- Email: [email protected]
Utility of initial EEG in first complex febrile seizure
Epilepsy & Behavior 52 (2015) 200–204
Contents lists available at ScienceDirect
Epilepsy & Behavior journal homepage: www.elsevier.com/locate/yebeh
Brief Communication
Utility of initial EEG in first complex febrile seizure Chellamani Harini a,⁎, Elanagan Nagarajan a, Amir A. Kimia b, Rachel Marin de Carvalho a, Sookee An a, Ann M. Bergin a, Masanori Takeoka a, Phillip L. Pearl a, Tobias Loddenkemper a a b
Division of Epilepsy and Clinical Neurophysiology, Department of Neurology, Boston Children's Hospital, Boston, MA, USA Division of Emergency Medicine, Boston Children's Hospital, Boston, MA, USA
a r t i c l e
i n f o
Article history: Received 3 August 2015 Revised 1 September 2015 Accepted 3 September 2015 Available online xxxx Keywords: Complex febrile seizures EEG Epilepsy Diagnosis
a b s t r a c t Objective: The risk of developing epilepsy following febrile seizures (FS) varies between 2% and 10%, with complex febrile seizures (CFS) having a higher risk. We examined the utility of detected epileptiform abnormalities on the initial EEG following a first CFS in predicting subsequent epilepsy. Methods: This was a retrospective study of consecutive patients (ages 6–60 months) who were neurologically healthy or mildly delayed, seen in the ED following a first CFS and had both an EEG and minimum of 2-year follow-up. Data regarding clinical characteristics, EEG report, development of subsequent epilepsy, and type of epilepsy were collected. Established clinical predictors for subsequent epilepsy in children with FS and EEG status were evaluated for potential correlation with the development of subsequent epilepsy. Sensitivity, specificity, and positive and negative predictive values of an abnormal EEG (epileptiform EEG) were calculated. Results: A group of 154 children met our inclusion criteria. Overall, 20 (13%) children developed epilepsy. The prevalence of epilepsy was 13% (CI 8.3–19.6%). Epileptiform abnormalities were noted in 21 patients (13.6%), EEG slowing in 23 patients (14.9%), and focal asymmetry in six (3.8%). Epileptiform EEGs were noted in 20% (4/20) of patients with epilepsy and 13% (17/134) of patients without epilepsy (p = 0.48). At an estimated risk of subsequent epilepsy of 10% (from population-based studies of children with FS), we determined that the PPV of an epileptiform EEG for subsequent epilepsy was 15%. None of the clinical variables (presence of more than 1 complex feature, family history of epilepsy, or status epilepticus) predicted epilepsy. Conclusions: An epileptiform EEG was not a sensitive measure and had a poor positive predictive value for the development of epilepsy among neurologically healthy or mildly delayed children with a first complex febrile seizure. The practice of obtaining routine EEG for predicting epilepsy after the first CFS needs clarification by welldefined prospective studies. © 2015 Elsevier Inc. All rights reserved.
1. Introduction Febrile seizures (FS) occur in 2–5% of children before the age of 5 years [1,2] with higher incidence rates of 8% in Japan and 14% in Guam [3]. Most febrile seizures are simple febrile seizures (SFS), and about 25–35% of FS are complex febrile seizures (CFS) [4,5]. The risk of developing epilepsy in children with febrile seizures varies between 2% and 10% depending on the length of follow-up and type of FS [7–9, 11]. Complex febrile seizures have a higher risk for developing epilepsy when compared with SFS, especially in CFS with multiple complex features [7] or occurring in the setting of neurodevelopmental delay [8]. The American Academy of Pediatrics guideline states that electroencephalogram (EEG) should not be performed in the evaluation of a neurologically healthy child with a SFS as there is no evidence that the EEG abnormality will predict recurrence of FS and/or the development of epilepsy [6]. No such guidelines exist for CFS. It is uncertain whether an ⁎ Corresponding author at: 300, Longwood Avenue, Boston MA-02115, USA. E-mail address: [email protected] (C. Harini).
http://dx.doi.org/10.1016/j.yebeh.2015.09.003 1525-5050/© 2015 Elsevier Inc. All rights reserved.
abnormal EEG after a first CFS provides predictive value for subsequent epilepsy. Some authors find no rationale in doing an EEG in children with CFS because of the paucity of evidence [14]. More recently, a few studies have reported that epileptiform discharges noted with CFS is a risk factor for the development of epilepsy [10,12,13]. However such findings may not alter clinical management. We examined the utility of epileptiform abnormalities (seen on an initial EEG) following the first CFS in predicting the subsequent development of epilepsy with a minimum follow-up of two years. Our secondary goal was to identify if clinical or other EEG characteristics predicted the development of epilepsy in children with a first CFS. 2. Methods 2.1. Study design This was a retrospective chart review of consecutive patients who were evaluated following the first episode of CFS at an urban tertiary care pediatric ED between 1996 and 2011.
C. Harini et al. / Epilepsy & Behavior 52 (2015) 200–204
The study was approved by the Boston Children's Hospital institutional review board. 2.2. Patient identification This was done in two steps. Computer-assisted keyword screening was conducted using ICD-9 codes. We manually screened this database for potential cases. We performed case identification in two phases. First, we created a computer-assisted keyword screening tool called regular-expression matching to search the electronic medical record and to identify potentially eligible ED encounters. This technique provides a more comprehensive and inclusive search than keyword searching by including misspelled and mistyped variations. Second, we refined the output of the search tool by manual medical chart review. 2.3. Patient population We included children aged 6 to 60 months evaluated in the ED after their first CFS who had both an EEG and a minimum follow-up of 2 years. The index febrile seizure was the first CFS. We excluded children with known severe neurologic disability, central nervous system infection/insult, history of unprovoked seizure at baseline, and the presence of acute electrolyte imbalance at the time of CFS. The presence of a prior simple febrile seizure (SFS) or mild developmental delay (as assessed by a neurologist after the CFS) was not exclusionary.
201
time to develop epilepsy, family history of epilepsy or unprovoked seizures in first or second degree relatives, perinatal history, past medical history, treatment with antiepileptic medication (AED), imaging result (if available), and developmental history as documented by the treating physician after the initial CFS. Most of the patients (except for 11 patients) were seen by a neurologist. 2.7. Outcome measure The primary outcome was epilepsy, defined as the occurrence of at least two episodes of unprovoked seizures N24 h apart. 2.8. Statistical analysis Established clinical predictors (for subsequent epilepsy in children with FS) and EEG status of children with a first CFS were evaluated for potential correlation with the development of epilepsy in the cohort by univariate analysis using Fisher's exact test. The sensitivity and specificity of epileptiform EEG in predicting epilepsy were calculated using descriptive statistics. Applying the sensitivity/specificity results, positive predictive value (PPV) of epileptiform EEG and negative predicative value (NPV) of nonepileptiform EEG were calculated for the estimated risk of developing epilepsy or prevalence of epilepsy in children with a history of complex febrile seizures (around 10–20% [15]). All analyses were conducted using the Statistical Package for the Social Sciences IBM Corp. Released 2012. IBM SPSS Statistics for Windows, Version 21.0. Armonk, NY: IBM Corp.
2.4. Definitions 3. Results Complex febrile seizure was defined as seizure activity occurring with a temperature of ≥ 38.4 °C and included either focal features, prolonged duration (≥15 min), recurrence within 24 h, or a combination thereof. Status epilepticus (SE) was defined as single seizure or multiple seizures without recovery lasting ≥30 min. Epilepsy was defined as the occurrence of at least two episodes of unprovoked seizures N24 h apart.
There were 720 children evaluated in the ED for their first CFS of which 154 children (45% females) met the entry criteria (Fig. 1). Patients were excluded because of lack of 2-year follow-up (46%) or no EEG data available (19%) in the 6-month period following presentation. The index CFS was the first FS in 88% of children. Median age at the first CFS was 1.5 years (0.5–4.58 years). Median duration of follow-up for all patients was 6.3 years (interquartile range/IQR: 3.8–10.4 years),
2.5. Procedure Report from the initial EEG performed as a part of evaluation of the first CFS was utilized for analysis. Patients with EEG done 6 or more months after their first CFS were excluded. We compared the clinical/ EEG characteristics between the groups of patients with a first CFS who developed epilepsy to those who did not develop epilepsy at follow-up. Electroencephalogram characteristics were dichotomized as epileptiform and nonepileptiform. Epileptiform abnormalities consisted of spikes, sharp waves, or spike–wave complexes, and were noted as focal or generalized. Nonepileptiform EEG abnormalities such as slowing and focal asymmetry (defined as depression of cerebral activity that is lateralized or regional, of any frequency with a difference of amplitude of ≥50% between the sides) were also noted. Established clinical predictors of epilepsy following febrile seizures (number of complex features, family history of epilepsy, developmental delay) were chosen based on the existing literature [4,7,8]. The physician record from the final encounter with the patient was used to evaluate for the development of epilepsy or single unprovoked seizure. Electroencephalographic studies were performed using the 10–20 international system with bipolar and referential montages. Activation procedures were performed if possible. Sleep was obtained if possible. Sedation for EEG was not used.
Computer based search
Patients screened for CFS
N= 720
Patients with unprovoked seizure N = 18 First time CFS with 2 year FU
N = 371
No EEG or Late EEG* N = 139
First time CFS with EEG/FU
N = 232
Excluded patients
Manual review
N= 78
Provoked seizures**
33
Other reasons ***
45
Included patients
N=
154
2.6. Data collection Data collected from medical records included demographics, seizure characteristics, number of complex features, duration of fever, EEG report, occurrence of epilepsy or unprovoked seizure, type of epilepsy,
Fig. 1. Flow sheet of patients screened, excluded, and reasons for exclusion. *Late EEG—EEG obtained N6 months from CFS. **Other provoked seizures causes include SDH, meningitis, encephalitis, head trauma, and metabolic abnormalities. ***Other reasons: SFS—5, developmental delay—4, inadequate data—21, no fever—1, no seizures—5, febrile myoclonus—6, and not a first CFS—3.
202
C. Harini et al. / Epilepsy & Behavior 52 (2015) 200–204
Table 1 Clinical and EEG data.
Table 3 PPV and NPV of EEG for development of epilepsy.
Variables
Gender Age in years at CFS (median) [IQR] Abnormal MRI History No. of complex features 1 vs N1 Status epilepticus FH of epilepsy in 1st & 2nd degree relativesa Developmental delay Prior SFS EEG Findings Epileptiform EEG Slowing on EEG
Patients who developed epilepsy (n = 20)
Patients who did not develop epilepsy (n = 134)
p value
8/20 (40%) 1.55 (0.18–2.75)
62/134 (46%) 1.4 (1.1–1.9)
0.6 0.41
3/17c
11/65c
17/20 (85%) 6/20 (30%) 6/20 (30%)
95/134 (71%) 27/134 (20.1%) 27/134 (20.1%)
0.29 0.38 0.38
4/20 (20%) 3/20 (15%)
18/134 (13.4%) 16/134 (12.2%)
0.49 0.72
4b/20 (20%) 3/20 (15%)
17/134 (12.7%) 20/134 (14.9%)
0.48 1.0
a
FH unavailable in 19 patients. 1 patient with epilepsy had epileptiform activity and slowing on EEG. c Not all patients had MRI—abnormalities included periventricular leucomalacia, gliosis, heterotopia, delayed myelination, cyst, and others.
Risk of epilepsy among children with FS
PPV of epileptiform EEG (confidence intervals)
NPV of nonepileptiform EEG (confidence intervals)
13%a 10%b
19% (6.3–42.6%) 15% (4–39%)
88% (81–93%) 91% (85–95%)
PPV—positive predictive value, NPV—negative predictive value, FS—febrile seizures. a From our cohort. b From prospective studies of febrile seizures.
with epilepsy had nonepileptiform abnormalities with slowing (focal/ bilateral) with or without asymmetry. The electroclinical syndrome and epilepsy classification are given in Table 4. The time to develop epilepsy was not clearly stated in two patients. In the rest, epilepsy developed after a median duration of 5 months (IQR: 3–14 months) following the first CFS. 3.3. Occurrence of single unprovoked seizure
b
and for children with subsequent epilepsy, median duration was 7 years (IQR: 3.7–13.8 years). Median age at follow-up was 10 years (IQR: 5.1– 12.4 years). Eighty-five percent of the patients had ≥3 years of followup. Refer to Table 1 for clinical and EEG data.
A single unprovoked seizure occurred in 6 (3.9%) patients with a median duration of follow-up of 9.5 years (4.4–13.5 years). Four patients had generalized tonic–clonic seizures, one had a focal dyscognitive seizure, and one had a single seizure, unclassified. Four patients had a single event, 3 patients had a brief staring spell, and another had a confusional spell which could not be further characterized, and the physician did not make a diagnosis of seizure in these patients. 3.4. EEG results
3.1. Descriptive statistics Status epilepticus (SE) accounted for 33 patients (21.4%) of the cohort. In 6 patients, the duration was at least 30 min, but information regarding the exact duration of the SE was not available from the records. In the remainder, the median duration of SE was 40 min (IQR: 30–56 min). Timing of EEG was similar between both groups (with and without subsequent epilepsy) and was obtained at a median of 3 days (0– 131 days). Most patients (n = 109) had their EEG within 2 weeks of the CFS; 22 patients had EEG between 2 weeks and 1 month of CFS. Twenty-three patients had EEG after 1 month. The past medical history was significant for 4 children being a product of twin gestation and 8 children with history of prematurity between 32 and 36 weeks except for 1 patient born at 25 weeks gestation. Other systemic diseases such as asthma, vesicoureteral reflux, and congenital cardiac defects were noted in some patients (n = 9). Mild developmental delay including mild language delay and/or motor delay was noted in 22 patients (14%). Six patients, who appeared to have mild delays at the time of CFS, later developed mild verbal apraxia (n = 1), learning disorder (n = 1), intellectual disability (n = 3), and autism (n = 1). 3.2. Development of epilepsy Overall, 20 (13%) children developed epilepsy, and the prevalence of epilepsy was 13% (CI 8.3–19.6%) in our cohort. Epileptiform activity was noted in 4 patients who developed epilepsy (all with focal spikes, parietal in 2, one each in frontocentral and midline region). Three patients Table 2 Sensitivity/specificity of epileptiform EEG for predicting epilepsy.
Forty-two patients (27%) had an abnormal EEG that included slowing, epileptiform discharges, asymmetry of cerebral activity, or a combination thereof. Epileptiform abnormalities were noted in 21 patients (13.6%) with focal epileptiform activity (n = 18) and generalized spikes (n = 4) (details given in Fig. 2). Nonepileptiform abnormalities included slowing seen in 23 patients (14.9%) (focal in 14 or generalized in 9) and focal asymmetry (n = 6). Slowing was accompanied by epileptiform activity in 3 patients and by asymmetry in 3 patients. Electroencephalogram was abnormal with focal epileptiform activity in 3 of the 8 cases with a well-defined electroclinical syndrome. Slowing was seen in 8/33 (24%) following SE. Focal slowing was seen in 4 (accompanied by focal asymmetry in 3), and the remainder had generalized slowing. Epileptiform abnormality was seen in 2 patients with SE. 3.5. Epileptiform EEG and subsequent epilepsy—sensitivity/specificity, PPV/NPV The sensitivity (20%) and specificity (87.5%) of epileptiform EEG discharges for detecting later development of epilepsy are presented in Table 2. Epileptiform abnormality had a PPV of 19% for predicting epilepsy, and a nonepileptiform EEG had a NPV of 88% for not developing epilepsy in our cohort. The prevalence of epilepsy in our cohort is skewed, since probably more children with epilepsy had follow-up than those who did not. Therefore, we also calculated the estimated PPV and NPV using estimated risk for the development of epilepsy in children with febrile seizures from population-based studies (around 10% from prospective studies). These results are presented in Table 3. Excluding SE from the cohort made little difference to the sensitivity and specificity analysis (data not shown). 3.6. Relationship between clinical/EEG data to subsequent epilepsy
EEG
Epilepsy (yes)
Epilepsy (No)
Sensitivity
Specificity
Epileptiform Nonepileptiform
4 16
17 117
20% (CI 6.5–44.5%) –
– 87.5% (CI 80–92%)
Four of the 20 patients (20%) with epilepsy had epileptiform discharges versus 17 of 134 patients without epilepsy (12.7%) (p = 0.48). Electroencephalogram abnormalities (epileptiform discharges or slowing/asymmetry) did not predict subsequent epilepsy.
C. Harini et al. / Epilepsy & Behavior 52 (2015) 200–204
203
Table 4 Epilepsy classification. Patients with epilepsy (total = 20)
Electroclinical syndrome or epilepsy classification
Seizure type
2 4 2 3 4 4 1
Childhood absence epilepsy Genetic epilepsy with febrile seizures plus Atypical Dravet syndrome Epilepsy with focal seizures of unknown etiology Epilepsy with generalized seizures of unknown etiology Epilepsy with focal and generalized seizures Epilepsy unclassified
Absences Generalized tonic–clonic, tonic, myoclonic, focal dyscognitive Generalized tonic, focal dyscognitive Focal dyscognitive Generalized tonic, tonic–clonic, atonic Generalized tonic–clonic, focal dyscognitive Unknown
Among those who developed epilepsy, 30% had a history of febrile SE (6/20). However, febrile SE was not a significant predictor for the development of epilepsy in our cohort (p = 0.38). None of the other clinical variables examined showed significant association with the risk of developing subsequent epilepsy. Refer to Table 1 for details. 3.7. AED treatment Treatment with daily antiepileptic medication (AED) was initiated in 31 patients following CFS, and 4 of these patients stopped taking their AED in less than 2 weeks. Of the 27 patients who continued on AED, 19 were treated with phenobarbital, 4 with carbamazepine (started because of agitation with phenobarbital or because of focality with febrile seizure), and 3 with phenytoin. Two patients stopped phenobarbital because of drug rash, 2 others changed phenobarbital because of drug rash. Children were treated with AED for a median of 1 year (2 months–5.6 years). Epilepsy (n = 7) or single unprovoked seizure (n = 1) developed in 8 of the patients treated with AED for CFS. Antiepileptic medication treatment was initiated in 15 patients either because of EEG abnormality (epileptiform and/or nonepileptiform abnormalities) alone or in combination with SE or recurrent FS. Two patients with epileptiform EEG and 2 others with nonepileptiform abnormalities who received AED treatment developed epilepsy. The reminder of patients with EEG abnormalities (n = 11) were followed for a minimum period of 5 years (except for 1 patient with 2-year follow-up) following discontinuation of AED treatment and did not develop epilepsy. 4. Discussion We identified a large cohort of neurologically healthy or mildly delayed children with first time complex febrile seizures who had both an EEG and a minimum of 2-year follow-up (median 6.3 years). Overall, 20 (13%) children developed epilepsy. In our cohort, an epileptiform
Location of Epileptiform Discharges 5
*
*
No. of patients.
4 3 2 1 0
Other: Temporal (1), Occipital (1), Centrotemporal (1), Multiregional (1), *1 patient had both generalized and multi regional spikes
Fig. 2. Location of Epileptiform Discharges.
EEG was not a sensitive measure [sensitivity of 20% (CI 6.5–44.5%)] and had a poor positive predictive value [estimated PPV of 19% (CI 6.3–42.6%)] in detecting the development of subsequent epilepsy in the follow-up period. Thus, the presence of an epileptiform EEG does not change the risk of developing epilepsy significantly (around 15%), given that the prior probability of developing epilepsy following CFS is around 10–20%. Additionally, data from our cohort did not support previous reports suggesting a correlation between the number of complex features and development of epilepsy [7]. Our study indicated that a nonepileptiform EEG was reassuring with most of these patients (88%) not developing epilepsy. However, published literature suggests that children with CFS have a 80–90% chance of not developing epilepsy [15]. Hence, the reassurance of a nonepileptiform EEG did not provide additional information beyond what is already known about the risk of developing epilepsy in CFS. The traditional clinical risk factors (family history of epilepsy, multiple complex features) [7–9,16] for subsequent epilepsy in patients with febrile seizures were not associated with an increased risk of epilepsy in our study. We considered complex features of the first CFS as opposed to complex features cumulated across recurrent febrile seizures [4]. Also, the family history of epilepsy was unavailable in a significant minority (12%) of our patients. These factors potentially influenced our findings. In the present study, febrile SE was not identified as a risk factor for epilepsy or mesial temporal sclerosis, which may be related to shorter duration of follow-up [17]. In previous studies, epileptiform abnormalities were rarely seen in early EEG in patients with FS, with detection rates ranging between 1.4 and 8.6% [20,23]. However, there is a report of increased likelihood of finding epileptiform abnormalities on an early EEG (b 7 days postictus) [24]. Abnormal EEG was noted in 27% of our patients (epileptiform abnormality in 13.6%) which is higher than the abovementioned studies, but similar to the rate reported in some retrospective cohorts with FS [12,13] (14–22%). Most children in our series had early EEG. Our cohort was derived from a tertiary medical center and may have selected more severe cases of CFS. The case selection rather than timing of EEG perhaps impacted our observation. Our results show that epileptiform abnormalities after the first CFS did not predict epilepsy (epileptiform EEG in 20% of patients with epilepsy versus 13% (17/134) of patients without epilepsy, p = 0.48) which is in agreement with the findings from the longitudinal study of EEG in FS [23]. Our study findings are in contrast with findings from some recent retrospective studies, where epileptiform discharges were noted to be a significant risk factor for subsequent epilepsy [10, 12,13]. Their conclusions were influenced by a number of factors including subjects with any type of FS (not purely CFS) [10,12], a mean age at FS of N2 years [10,12,13] (older than our cohort), and varying duration of follow-up of the control versus the study group [10]. Although epileptiform discharges were identified as a risk factor, the PPV was only 31% [13] (PPV—19% in our study). The frontal location of the epileptiform discharges was associated with increased risk of epilepsy in one [12] but not in the other [10] study. Complex febrile seizures may be the initial presentation of certain rare epilepsy syndromes such as Dravet syndrome [19], PCDH19-related epilepsy [18], and genetic epilepsy with febrile seizures, and therefore,
204
C. Harini et al. / Epilepsy & Behavior 52 (2015) 200–204
some may consider getting an EEG following the initial CFS. Many of the patients with these epileptic syndromes can have initial normal EEG. The diagnosis of these syndromes is first based on clinical criteria, and some of these syndromes are confirmed with genetic testing. As the role of genetic testing in children with febrile seizures/epilepsy is refined, the indications for EEG based on genetic diagnosis and age will be clarified. Electroencephalogram abnormalities can be seen intermittently, and a single recording can be subject to sampling error. There may be social implication of notifying parents about an abnormal EEG. Epileptiform EEG may prompt AED therapy. There is evidence that early intervention with standard AEDs does not interfere with the process of long-term epileptogenesis [22]. There are side effects of AEDs. Hence Epileptiform EEG does not necessarily imply AED therapy. The poor predictive value of epileptiform EEG for predicting epilepsy in the short term follow-up, as shown in our population, does not lend support to the practice of obtaining a routine EEG in neurologically healthy or mildly delayed children following a first CFS. The FEBSTAT study, which is a prospective study being conducted in children with febrile status epilepticus, will shed light on the role of EEG in febrile status epilepticus [21]. In other situations, EEG after the first CFS can be considered for acute clinical indications (persistent seizures or altered mental status) or for research purposes to study the role of EEG as a biomarker for outcome. 5. Limitations Our study has the limitations inherent to any retrospective series. Patients with inadequate data can introduce bias. The prevalence of epilepsy in children with FS increases with longer duration of follow-up, and including patients with shorter duration of follow-up can affect the results. On the other hand, our hospital-based study may have selected more severe forms of CFS, especially those who developed epilepsy, thereby introducing selection bias leading to overestimated prevalence of epilepsy. Timing of acquisition of EEG (early versus late) and age of the patient at the time of EEG can affect the results. However, we wanted to be pragmatic and used the EEG that was obtained for the initial evaluation/clinical decision-making following the first CFS. Finally, bias resulting from unknown confounders may have affected our results. 6. Conclusions We found that epileptiform abnormalities seen on an initial EEG after a first CFS were a poor predictor of subsequent epilepsy. Electroencephalogram slowing was also not a significant predictor for epilepsy. Thus, early EEG abnormalities after a first CFS are not likely to identify patients at risk for epilepsy. There are probably multiple factors influencing the occurrence of abnormalities seen on the EEG including age of the patient, timing of EEG, and genetic syndrome. The relationship between EEG abnormalities and subsequent epilepsy in febrile SE is being studied. We are in need of prospective studies in children with CFS with EEG obtained at specified time periods to help determine whether EEG can be a biomarker in terms of predicting outcomes, including the development of recurrent unprovoked seizures. Until the time that more definitive answers are available, the practice of obtaining routine EEG in patients with a first CFS may not be warranted in most circumstances. Conflict of interest statement Tobias Loddenkemper serves on the Laboratory Accreditation Board for Long Term (Epilepsy and Intensive Care Unit) Monitoring, on the council (and as treasurer) of the American Clinical Neurophysiology Society, on the American Board of Clinical Neurophysiology, as an associate Editor for Seizure, as contributing editor for Epilepsy Currents, and as an associate editor for Wyllie’s Treatment of Epilepsy 6th edition.
He is part of pending patent applications to detect and predict seizures and to diagnose epilepsy. He receives research support from the Epilepsy Research Foundation, the American Epilepsy Society, the Epilepsy Foundation of America, the Epilepsy Therapy Project, PCORI, the Pediatric Epilepsy Research Foundation, Cure, and HHV-6 Foundation and received research grants from Lundbeck, Eisai, Upsher-Smith, and Pfizer. He serves on a scientific advisory board for Upsher Smith. He performs long-term video electroencephalogram and ICU monitoring, electroencephalograms, and other electrophysiological studies at Boston Children's Hospital and affiliated hospitals and bills for these procedures, and he evaluates pediatric neurology patients and bills for clinical care. He has received speaker honorariums from national societies including the AAN, AES, and ACNS and for grand rounds at various academic centers. His wife, Dr. Karen Stannard, is a pediatric neurologist, and she performs long-term video electroencephalogram and ICU monitoring, electroencephalograms, and other electrophysiological studies and bills for these procedures, and she evaluates pediatric neurology patients and bills for clinical care. The rest of the authors have no COI.
References [1] Nelson KB, Ellenberg JH. Predisposing and causative factors in childhood epilepsy. Epilepsia 1987;28(Suppl. 1):S16–24. [2] Verity CM, Butler NR, Golding J. Febrile convulsions in a national cohort followed up from birth. I: Prevalence and recurrence in the first five years of life. Br Med J (Clin Res Ed) 1985;290:1307–10. [3] Hauser WA. The prevalence and incidence of convulsive disorders in children. Epilepsia 1994;35(Suppl. 2):S1–6. [4] Berg AT, Shinnar S. Complex febrile seizures. Epilepsia 1996;37:126–33. [5] 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. [6] Nelson KB, Ellenberg JH. Predictors οf epilepsy in children who have experienced febrile seizures. N Engl J Med 1976;295:1029–32. [7] Annegers JF, Hauser WA, Elveback LR, Kurland LT. The risk of epilepsy following febrile convulsions. Neurology 1979;29:297–303. [8] 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. [9] Verity CM, Golding J. Risk of epilepsy after febrile convulsions. A national cohort study. Br Med J 1991;303:1373–6. [10] Neligan A, Bell GS, Giavasi C, Johnson AL, Goodridge DM, Shovron SD, et al. Longterm risk of developing epilepsy after febrile seizures: a prospective cohort study. Neurology 2012;78:1166–70. [11] Kim H, Byun S-H, Kim J-S, Lim B-C, Chae J-H, Choi J, et al. Clinical and EEG risk factors for subsequent epilepsy in patients with complex febrile seizures. Epilepsy Res 2013;105:158–63. [12] Sadleir LG, Scheffer IE. Febrile seizures. BMJ 2007;334:307–11. [13] Waruiru C, Appleton R. Febrile seizures: an update. Arch Dis Child 2004;89(8): 751–6. [14] Kanemura H, Mizoroqi S, Aoyaqi K, Suqita K, Aihara M. EEG characteristics predict subsequent epilepsy in children with febrile seizure. Brain Dev 2012;34:302–7. [15] Chungath M, Shorvon S. The mortality and morbidity of febrile seizures. Nat Clin Pract Neurol 2008;4:610–21. [16] Berg AT, Shinnar S. Unprovoked seizures in children with febrile seizures: short term outcome. Neurology 1996;47:562–8. [17] Hesdorffer DC, Shinnar S, Lewis DV, Moshé SL, Nordli Jr DR, Pellock JM, et al. Design and phenomenology of the FEBSTAT study. Epilepsia 2012;53:1471–80. [18] van Harssel JJ, Weckhuysen S, van Kempen MJ, Hardies K, Verbeek NE, de Kovel CG, et al. Clinical and genetic aspects of PCDH19-related epilepsy syndromes and the possible role of PCDH19 mutations in males with autism spectrum disorders. Neurogenetics 2013;14(1):23–34. [19] Brunklaus A, Zuberi SM. Dravet syndrome—from epileptic encephalopathy to channelopathy. Epilepsia 2014;55(7):979–84. [20] Maytal J, Steele R, Eviatar L, Novak G. The value of early postictal EEG in children with complex febrile seizures. Epilepsia 2000;41(2):219–21. [21] Nordli Jr DR, Moshe SL, Shinnar S, Hesdorffer DC, Sogawa Y, Pellock JM, et al. Acute EEG findings in children with febrile status epilepticus: results of the FEBSTAT study. Neurology 2012;79(22):2180–6. [22] Marson A, Jacoby A, Johnson A, Kim L, Gamble C, Chadwick D. Immediate versus deferred antiepileptic drug treatment for early epilepsy and single seizures: a randomised controlled trial. Lancet 2005;365(9476):2007–13. [23] Frantzen E, Lennox-Buchthal M, Nygaard A. Longitudinal EEG and clinical study of children with febrile convulsions. Electroencephalogr Clin Neurophysiol 1968;24: 197–212. [24] Joshi C, Wawrykow T, Patrick J, Prasad A. Do clinical variables predict an abnormal EEG in patients with complex febrile seizures? Seizure 2005;14(6):429–34.
Contents lists available at ScienceDirect
Epilepsy & Behavior journal homepage: www.elsevier.com/locate/yebeh
Brief Communication
Utility of initial EEG in first complex febrile seizure Chellamani Harini a,⁎, Elanagan Nagarajan a, Amir A. Kimia b, Rachel Marin de Carvalho a, Sookee An a, Ann M. Bergin a, Masanori Takeoka a, Phillip L. Pearl a, Tobias Loddenkemper a a b
Division of Epilepsy and Clinical Neurophysiology, Department of Neurology, Boston Children's Hospital, Boston, MA, USA Division of Emergency Medicine, Boston Children's Hospital, Boston, MA, USA
a r t i c l e
i n f o
Article history: Received 3 August 2015 Revised 1 September 2015 Accepted 3 September 2015 Available online xxxx Keywords: Complex febrile seizures EEG Epilepsy Diagnosis
a b s t r a c t Objective: The risk of developing epilepsy following febrile seizures (FS) varies between 2% and 10%, with complex febrile seizures (CFS) having a higher risk. We examined the utility of detected epileptiform abnormalities on the initial EEG following a first CFS in predicting subsequent epilepsy. Methods: This was a retrospective study of consecutive patients (ages 6–60 months) who were neurologically healthy or mildly delayed, seen in the ED following a first CFS and had both an EEG and minimum of 2-year follow-up. Data regarding clinical characteristics, EEG report, development of subsequent epilepsy, and type of epilepsy were collected. Established clinical predictors for subsequent epilepsy in children with FS and EEG status were evaluated for potential correlation with the development of subsequent epilepsy. Sensitivity, specificity, and positive and negative predictive values of an abnormal EEG (epileptiform EEG) were calculated. Results: A group of 154 children met our inclusion criteria. Overall, 20 (13%) children developed epilepsy. The prevalence of epilepsy was 13% (CI 8.3–19.6%). Epileptiform abnormalities were noted in 21 patients (13.6%), EEG slowing in 23 patients (14.9%), and focal asymmetry in six (3.8%). Epileptiform EEGs were noted in 20% (4/20) of patients with epilepsy and 13% (17/134) of patients without epilepsy (p = 0.48). At an estimated risk of subsequent epilepsy of 10% (from population-based studies of children with FS), we determined that the PPV of an epileptiform EEG for subsequent epilepsy was 15%. None of the clinical variables (presence of more than 1 complex feature, family history of epilepsy, or status epilepticus) predicted epilepsy. Conclusions: An epileptiform EEG was not a sensitive measure and had a poor positive predictive value for the development of epilepsy among neurologically healthy or mildly delayed children with a first complex febrile seizure. The practice of obtaining routine EEG for predicting epilepsy after the first CFS needs clarification by welldefined prospective studies. © 2015 Elsevier Inc. All rights reserved.
1. Introduction Febrile seizures (FS) occur in 2–5% of children before the age of 5 years [1,2] with higher incidence rates of 8% in Japan and 14% in Guam [3]. Most febrile seizures are simple febrile seizures (SFS), and about 25–35% of FS are complex febrile seizures (CFS) [4,5]. The risk of developing epilepsy in children with febrile seizures varies between 2% and 10% depending on the length of follow-up and type of FS [7–9, 11]. Complex febrile seizures have a higher risk for developing epilepsy when compared with SFS, especially in CFS with multiple complex features [7] or occurring in the setting of neurodevelopmental delay [8]. The American Academy of Pediatrics guideline states that electroencephalogram (EEG) should not be performed in the evaluation of a neurologically healthy child with a SFS as there is no evidence that the EEG abnormality will predict recurrence of FS and/or the development of epilepsy [6]. No such guidelines exist for CFS. It is uncertain whether an ⁎ Corresponding author at: 300, Longwood Avenue, Boston MA-02115, USA. E-mail address: [email protected] (C. Harini).
http://dx.doi.org/10.1016/j.yebeh.2015.09.003 1525-5050/© 2015 Elsevier Inc. All rights reserved.
abnormal EEG after a first CFS provides predictive value for subsequent epilepsy. Some authors find no rationale in doing an EEG in children with CFS because of the paucity of evidence [14]. More recently, a few studies have reported that epileptiform discharges noted with CFS is a risk factor for the development of epilepsy [10,12,13]. However such findings may not alter clinical management. We examined the utility of epileptiform abnormalities (seen on an initial EEG) following the first CFS in predicting the subsequent development of epilepsy with a minimum follow-up of two years. Our secondary goal was to identify if clinical or other EEG characteristics predicted the development of epilepsy in children with a first CFS. 2. Methods 2.1. Study design This was a retrospective chart review of consecutive patients who were evaluated following the first episode of CFS at an urban tertiary care pediatric ED between 1996 and 2011.
C. Harini et al. / Epilepsy & Behavior 52 (2015) 200–204
The study was approved by the Boston Children's Hospital institutional review board. 2.2. Patient identification This was done in two steps. Computer-assisted keyword screening was conducted using ICD-9 codes. We manually screened this database for potential cases. We performed case identification in two phases. First, we created a computer-assisted keyword screening tool called regular-expression matching to search the electronic medical record and to identify potentially eligible ED encounters. This technique provides a more comprehensive and inclusive search than keyword searching by including misspelled and mistyped variations. Second, we refined the output of the search tool by manual medical chart review. 2.3. Patient population We included children aged 6 to 60 months evaluated in the ED after their first CFS who had both an EEG and a minimum follow-up of 2 years. The index febrile seizure was the first CFS. We excluded children with known severe neurologic disability, central nervous system infection/insult, history of unprovoked seizure at baseline, and the presence of acute electrolyte imbalance at the time of CFS. The presence of a prior simple febrile seizure (SFS) or mild developmental delay (as assessed by a neurologist after the CFS) was not exclusionary.
201
time to develop epilepsy, family history of epilepsy or unprovoked seizures in first or second degree relatives, perinatal history, past medical history, treatment with antiepileptic medication (AED), imaging result (if available), and developmental history as documented by the treating physician after the initial CFS. Most of the patients (except for 11 patients) were seen by a neurologist. 2.7. Outcome measure The primary outcome was epilepsy, defined as the occurrence of at least two episodes of unprovoked seizures N24 h apart. 2.8. Statistical analysis Established clinical predictors (for subsequent epilepsy in children with FS) and EEG status of children with a first CFS were evaluated for potential correlation with the development of epilepsy in the cohort by univariate analysis using Fisher's exact test. The sensitivity and specificity of epileptiform EEG in predicting epilepsy were calculated using descriptive statistics. Applying the sensitivity/specificity results, positive predictive value (PPV) of epileptiform EEG and negative predicative value (NPV) of nonepileptiform EEG were calculated for the estimated risk of developing epilepsy or prevalence of epilepsy in children with a history of complex febrile seizures (around 10–20% [15]). All analyses were conducted using the Statistical Package for the Social Sciences IBM Corp. Released 2012. IBM SPSS Statistics for Windows, Version 21.0. Armonk, NY: IBM Corp.
2.4. Definitions 3. Results Complex febrile seizure was defined as seizure activity occurring with a temperature of ≥ 38.4 °C and included either focal features, prolonged duration (≥15 min), recurrence within 24 h, or a combination thereof. Status epilepticus (SE) was defined as single seizure or multiple seizures without recovery lasting ≥30 min. Epilepsy was defined as the occurrence of at least two episodes of unprovoked seizures N24 h apart.
There were 720 children evaluated in the ED for their first CFS of which 154 children (45% females) met the entry criteria (Fig. 1). Patients were excluded because of lack of 2-year follow-up (46%) or no EEG data available (19%) in the 6-month period following presentation. The index CFS was the first FS in 88% of children. Median age at the first CFS was 1.5 years (0.5–4.58 years). Median duration of follow-up for all patients was 6.3 years (interquartile range/IQR: 3.8–10.4 years),
2.5. Procedure Report from the initial EEG performed as a part of evaluation of the first CFS was utilized for analysis. Patients with EEG done 6 or more months after their first CFS were excluded. We compared the clinical/ EEG characteristics between the groups of patients with a first CFS who developed epilepsy to those who did not develop epilepsy at follow-up. Electroencephalogram characteristics were dichotomized as epileptiform and nonepileptiform. Epileptiform abnormalities consisted of spikes, sharp waves, or spike–wave complexes, and were noted as focal or generalized. Nonepileptiform EEG abnormalities such as slowing and focal asymmetry (defined as depression of cerebral activity that is lateralized or regional, of any frequency with a difference of amplitude of ≥50% between the sides) were also noted. Established clinical predictors of epilepsy following febrile seizures (number of complex features, family history of epilepsy, developmental delay) were chosen based on the existing literature [4,7,8]. The physician record from the final encounter with the patient was used to evaluate for the development of epilepsy or single unprovoked seizure. Electroencephalographic studies were performed using the 10–20 international system with bipolar and referential montages. Activation procedures were performed if possible. Sleep was obtained if possible. Sedation for EEG was not used.
Computer based search
Patients screened for CFS
N= 720
Patients with unprovoked seizure N = 18 First time CFS with 2 year FU
N = 371
No EEG or Late EEG* N = 139
First time CFS with EEG/FU
N = 232
Excluded patients
Manual review
N= 78
Provoked seizures**
33
Other reasons ***
45
Included patients
N=
154
2.6. Data collection Data collected from medical records included demographics, seizure characteristics, number of complex features, duration of fever, EEG report, occurrence of epilepsy or unprovoked seizure, type of epilepsy,
Fig. 1. Flow sheet of patients screened, excluded, and reasons for exclusion. *Late EEG—EEG obtained N6 months from CFS. **Other provoked seizures causes include SDH, meningitis, encephalitis, head trauma, and metabolic abnormalities. ***Other reasons: SFS—5, developmental delay—4, inadequate data—21, no fever—1, no seizures—5, febrile myoclonus—6, and not a first CFS—3.
202
C. Harini et al. / Epilepsy & Behavior 52 (2015) 200–204
Table 1 Clinical and EEG data.
Table 3 PPV and NPV of EEG for development of epilepsy.
Variables
Gender Age in years at CFS (median) [IQR] Abnormal MRI History No. of complex features 1 vs N1 Status epilepticus FH of epilepsy in 1st & 2nd degree relativesa Developmental delay Prior SFS EEG Findings Epileptiform EEG Slowing on EEG
Patients who developed epilepsy (n = 20)
Patients who did not develop epilepsy (n = 134)
p value
8/20 (40%) 1.55 (0.18–2.75)
62/134 (46%) 1.4 (1.1–1.9)
0.6 0.41
3/17c
11/65c
17/20 (85%) 6/20 (30%) 6/20 (30%)
95/134 (71%) 27/134 (20.1%) 27/134 (20.1%)
0.29 0.38 0.38
4/20 (20%) 3/20 (15%)
18/134 (13.4%) 16/134 (12.2%)
0.49 0.72
4b/20 (20%) 3/20 (15%)
17/134 (12.7%) 20/134 (14.9%)
0.48 1.0
a
FH unavailable in 19 patients. 1 patient with epilepsy had epileptiform activity and slowing on EEG. c Not all patients had MRI—abnormalities included periventricular leucomalacia, gliosis, heterotopia, delayed myelination, cyst, and others.
Risk of epilepsy among children with FS
PPV of epileptiform EEG (confidence intervals)
NPV of nonepileptiform EEG (confidence intervals)
13%a 10%b
19% (6.3–42.6%) 15% (4–39%)
88% (81–93%) 91% (85–95%)
PPV—positive predictive value, NPV—negative predictive value, FS—febrile seizures. a From our cohort. b From prospective studies of febrile seizures.
with epilepsy had nonepileptiform abnormalities with slowing (focal/ bilateral) with or without asymmetry. The electroclinical syndrome and epilepsy classification are given in Table 4. The time to develop epilepsy was not clearly stated in two patients. In the rest, epilepsy developed after a median duration of 5 months (IQR: 3–14 months) following the first CFS. 3.3. Occurrence of single unprovoked seizure
b
and for children with subsequent epilepsy, median duration was 7 years (IQR: 3.7–13.8 years). Median age at follow-up was 10 years (IQR: 5.1– 12.4 years). Eighty-five percent of the patients had ≥3 years of followup. Refer to Table 1 for clinical and EEG data.
A single unprovoked seizure occurred in 6 (3.9%) patients with a median duration of follow-up of 9.5 years (4.4–13.5 years). Four patients had generalized tonic–clonic seizures, one had a focal dyscognitive seizure, and one had a single seizure, unclassified. Four patients had a single event, 3 patients had a brief staring spell, and another had a confusional spell which could not be further characterized, and the physician did not make a diagnosis of seizure in these patients. 3.4. EEG results
3.1. Descriptive statistics Status epilepticus (SE) accounted for 33 patients (21.4%) of the cohort. In 6 patients, the duration was at least 30 min, but information regarding the exact duration of the SE was not available from the records. In the remainder, the median duration of SE was 40 min (IQR: 30–56 min). Timing of EEG was similar between both groups (with and without subsequent epilepsy) and was obtained at a median of 3 days (0– 131 days). Most patients (n = 109) had their EEG within 2 weeks of the CFS; 22 patients had EEG between 2 weeks and 1 month of CFS. Twenty-three patients had EEG after 1 month. The past medical history was significant for 4 children being a product of twin gestation and 8 children with history of prematurity between 32 and 36 weeks except for 1 patient born at 25 weeks gestation. Other systemic diseases such as asthma, vesicoureteral reflux, and congenital cardiac defects were noted in some patients (n = 9). Mild developmental delay including mild language delay and/or motor delay was noted in 22 patients (14%). Six patients, who appeared to have mild delays at the time of CFS, later developed mild verbal apraxia (n = 1), learning disorder (n = 1), intellectual disability (n = 3), and autism (n = 1). 3.2. Development of epilepsy Overall, 20 (13%) children developed epilepsy, and the prevalence of epilepsy was 13% (CI 8.3–19.6%) in our cohort. Epileptiform activity was noted in 4 patients who developed epilepsy (all with focal spikes, parietal in 2, one each in frontocentral and midline region). Three patients Table 2 Sensitivity/specificity of epileptiform EEG for predicting epilepsy.
Forty-two patients (27%) had an abnormal EEG that included slowing, epileptiform discharges, asymmetry of cerebral activity, or a combination thereof. Epileptiform abnormalities were noted in 21 patients (13.6%) with focal epileptiform activity (n = 18) and generalized spikes (n = 4) (details given in Fig. 2). Nonepileptiform abnormalities included slowing seen in 23 patients (14.9%) (focal in 14 or generalized in 9) and focal asymmetry (n = 6). Slowing was accompanied by epileptiform activity in 3 patients and by asymmetry in 3 patients. Electroencephalogram was abnormal with focal epileptiform activity in 3 of the 8 cases with a well-defined electroclinical syndrome. Slowing was seen in 8/33 (24%) following SE. Focal slowing was seen in 4 (accompanied by focal asymmetry in 3), and the remainder had generalized slowing. Epileptiform abnormality was seen in 2 patients with SE. 3.5. Epileptiform EEG and subsequent epilepsy—sensitivity/specificity, PPV/NPV The sensitivity (20%) and specificity (87.5%) of epileptiform EEG discharges for detecting later development of epilepsy are presented in Table 2. Epileptiform abnormality had a PPV of 19% for predicting epilepsy, and a nonepileptiform EEG had a NPV of 88% for not developing epilepsy in our cohort. The prevalence of epilepsy in our cohort is skewed, since probably more children with epilepsy had follow-up than those who did not. Therefore, we also calculated the estimated PPV and NPV using estimated risk for the development of epilepsy in children with febrile seizures from population-based studies (around 10% from prospective studies). These results are presented in Table 3. Excluding SE from the cohort made little difference to the sensitivity and specificity analysis (data not shown). 3.6. Relationship between clinical/EEG data to subsequent epilepsy
EEG
Epilepsy (yes)
Epilepsy (No)
Sensitivity
Specificity
Epileptiform Nonepileptiform
4 16
17 117
20% (CI 6.5–44.5%) –
– 87.5% (CI 80–92%)
Four of the 20 patients (20%) with epilepsy had epileptiform discharges versus 17 of 134 patients without epilepsy (12.7%) (p = 0.48). Electroencephalogram abnormalities (epileptiform discharges or slowing/asymmetry) did not predict subsequent epilepsy.
C. Harini et al. / Epilepsy & Behavior 52 (2015) 200–204
203
Table 4 Epilepsy classification. Patients with epilepsy (total = 20)
Electroclinical syndrome or epilepsy classification
Seizure type
2 4 2 3 4 4 1
Childhood absence epilepsy Genetic epilepsy with febrile seizures plus Atypical Dravet syndrome Epilepsy with focal seizures of unknown etiology Epilepsy with generalized seizures of unknown etiology Epilepsy with focal and generalized seizures Epilepsy unclassified
Absences Generalized tonic–clonic, tonic, myoclonic, focal dyscognitive Generalized tonic, focal dyscognitive Focal dyscognitive Generalized tonic, tonic–clonic, atonic Generalized tonic–clonic, focal dyscognitive Unknown
Among those who developed epilepsy, 30% had a history of febrile SE (6/20). However, febrile SE was not a significant predictor for the development of epilepsy in our cohort (p = 0.38). None of the other clinical variables examined showed significant association with the risk of developing subsequent epilepsy. Refer to Table 1 for details. 3.7. AED treatment Treatment with daily antiepileptic medication (AED) was initiated in 31 patients following CFS, and 4 of these patients stopped taking their AED in less than 2 weeks. Of the 27 patients who continued on AED, 19 were treated with phenobarbital, 4 with carbamazepine (started because of agitation with phenobarbital or because of focality with febrile seizure), and 3 with phenytoin. Two patients stopped phenobarbital because of drug rash, 2 others changed phenobarbital because of drug rash. Children were treated with AED for a median of 1 year (2 months–5.6 years). Epilepsy (n = 7) or single unprovoked seizure (n = 1) developed in 8 of the patients treated with AED for CFS. Antiepileptic medication treatment was initiated in 15 patients either because of EEG abnormality (epileptiform and/or nonepileptiform abnormalities) alone or in combination with SE or recurrent FS. Two patients with epileptiform EEG and 2 others with nonepileptiform abnormalities who received AED treatment developed epilepsy. The reminder of patients with EEG abnormalities (n = 11) were followed for a minimum period of 5 years (except for 1 patient with 2-year follow-up) following discontinuation of AED treatment and did not develop epilepsy. 4. Discussion We identified a large cohort of neurologically healthy or mildly delayed children with first time complex febrile seizures who had both an EEG and a minimum of 2-year follow-up (median 6.3 years). Overall, 20 (13%) children developed epilepsy. In our cohort, an epileptiform
Location of Epileptiform Discharges 5
*
*
No. of patients.
4 3 2 1 0
Other: Temporal (1), Occipital (1), Centrotemporal (1), Multiregional (1), *1 patient had both generalized and multi regional spikes
Fig. 2. Location of Epileptiform Discharges.
EEG was not a sensitive measure [sensitivity of 20% (CI 6.5–44.5%)] and had a poor positive predictive value [estimated PPV of 19% (CI 6.3–42.6%)] in detecting the development of subsequent epilepsy in the follow-up period. Thus, the presence of an epileptiform EEG does not change the risk of developing epilepsy significantly (around 15%), given that the prior probability of developing epilepsy following CFS is around 10–20%. Additionally, data from our cohort did not support previous reports suggesting a correlation between the number of complex features and development of epilepsy [7]. Our study indicated that a nonepileptiform EEG was reassuring with most of these patients (88%) not developing epilepsy. However, published literature suggests that children with CFS have a 80–90% chance of not developing epilepsy [15]. Hence, the reassurance of a nonepileptiform EEG did not provide additional information beyond what is already known about the risk of developing epilepsy in CFS. The traditional clinical risk factors (family history of epilepsy, multiple complex features) [7–9,16] for subsequent epilepsy in patients with febrile seizures were not associated with an increased risk of epilepsy in our study. We considered complex features of the first CFS as opposed to complex features cumulated across recurrent febrile seizures [4]. Also, the family history of epilepsy was unavailable in a significant minority (12%) of our patients. These factors potentially influenced our findings. In the present study, febrile SE was not identified as a risk factor for epilepsy or mesial temporal sclerosis, which may be related to shorter duration of follow-up [17]. In previous studies, epileptiform abnormalities were rarely seen in early EEG in patients with FS, with detection rates ranging between 1.4 and 8.6% [20,23]. However, there is a report of increased likelihood of finding epileptiform abnormalities on an early EEG (b 7 days postictus) [24]. Abnormal EEG was noted in 27% of our patients (epileptiform abnormality in 13.6%) which is higher than the abovementioned studies, but similar to the rate reported in some retrospective cohorts with FS [12,13] (14–22%). Most children in our series had early EEG. Our cohort was derived from a tertiary medical center and may have selected more severe cases of CFS. The case selection rather than timing of EEG perhaps impacted our observation. Our results show that epileptiform abnormalities after the first CFS did not predict epilepsy (epileptiform EEG in 20% of patients with epilepsy versus 13% (17/134) of patients without epilepsy, p = 0.48) which is in agreement with the findings from the longitudinal study of EEG in FS [23]. Our study findings are in contrast with findings from some recent retrospective studies, where epileptiform discharges were noted to be a significant risk factor for subsequent epilepsy [10, 12,13]. Their conclusions were influenced by a number of factors including subjects with any type of FS (not purely CFS) [10,12], a mean age at FS of N2 years [10,12,13] (older than our cohort), and varying duration of follow-up of the control versus the study group [10]. Although epileptiform discharges were identified as a risk factor, the PPV was only 31% [13] (PPV—19% in our study). The frontal location of the epileptiform discharges was associated with increased risk of epilepsy in one [12] but not in the other [10] study. Complex febrile seizures may be the initial presentation of certain rare epilepsy syndromes such as Dravet syndrome [19], PCDH19-related epilepsy [18], and genetic epilepsy with febrile seizures, and therefore,
204
C. Harini et al. / Epilepsy & Behavior 52 (2015) 200–204
some may consider getting an EEG following the initial CFS. Many of the patients with these epileptic syndromes can have initial normal EEG. The diagnosis of these syndromes is first based on clinical criteria, and some of these syndromes are confirmed with genetic testing. As the role of genetic testing in children with febrile seizures/epilepsy is refined, the indications for EEG based on genetic diagnosis and age will be clarified. Electroencephalogram abnormalities can be seen intermittently, and a single recording can be subject to sampling error. There may be social implication of notifying parents about an abnormal EEG. Epileptiform EEG may prompt AED therapy. There is evidence that early intervention with standard AEDs does not interfere with the process of long-term epileptogenesis [22]. There are side effects of AEDs. Hence Epileptiform EEG does not necessarily imply AED therapy. The poor predictive value of epileptiform EEG for predicting epilepsy in the short term follow-up, as shown in our population, does not lend support to the practice of obtaining a routine EEG in neurologically healthy or mildly delayed children following a first CFS. The FEBSTAT study, which is a prospective study being conducted in children with febrile status epilepticus, will shed light on the role of EEG in febrile status epilepticus [21]. In other situations, EEG after the first CFS can be considered for acute clinical indications (persistent seizures or altered mental status) or for research purposes to study the role of EEG as a biomarker for outcome. 5. Limitations Our study has the limitations inherent to any retrospective series. Patients with inadequate data can introduce bias. The prevalence of epilepsy in children with FS increases with longer duration of follow-up, and including patients with shorter duration of follow-up can affect the results. On the other hand, our hospital-based study may have selected more severe forms of CFS, especially those who developed epilepsy, thereby introducing selection bias leading to overestimated prevalence of epilepsy. Timing of acquisition of EEG (early versus late) and age of the patient at the time of EEG can affect the results. However, we wanted to be pragmatic and used the EEG that was obtained for the initial evaluation/clinical decision-making following the first CFS. Finally, bias resulting from unknown confounders may have affected our results. 6. Conclusions We found that epileptiform abnormalities seen on an initial EEG after a first CFS were a poor predictor of subsequent epilepsy. Electroencephalogram slowing was also not a significant predictor for epilepsy. Thus, early EEG abnormalities after a first CFS are not likely to identify patients at risk for epilepsy. There are probably multiple factors influencing the occurrence of abnormalities seen on the EEG including age of the patient, timing of EEG, and genetic syndrome. The relationship between EEG abnormalities and subsequent epilepsy in febrile SE is being studied. We are in need of prospective studies in children with CFS with EEG obtained at specified time periods to help determine whether EEG can be a biomarker in terms of predicting outcomes, including the development of recurrent unprovoked seizures. Until the time that more definitive answers are available, the practice of obtaining routine EEG in patients with a first CFS may not be warranted in most circumstances. Conflict of interest statement Tobias Loddenkemper serves on the Laboratory Accreditation Board for Long Term (Epilepsy and Intensive Care Unit) Monitoring, on the council (and as treasurer) of the American Clinical Neurophysiology Society, on the American Board of Clinical Neurophysiology, as an associate Editor for Seizure, as contributing editor for Epilepsy Currents, and as an associate editor for Wyllie’s Treatment of Epilepsy 6th edition.
He is part of pending patent applications to detect and predict seizures and to diagnose epilepsy. He receives research support from the Epilepsy Research Foundation, the American Epilepsy Society, the Epilepsy Foundation of America, the Epilepsy Therapy Project, PCORI, the Pediatric Epilepsy Research Foundation, Cure, and HHV-6 Foundation and received research grants from Lundbeck, Eisai, Upsher-Smith, and Pfizer. He serves on a scientific advisory board for Upsher Smith. He performs long-term video electroencephalogram and ICU monitoring, electroencephalograms, and other electrophysiological studies at Boston Children's Hospital and affiliated hospitals and bills for these procedures, and he evaluates pediatric neurology patients and bills for clinical care. He has received speaker honorariums from national societies including the AAN, AES, and ACNS and for grand rounds at various academic centers. His wife, Dr. Karen Stannard, is a pediatric neurologist, and she performs long-term video electroencephalogram and ICU monitoring, electroencephalograms, and other electrophysiological studies and bills for these procedures, and she evaluates pediatric neurology patients and bills for clinical care. The rest of the authors have no COI.
References [1] Nelson KB, Ellenberg JH. Predisposing and causative factors in childhood epilepsy. Epilepsia 1987;28(Suppl. 1):S16–24. [2] Verity CM, Butler NR, Golding J. Febrile convulsions in a national cohort followed up from birth. I: Prevalence and recurrence in the first five years of life. Br Med J (Clin Res Ed) 1985;290:1307–10. [3] Hauser WA. The prevalence and incidence of convulsive disorders in children. Epilepsia 1994;35(Suppl. 2):S1–6. [4] Berg AT, Shinnar S. Complex febrile seizures. Epilepsia 1996;37:126–33. [5] 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. [6] Nelson KB, Ellenberg JH. Predictors οf epilepsy in children who have experienced febrile seizures. N Engl J Med 1976;295:1029–32. [7] Annegers JF, Hauser WA, Elveback LR, Kurland LT. The risk of epilepsy following febrile convulsions. Neurology 1979;29:297–303. [8] 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. [9] Verity CM, Golding J. Risk of epilepsy after febrile convulsions. A national cohort study. Br Med J 1991;303:1373–6. [10] Neligan A, Bell GS, Giavasi C, Johnson AL, Goodridge DM, Shovron SD, et al. Longterm risk of developing epilepsy after febrile seizures: a prospective cohort study. Neurology 2012;78:1166–70. [11] Kim H, Byun S-H, Kim J-S, Lim B-C, Chae J-H, Choi J, et al. Clinical and EEG risk factors for subsequent epilepsy in patients with complex febrile seizures. Epilepsy Res 2013;105:158–63. [12] Sadleir LG, Scheffer IE. Febrile seizures. BMJ 2007;334:307–11. [13] Waruiru C, Appleton R. Febrile seizures: an update. Arch Dis Child 2004;89(8): 751–6. [14] Kanemura H, Mizoroqi S, Aoyaqi K, Suqita K, Aihara M. EEG characteristics predict subsequent epilepsy in children with febrile seizure. Brain Dev 2012;34:302–7. [15] Chungath M, Shorvon S. The mortality and morbidity of febrile seizures. Nat Clin Pract Neurol 2008;4:610–21. [16] Berg AT, Shinnar S. Unprovoked seizures in children with febrile seizures: short term outcome. Neurology 1996;47:562–8. [17] Hesdorffer DC, Shinnar S, Lewis DV, Moshé SL, Nordli Jr DR, Pellock JM, et al. Design and phenomenology of the FEBSTAT study. Epilepsia 2012;53:1471–80. [18] van Harssel JJ, Weckhuysen S, van Kempen MJ, Hardies K, Verbeek NE, de Kovel CG, et al. Clinical and genetic aspects of PCDH19-related epilepsy syndromes and the possible role of PCDH19 mutations in males with autism spectrum disorders. Neurogenetics 2013;14(1):23–34. [19] Brunklaus A, Zuberi SM. Dravet syndrome—from epileptic encephalopathy to channelopathy. Epilepsia 2014;55(7):979–84. [20] Maytal J, Steele R, Eviatar L, Novak G. The value of early postictal EEG in children with complex febrile seizures. Epilepsia 2000;41(2):219–21. [21] Nordli Jr DR, Moshe SL, Shinnar S, Hesdorffer DC, Sogawa Y, Pellock JM, et al. Acute EEG findings in children with febrile status epilepticus: results of the FEBSTAT study. Neurology 2012;79(22):2180–6. [22] Marson A, Jacoby A, Johnson A, Kim L, Gamble C, Chadwick D. Immediate versus deferred antiepileptic drug treatment for early epilepsy and single seizures: a randomised controlled trial. Lancet 2005;365(9476):2007–13. [23] Frantzen E, Lennox-Buchthal M, Nygaard A. Longitudinal EEG and clinical study of children with febrile convulsions. Electroencephalogr Clin Neurophysiol 1968;24: 197–212. [24] Joshi C, Wawrykow T, Patrick J, Prasad A. Do clinical variables predict an abnormal EEG in patients with complex febrile seizures? Seizure 2005;14(6):429–34.