Predictive value of EEG for febrile seizure recurrence

Brain & Development xxx (2017) xxx–xxx www.elsevier.com/locate/braindev

Original article

Predictive value of EEG for febrile seizure recurrence Alberto M. Cappellari a,⇑, Carolina Brizio b, Marta B. Mazzoni b, Giuseppe Bertolozzi b, Federica Vianello b, Alessia Rocchi b, Massimo Belli a, Andrea Nossa a, Dario Consonni c, Gregorio P. Milani d, Emilio F. Fossali b a Department of Neuroscience, Fondazione IRCCS Ca’ Granda, Ospedale Maggiore Policlinico, Milan, Italy Department of Pediatric Emergency, Fondazione IRCCS Ca’ Granda, Ospedale Maggiore Policlinico, Clinica De Marchi, Milan, Italy c Epidemiology Unit, Fondazione IRCCS Ca’ Granda, Ospedale Maggiore Policlinico, Milan, Italy d Pediatric Unit, Fondazione IRCCS Ca’ Granda Ospedale Maggiore Policlinico and Department of Clinical Sciences and Community Health, Universita‘ degli Studi di Milano, Milan, Italy

b

Received 10 October 2017; received in revised form 2 December 2017; accepted 5 December 2017

Abstract Objective: To define the role of the EEG in predicting recurrence of febrile seizures (FS) in children after a first FS. Methods: Children with a first simple or complex FS who underwent EEG at our hospital were retrospectively enrolled. EEG recordings were classified in three groups: normal, abnormal (slow activity or epileptiform discharges), and pseudo-petit mal discharge (PPMD) pattern. Children were followed-up for at least three years. Results: A total of 126 patients met the entry criteria, and 113 of them completed the follow-up. Risk of FS recurrence decreased linearly with increasing age ( 2% per month). The risk was higher among patients with PPMD pattern (absolute risk 86%, adjusted relative risk 2.00) and abnormal EEG (epileptiform discharges: absolute risk 71%, adjusted relative risk 2.00; slow activity: absolute risk 56%, adjusted relative risk 1.44), compared with those with normal EEG (absolute risk 41%). Conclusions: PPMD and abnormal EEG should be considered as an independent risk factor for FS recurrence. Ó 2017 The Japanese Society of Child Neurology. Published by Elsevier B.V. All rights reserved.

Keywords: Febrile seizures; EEG; Pseudo-petit mal discharge

1. Introduction Febrile seizures (FS) are the most common type of seizures in childhood [1]. Most of them are simple FS, while 25–35% are complex FS [2]. The American Academy of Pediatrics guidelines state that electroencephalogram (EEG) is not justified in the evaluation of a neurologically healthy child with a simple FS and ⇑ Corresponding author at: Fondazione IRCCS Ca’ Granda, Ospedale Maggiore Policlinico, Via Sforza 35, 20122 Milan, Italy. E-mail address: [email protected] (A.M. Cappellari).

normal neurodevelopment [3], while the value of EEG in patients with complex FS remains controversial [4]. However, patients with simple FS are occasionally investigated by EEG because of parental demand, FS recurrence or their physician’s recommendation [5]. Furthermore, even in children with simple FS, abnormal EEG findings may increase the risk of subsequent unprovoked seizures [6]. There is longstanding debate on the usefulness of EEG in children with FS to predict long-term outcome [7], and some authors reported that there is no evidence that EEG abnormalities help to predict either the recurrence of FS or the development of epilepsy [8,9]. However, the predictive value of EEG

https://doi.org/10.1016/j.braindev.2017.12.004 0387-7604/Ó 2017 The Japanese Society of Child Neurology. Published by Elsevier B.V. All rights reserved.

Please cite this article in press as: Cappellari AM et al. Predictive value of EEG for febrile seizure recurrence. Brain Dev (2017), https://doi.org/ 10.1016/j.braindev.2017.12.004

2

A.M. Cappellari et al. / Brain & Development xxx (2017) xxx–xxx

patterns different from epileptiform discharges has not been fully investigated. The main aim of our study was to evaluate the possible role of EEG in predicting recurrence of FS in children after a first FS, during a followup of at least three years. 2. Patients and methods

for age, without any abnormality, (2) abnormal: focal or generalized slow activity, or epileptiform discharges, (3) PPMD pattern: paroxysmal nearly generalized high voltage 3–4/sec waves with poorly developed spikes between the slow waves, occurring in drowsiness [11] (Fig. 1). Information on FS recurrence and occurrence of unprovoked seizures was obtained by a telephone interview, during a follow-up of at least 3 years.

2.1. Study design 2.5. Outcome measures This was a retrospective chart review of consecutive children with a first FS who underwent EEG at our hospital from January 2007 to January 2013. Informed consent was obtained from parents of all patients.

Main outcome measure was the recurrence of febrile seizures (FS) in children with a first FS. 2.6. Statistical analysis

2.2. Patient population We included children aged 6–72 months evaluated in our hospital after a first FS, who had both an EEG and a follow-up of at least three years. Although FS is usually defined to occur in infants or children 6–60 months of age, we included infants until 72 months owing to possible occurrence of simple FS in Genetic Epilepsy with Febrile Seizures Plus (GEFS+), whose onset is typically between 6 months and 6 years of age [10]. Children with previous unprovoked seizures or known neurological disorders (two patients with cerebral palsy) were excluded from the study. 2.3. Definitions FS was defined as a seizure occurring in association with a body temperature 38 °C in the absence of central nervous system infection, metabolic disturbance, or acute electrolyte imbalance. Simple FS were defined as generalized, with duration <15 min and not recurrent within 24 h. Complex FS were either focal, repetitive within 24 h or with a duration >15 min. Recurrent FS was defined as a seizure occurring during a new episode of fever defined as above in a child who had already experienced at least one previous FS. Epilepsy was defined as two or more unprovoked seizures occurring >24 h apart. 2.4. Procedure EEG studies were performed using the 10–20 international system with bipolar montages, after a median time interval of 30 days (range 10–60). EEG was obtained without pharmacological sedation. Each recording included wake and sleep stages, and activation procedures were performed when possible. EEG was interpreted by a pediatric neurologist and neurophysiologist (AMC). EEG findings were classified in three groups: (1) normal: background activity appropriate

Data were analyzed by chi-squared test for categorical variables and Wilcoxon-Mann-Whitney test for the continuous ones. Univariate and multiple Poisson regression models with robust variance were fitted to calculate relative risks (RR) of FS recurrence and 95% confidence intervals (95% CI) [12]. Statistical analyses were performed using Stata 13 (StataCorp. 2013. Stata: Release 13. Statistical Software. College Station, TX: StataCorp LP). 3. Results A total of 126 patients met the entry criteria, and 113 of them completed the follow-up. Clinical and EEG characteristics at admission of 113 children with complete follow-up are illustrated in Table 1, and their risks of recurrence of FS are showed in Table 2. First seizure duration varied from 0.5 to 40 min (median 2, mean 3.5), 6 patients presented with seizure >15 min, while data on the second seizure duration were not available in all patients. Risk of FS recurrence decreased linearly with increasing age ( 2% per month). Mean age at the time of recurrence of FS was 30 months. Time between the onset of fever and the initial seizure was <1-h only in 4 children. However, we lacked data on 25 children. In any case, we verified that risk did not vary much (RR = 1.01 per hour). For these reasons we did not include this variable in the final regression model. Children with PPMD and abnormal EEG pattern showed a higher risk of FS recurrence (PPMD: absolute risk 86%, adjusted relative risk 2.00; epileptiform discharges: absolute risk 71%, adjusted relative risk 2.00; slow activity: absolute risk 56%, adjusted relative risk 1.44) compared to children with normal EEG (absolute risk 41%). Children with PPMD and epileptiform discharges had more frequently a family history of FS than children with normal EEG, about three times, but we evaluated risk of FS adjusted for family history of FS.

Please cite this article in press as: Cappellari AM et al. Predictive value of EEG for febrile seizure recurrence. Brain Dev (2017), https://doi.org/ 10.1016/j.braindev.2017.12.004

A.M. Cappellari et al. / Brain & Development xxx (2017) xxx–xxx

3

Fig. 1. Pseudo-petit mal discharge: paroxysmal generalized high voltage slow-waves with intermixed poorly developed spikes, occurring in drowsiness.

Table 1 Clinical and EEG characteristics at admission in 113 children with complete follow-up. Variable

All

EEG Normal

PPMD

Slow

Epileptic

Patients Age (months) Gender (F/M)

113 24 [8–71] 51/62

68 (100) 23.5 [8–71] 31/37

29 (100) 26 [8–62] 12/17

9 (100) 31 [11–70] 4/5

7 (100) 28 [19–62] 4/3

Type of FS Simple Complex Focal* Repetitive within 24 h* Duration > 15 min* Family history of FS Family history of epilepsy

87 (77) 26 (23) 10 (38) 10 (38) 6 (23) 20 (18) 6 (5)

52 (76) 16 (24) 5 (31) 8 (50) 3 (19) 8 (12) 3 (4)

26 (89) 3 (11) 2 (67) 1 (33) 0 9 (31) 2 (7)

5 4 1 1 2 1 0

4 3 2 0 1 2 1

Maximal axillary temperature <39 °C 39–39.9*C 40* C

52 (46) 45 (40) 16 (14)

35 (51) 22 (32) 11 (16)

9 (31) 15 (52) 5 (17)

5 (56) 4 (44) 0 (0)

(56) (44) (25) (25) (50) (11) (0)

(57) (43) (67) (33) (29) (14)

3 (43) 4 (57) 0 (0)

Abbreviations: FS, febrile seizures; PPMD, pseudo-petit mal discharge. Results are reported as median [range], or as number (%). * Percentages calculated over children with complex febrile seizures.

Please cite this article in press as: Cappellari AM et al. Predictive value of EEG for febrile seizure recurrence. Brain Dev (2017), https://doi.org/ 10.1016/j.braindev.2017.12.004

4

A.M. Cappellari et al. / Brain & Development xxx (2017) xxx–xxx

Table 2 Absolute and relative risks (RR) and 95% confidence intervals (CI) of recurrence of febrile seizures according to selected characteristics, in 113 children with complete follow-up. Variable

FS recurrence N (%)

RR, crude

95% CI

RR, adjusted*

95% CI

Age, months

63 (56)

0.98

0.97–99

0.98

0.97–0.99

Gender Female Male

28 (55) 35 (56)

1.00 1.03

Reference 0.74–1.43

1.00 1.13

Reference 0.81–1.56

Family history of FS No Yes

49 (53) 14 (70)

1.00 1.33

Reference 0.94–1.88

1.00 1.06

Reference 0.73–1.54

Family history of epilepsy No 59 (55) Yes 4 (67)

1.00 1.21

Reference 0.67–2.19

1.00 1.06

Reference 0.61–1.84

Type of FS Simple Complex

49 (56) 14 (54)

1.00 0.96

Reference 0.64–1.43

1.00 0.92

Reference 0.62–1.37

Maximal axillary temperature <39 °C 27 (52) 39–39.9 °C 28 (62) 40 °C 8 (50)

1.00 1.20 0.96

Reference 0.85–1.70 0.55–1.68

1.00 1.03 0.88

Reference 0.74–1.44 0.53–1.47

EEG pattern Normal PPMD Slow Epileptiform

1.00 2.09 1.35 1.73

Reference 1.52–2.88 0.70–2.59 1.00–3.01

1.00 2.00 1.44 2.00

Reference 1.42–2.82 0.80–2.57 1.20–3.32

28 (41) 25 (86) 5 (56) 5 (71)

Abbreviations: FS, febrile seizures; PPMD, pseudo-petit mal discharge. * Each variable adjusted for the others.

4. Discussion Our study indicates that PPMD pattern and abnormal EEG are predictive of FS recurrence. In 1964, Gibbs and Gibbs described PPMD as a paroxysmal discharge consisting of generalized or nearly generalized high voltage 3–4/sec waves with a poorly developed spike in the positive trough between the slow waves occurring in drowsiness only. A history of FS was quite common in these children [11]. In 1983, Alvarez et al. described a similar pattern, which they called ‘‘hypnagogic paroxysmal spike and wave activity”. This finding was found in 23% of patients with FS, while it was not observed in the normal control group [13]. However, the usefulness of PPMD as a predictive marker for recurrence of FS has not been subsequently investigated. In our study, the incidence of PPMD in children with FS (25.7%) was similar to that reported by Alvarez et al. [13]. Furthermore, absolute risk of FS recurrence was higher among those with PPMD (86%) and abnormal EEG (epileptiform discharges: 71%; slow activity: 56%), compared with those with normal EEG (41%). Patients with PPMD and epileptiform discharges could present with simple or complex FS. Interestingly, PPMD had a very low incidence of complex FS (11%) compared with abnormal EEG (epileptiform discharges

43%, slow activity 44%). Furthermore, PPMD was never detected in the subgroup of complex FS with duration >15 min. We found no relationship between the location of epileptiform discharges and recurrence of FS. Type of recurrent FS, simple or complex, was not fully investigated at follow-up. Therefore, we cannot make a relationship between EEG at admission and type of recurrent FS. In our population, 41% of patients with normal EEG had FS recurrence. We run the regression models among children with normal EEG: apart from age we did not find other important risk factors of FS recurrence. Possibly, our patients represent a selected sample admitted to a tertiary care hospital. Some authors reported no significant correlation between the recurrence rate of FS in patients with normal EEG and that of patients whose EEGs showed epileptiform discharges [5]. Clinical predictors of recurrent FS are well-known since the work by Berg et al., consisting of young age at onset, history of FS in a first-degree relative, low degree of fever while in the emergency department and brief duration (<1-h) of recognized fever prior to the initial seizure [14]. As reported by the authors, age of onset was examined both as a simple dichotomous factor (<18 vs  18 month) and as an ordinal variable measured in 6-month increments after the age of 18 months [14].

Please cite this article in press as: Cappellari AM et al. Predictive value of EEG for febrile seizure recurrence. Brain Dev (2017), https://doi.org/ 10.1016/j.braindev.2017.12.004

A.M. Cappellari et al. / Brain & Development xxx (2017) xxx–xxx

We also found that younger age was a risk factor for FS recurrence, and the risk decreased linearly with increasing age (-2% per month). Nonetheless, multivariate analysis identified PPMD as an independent risk factor of FS recurrence. Our study has several limitations, including its retrospective nature, short duration of follow-up and timing of EEG. The time elapsed from FS is one of several variables reported to explain the wide range (2–86%) of EEG abnormalities in FS [15]. However, a recent study suggests that distribution of abnormalities in early EEG (24–48 h) and late EEG (2 weeks) recording was the same in patients with FS [16]. There is a debate on the role of EEG in children with FS to predict long-term outcome [7]. Some authors suggest that epileptiform discharges on EEG are not predictive factors of the recurrence of FS, but important predictors of epilepsy development [5,6]. However, to our knowledge, the predictive value of other EEG findings such as PPMD has not been fully investigated in the literature. FS recurrence is an important issue for both parents and clinicians because it is a frightening experience. Young age at onset, family history of FS, low degree of fever and short interval between fever onset and initial seizure has been reported as main predictors of recurrent FS [14]. Our study indicates that PPMD pattern and abnormal EEG should be considered as another independent risk factor for FS recurrence. 5. Conclusion EEG may have a role in predicting recurrence of FS in children after a first FS. Although the limitations of our study, we suggest that the practice of obtaining EEG after a first FS needs better clarification by prospective studies. Conflicts of interest None. Financial support None. This research did not receive any specific grant from funding agencies in the public, commercial or not-forprofit sectors.

5

References [1] Carapetian S, Hageman J, Lyons E, Leonard D, Janies K, Kelley K, Fuchs S. Emergency department evaluation and management of children with simple febrile seizures. Clin Pediatr 2015;54(10): 992–8. [2] Harini C, Nagarajan E, Kimia AA, de Carvalho RM, An S, Bergin AM, Takeoka M, Pearl PL, Loddenkemper T. Utility of initial EEG in first complex febrile seizure. Epilepsy Behav 2015;52:200–4. [3] American Academy of Pediatrics (AAP). Febrile seizures: guideline for the neurodiagnostic evaluation of the child with a simple febrile seizure. Subcommittee on febrile seizures. Pediatrics 2011;127:389–94. [4] Kim H, Byun SH, Kim JS, Lim BC, Chae JH, Choi J, Kim KJ, Hwang YS, Hwang H. Clinical and EEG risk factors for subsequent epilepsy in patients with complex febrile seizures. Epilepsy Res 2013;105:158–63. [5] Wo SB, Lee JH, Lee YJ, Sung TJ, Lee KH, Kim SK. Risk for developing epilepsy and epileptiform discharges on EEG in patients with febrile seizures. Brain Dev 2013;35:307–11. [6] Hwang G, Kang HS, Park SY, Han KH, Kim SH. Predictors of unprovoked seizure after febrile seizure: short-term outcomes. Brain Dev 2015;37:315–21. [7] Natsume J, Hamano SI, Iyoda K, Kanemura H, Kubota M, Mimaki M, Niijima S, Tanabe T, Yoshinaga H, Kojimahara N, Komaki H, Sugai K, Fukuda T, Maegaki Y, Sugie H. New guidelines for management of febrile seizures in Japan. Brain Dev 2017;39:2–9. [8] Kuturec M, Emoto SE, Sofijanov N, Dukovski M, Duma F, Ellenberg JH, Hirtz DG, Nelson KB. Febrile seizures: is the EEG a useful predictor of recurrences? Clin Pediatr 1997;36:31–6. [9] Pavlidou E, Panteliadis C. Prognostic factors for subsequent epilepsy in children with febrile seizures. Epilepsia 2013;54:2101–7. [10] Wirrell EC, Nickels KC. Electroclinical syndromes: infantile onset. In: Swaiman KF, Ashwal S, Ferriero DM, Schor NF, Finkel RS, Gropman AL, Pearl PL, Shevell MI, editors. Swaiman’s Pediatric Neurology. Sixth Edition: Elsevier; 2017. p. 557–68. [11] Niedermeyer E, Lopes da Silva F. Electroencephalography. 2nd ed. Urban and Schwarzenberg;1987. [12] Zou G. A modified poisson regression approach to prospective studies with binary data. Am J Epidemiol 2004;159:702–6. [13] Alvarez N, Lombroso CT, Medina C, Cantlon B. Paroxysmal spike and wave activity in drowsiness in young children: Its relationship to febrile convulsions. Electroencephalogr Clin Neurophysiol 1983;56:406–13. [14] Berg AT, Shinnar S, Darefsky AS, Holford TR, Shapiro ED, Salomon ME, Crain EF, Hauser AW. Predictors of recurrent febrile seizures. A prospective cohort study. Arch Pediatr Adolesc Med 1997 Apr;151(4):371–8. ucel O, Aka S, Yazicioglu L, Ceran O. Role of early EEG and [15] Y} neuroimaging in determination of prognosis in children with complex febrile seizure. Pediatr Int 2004;46:463–7. [16] Karimzadeh P, Rezayi A, Togha M, Ahmadabadi F, Derakhshanfar H, Azargashb E, Khodaei F. The best time for EEG recording in febrile seizure. Iran J Child Neurol 2014;8:20–5.

Please cite this article in press as: Cappellari AM et al. Predictive value of EEG for febrile seizure recurrence. Brain Dev (2017), https://doi.org/ 10.1016/j.braindev.2017.12.004