infections or nonspecific febrile illnesses. The recurrence risks in our study were consistent with previous studies not only the two in similar settings, 1,3 but also the following only on unprovoked seizures or on seizures provoked by mild gastroenteritis.

6 1006 T. Zhang et al. Table 4. Kinds of recurrent seizure, by kind of first afebrile seizure Second seizure First seizure Gastrointestinal infection Nongastrointestinal infection Unprovoked Total AS-GI AS-nGI US AS-nGI, afebrile seizure associated with nongastrointestinal infection; AS-GI, afebrile seizure associated with gastrointestinal infection; US, Unprovoked seizure. There are many reports on recurrence risk following a first unprovoked seizure. In a quantitative review by Berg and Shinnar, 14 the reported risks of recurrence of unprovoked seizure after the first unprovoked seizure varied from 27% to 71%. In our study, the rate of recurrent unprovoked seizure after first US was 33.8%. The occurrence of seizures provoked by mild gastroenteritis is not a new concept. In 1982, Morooka 15 reported a clinical condition in which mild diarrhea could trigger afebrile convulsions with an excellent prognosis. Since then, CwG has been mostly reported in Japan, 5,10,16,17 as well as in China, United States, 21,22 and Europe. 6 The word benign indicates a favorable prognosis: % patients may have recurrence of CwG in early childhood, 5,6,23 but patients have no recurrence of unprovoked seizures and have normal psychomotor development. In our study, 100 out of 101 patients in the AS-nGI group met the diagnostic criteria for CwG. 6 The excluded patient had a family history of epilepsy and developed unprovoked seizures. Of the 100 CwG patients, 6 (6%) had recurrent seizure associated with gastrointestinal infection. In summary, all the previous studies of different settings and our study showed that patients with first AS-GI had a relatively lower recurrence risk than those with first US and first AS-nGI, which indicated that AS-GI is a distinct entity of seizure with a prognosis better than that of US and ASnGI. Based on the limited number of studies on AS-nGI, its recurrence risk is between that of AS-GI and US. The other major difference between AS-GI and AS-nGI found by our study was the conditions associated with the recurrent seizure. For patients having first AS-GI and a recurrent seizure, all but one had recurrent provoked seizures, whereas patients with first AS-nGI or first US were more likely to have recurrent unprovoked seizures rather than recurrent provoked seizures. What is more, all of the provoked recurrences of first AS-GI patients were concurrent with gastrointestinal infection, whereas for first ASnGI patients and first US patients, the few recurrent provoked seizures were all concurrent with nongastrointestinal infection. If we exclude the sole patient in the AS-GI group who did not meet the diagnostic criteria for CwG, this finding could be condensed: all of the patients with CwG in our study had only recurrent provoked seizures, and these only in conjunction with gastroenteritis. We support findings from studies on CwG: a few patients had recurrence of CwG, but no patients have recurrent unprovoked seizures. 5,6,23 In other words, gastroenteritis is a necessary condition of recurrent seizure for patients with CwG. As Hauser and Beghi 24 put it, one of the important aspects of the differences between acute symptomatic seizures (also called situation-related seizures) and epilepsy is that acute symptomatic seizures are not necessarily characterized by a tendency for recurrence unless there is recurrence of the underlying acute causal condition. In addition, in the opinion of Uemura et al., 5 Verrotti et al., 4 and Specchio and Vigevano, 7 CwG is more likely to be categorized as situation-related seizures than epilepsy, although it is not described in the ILAE classification. Thus, AS-GI, if free of risk factors such as family history of epilepsy, is therefore an entity of acute symptomatic seizures. In terms of acute symptomatic seizures, AS-GI shares several characteristics of the most common acute symptomatic seizures in febrile seizures 25 : (1) diarrhea or fever is an identifiable event preceding or concurrent with the seizure; (2) the rate of recurrent seizures is far lower than first unprovoked seizure; (3) the recurrent seizures are accompanied by a necessary condition; and (4) a small proportion of patients who have some risk factors, like a family history of epilepsy, may finally develop epilepsy. In all these aspects, the AS-GI is in all probability an entity in the same category with febrile seizures of acute symptomatic seizures. For the same reason, AS-nGI is likely distinct from acute symptomatic seizures, since 78.9% of the recurrent seizures after first AS-nGI were unprovoked seizures, which showed no statistically significant differences from the first US in our study. Martin et al. 3 also found that only patients with first US or first AS-nGI, not patients with first AS-GI, had recurrent unprovoked seizure. The evidence above strongly supported that patients with first AS-GI, if free of risk factors such as family history of epilepsy and meeting the diagnostic criteria of CwG, had a better prognosis and should be categorized as acute symptomatic seizures; whereas patients with first AS-nGI may have recurrent unprovoked seizures like patients with first US, which suggest its essential difference from AS-GI. However, the mechanism between gastrointestinal infection and seizures remains unclear, unlike other related conditions such as head injury, stroke, and central nervous system infection. Although several case reports revealed

7 1007 Categorization of First Afebrile Seizure that rotavirus was an identified etiology of nonfebrile seizures by evidences of its neurotropism, lacking of control group and the implicated viruses are common, it remains possible that chance alone may lead some children who develop seizures to have a concomitant common viral illness. 1,2,4 Our study supports this opinion, because only a small proportion (29.7%) of children with first AS-GI had evidence of rotavirus infection. We believe that genetic predisposition may play a role in this mechanism. In the Singapore study, 1 there was a significant excess of Chinese patients with provoked seizures compared with febrile seizures and unprovoked seizures. In a study of 105 Japanese CwG patients by Uemura et al., 5 family histories of febrile or afebrile seizures were noted in 7% and 6% of the patients with CwG, respectively; a familial case of CwG was documented in a pair of identical twins. Uemura and Okumura 26 also pointed out that reports on CwG have been rare in the United States and European countries. It may be that CwG is overlooked in United States and European countries, but possibly it simply is relatively rare among the main ethnic groups there. In our study, 12.9% patients with AS-GI had a family history of seizure, which is higher than Uemura et al. found, but still too small to be probative. Further studies on the mechanism of gastrointestinal infection and seizures may help us better understand CwG and seizure generation. Our study was also able to confirm several previous findings on CwG. For the 100 CwG patients in our study, 96 patients first seizures occurred within 5 days after the onset of gastroenteritis, though four preceded it by h; 64% of them had seizures in clusters, the duration of seizures was within 5 min in 91% of them, and one patient had status epilepticus. These match the typical seizure features of CwG in previous studies 4 6,10,20,23,27 29 : seizures tending to occur repetitively in clusters, short seizures occurring between the first and fifth day of viral gastroenteritis, though some preceded the onset of diarrhea for 24 h. The seizure type and features of EEG of CwG have been controversial in previous studies. Some have reported that all 16,27 or most 5,28,29 of patients with CwG experienced symmetric generalized seizures, while patients may sometimes have seizures resembling complex partial ones. 5,16 These studies also found that most 20 or all 5,28,29 of the interictal EEG studies were normal. However, Komori et al. 10 and Kawano et al. 29 believed that generalized seizure in CwG might involve partial seizures with secondary generalization, and video-eeg monitoring should be employed for further studies. This idea was supported by several studies on video-eeg. 6,17,30 The ictal EEG studies revealed that all seizures were of focal onset with or without secondarily generalized seizure. In our study, we met with the same situation that 97% of patients with CwG showed apparent symmetric generalized seizures; however, most of the results were based on recall by parents who might have been in shock, which were not reliable as most of them could only describe the climax performance but ignored the onset, which is important for judgments of seizure type. At the same time, the ictal EEG, the other crucial way to confirm seizure type, had very low positive rate. Of 95 CwG patients having EEG examination, only three had ictal EEGs. However, all three patients ictal EEGs showed focal seizures with secondary generalization (Table 2). Based on previous opinions and our finding, there was probably a large part of focal seizures with secondary generalization in the majority rest of patients who acted as apparent generalized seizures while having negative ictal EEG. Investigators previously reported that % of CwG patients had abnormal interictal EEG recordings, 5,6,27 29 whereas 48.4% of the interictal EEG findings of CwG appeared abnormal in our study. Our higher rate may be attributed to our patients undergoing 3-h EEG examination within 1 week after the first seizure, which was longer and earlier than in other studies. Conclusion AS-GI, if free of risk factors such as family history of epilepsy and meeting the diagnostic criteria of CwG, had a better prognosis and should be categorized as acute symptomatic seizures, whereas patients with first AS-nGI, like patients with first US, may have recurrent unprovoked seizures. Afebrile seizures associated with minor infections indeed are of two distinguishable kinds. Ours is one of the few available prospective studies about the recurrence risk after first seizure in childhood. Limitations Two similar studies were based on patients in emergency department, 2,3 whereas our study s participants were inpatients. Because consideration of children s safety led to admission of almost all of those presenting with first seizure, we are confident that our sample is representative of the target patients. However, as a hospital-based study, this study had an ineluctable selection bias. For unprovoked seizures, the rate of multiple seizures within 24 h was 79.7% in our study, much higher than in previous studies. 12,31,32 The reason might be that patients with only one unprovoked seizure may choose to stay in community hospitals or clinics. However, since multiple seizures within 24 h do not seem to carry an excess risk of recurrence when they are idiopathic or cryptogenic in origin, 9 this selection bias apparently had little influence on the recurrence rate. Acknowledgments We d like to express our sincere thanks to all the patients and their family involved in this study, as well as all the medical and allied health staff members who cared for the study children.

8 1008 T. Zhang et al. None. Funding Disclosure None of the authors has any conflict of interest to disclose. We confirm that we have read the Journal s position on issues involved in ethical publication and affirm that this report is consistent with those guidelines. References 1. Lee WL, Ong HT. Afebrile seizures associated with minor infections: comparison with febrile seizures and unprovoked seizures. Pediatr Neurol 2004;31: Zerr DM, Blume HK, Berg AT, et al. Nonfebrile illness seizures: a unique seizure category? Epilepsia 2005;46: Martin ET, Kerin T, Christakis DA, et al. Redefining outcome of first seizures by acute illness. Pediatrics 2010;126:e1477 e Verrotti A, Tocco AM, Coppola GG, et al. Afebrile benign convulsions with mild gastroenteritis: a new entity? Acta Neurol Scand 2009;120: Uemura N, Okumura A, Negoro T, et al. Clinical features of benign convulsions with mild gastroenteritis. Brain Dev 2002;24: Verrotti A, Nanni G, Agostinelli S, et al. Benign convulsions associated with mild gastroenteritis: a multicenter clinical study. Epilepsy Res 2011;93: Specchio N, Vigevano F. The spectrum of benign infantile seizures. Epilepsy Res 2006;70(Suppl. 1):S156 S Bancaud J, Henriksen O, Rubio-Donnadieu F, et al. Proposal for revised clinical and electroencephalographic classification of epileptic seizures. Epilepsia 1981;22: Seneviratne U. Management of the first seizure: an evidence based approach. Postgrad Med J 2009;85: Komori H, Wada M, Eto M, et al. Benign convulsions with mild gastroenteritis: a report of 10 recent cases detailing clinical varieties. Brain Dev 1995;17: Fisher RS, Leppik I. Debate: when does a seizure imply epilepsy? Epilepsia 2008;49(Suppl. 9): Ramos Lizana J, Cassinello Garcia E, Carrasco Marina LL, et al. Seizure recurrence after a first unprovoked seizure in childhood: a prospective study. Epilepsia 2000;41: Shinnar S, Berg AT, Moshe SL, et al. The risk of seizure recurrence after a first unprovoked afebrile seizure in childhood: an extended follow-up. Pediatrics 1996;98: Berg AT, Shinnar S. The risk of seizure recurrence following a first unprovoked seizure: a quantitative review. Neurology 1991;41: Morooka K. Convulsions and mild diarrhea. Shonika 1982;35: Abe T, Kobayashi M, Araki K, et al. Infantile convulsions with mild gastroenteritis. Brain Dev 2000;22: Imai K, Otani K, Yanagihara K, et al. Ictal video-eeg recording of three partial seizures in a patient with the benign infantile convulsions associated with mild gastroenteritis. Epilepsia 1999;40: Wong V. Acute gastroenteritis-related encephalopathy. J Child Neurol 2001;16: Wang YC, Hung KL. Benign seizures associated with mild diarrhea: clinical analysis of 20 cases. Zhonghua Min Guo Xiao Er Ke Yi Xue Hui Za Zhi 1993;34: Li T, Hong S, Peng X, et al. Benign infantile convulsions associated with mild gastroenteritis: an electroclinical study of 34 patients. Seizure 2014;23: Contino MF, Lebby T, Arcinue EL. Rotaviral gastrointestinal infection causing afebrile seizures in infancy and childhood. Am J Emerg Med 1994;12: Iyadurai S, Troester M, Harmala J, et al. Benign afebrile seizures in acute gastroenteritis: is rotavirus the culprit? J Child Neurol 2007;22: Caraballo RH, Ganez L, Santos CdeL, et al. Benign infantile seizures with mild gastroenteritis: study of 22 patients. Seizure 2009;18: Hauser WA, Beghi E. First seizure definitions and worldwide incidence and mortality. Epilepsia 2008;49(Suppl. 1): Graves RC, Oehler K, Tingle LE. Febrile seizures: risks, evaluation, and prognosis. Am Fam Physician 2012;85: Uemura N, Okumura A. Benign convulsions with mild gastroenteritis a worldwide clinical entity. Brain Dev 2005;27: Hung JJ, Wen HY, Yen MH, et al. Rotavirus gastroenteritis associated with afebrile convulsion in children: clinical analysis of 40 cases. Chang Gung Med J 2003;26: Narchi H. Benign afebrile cluster convulsions with gastroenteritis: an observational study. BMC Pediatr 2004;4: Kawano G, Oshige K, Syutou S, et al. Benign infantile convulsions associated with mild gastroenteritis: a retrospective study of 39 cases including virological tests and efficacy of anticonvulsants. Brain Dev 2007;29: Maruyama K, Okumura A, Sofue A, et al. Ictal EEG in patients with convulsions with mild gastroenteritis. Brain Dev 2007;29: Scotoni AE, Manreza ML, Guerreiro MM. Recurrence after a first unprovoked cryptogenic/idiopathic seizure in children: a prospective study from Sao Paulo, Brazil. Epilepsia 2004;45: Marson A, Jacoby A, Johnson A, et al. Immediate versus deferred antiepileptic drug treatment for early epilepsy and single seizures: a randomised controlled trial. Lancet 2005;365: Supporting Information Additional Supporting Information may be found in the online version of this article: Figure S1. Trial flow diagram.