- Email: [email protected]
Electroencephalogram confirmatory rate in neonatal seizures
Electroencephalogram Confirmatory Rate in Neonatal Seizures Raj D. Sheth, MD The electroencephalogram (EEG) is confirmatory in 70% of children and adults with seizures, although gestation- and etiology-specific EEG confirmatory rates in neonates have not been well defined. All neonates treated for seizures and who underwent EEG were identified from 4,575 neonates admitted between 1985 and 1996 to a neonatal intensive care unit. The relationship between EEG findings (epileptiform discharges and background abnormalities) and gestation, mortality rate, and seizure etiology was examined using the Student t test. One hundred eighty-three neonates treated for seizures underwent a total of 352 EEGs: 144 of these neonates (79%) had an abnormal EEG (epileptiform discharges in 113 (60%) and nonepileptiform background abnormalities in 31). The EEG confirmatory rate increased with gestation (63% at 28 weeks vs 77% at term, P < 0.02). Etiology for seizures also influenced the EEG confirmatory rate: central nervous system (CNS) infection 95% (P < 0.05), hypoxic-ischemic encephalopathy 80% (P < 0.05), germinal matrix-intraventricular hemorrhage 65%, and CNS malformations 65%. The EEG confirmatory rate was predictive of neonatal mortality (19% vs 6%, P < 0.03). The EEG was directly confirmatory (epileptiform discharges) in 60% and supportive (nonepileptiform background abnormalities) in a further 17% of neonates with seizures. Gestation and etiology influence the EEG confirmatory rate in neonatal seizures. © 1999 by Elsevier Science Inc. All rights reserved. Sheth RD. Electroencephalogram confirmatory rate in neonatal seizures. Pediatr Neurol 1999;20:27-30.
and often are difficult to classify [10,11]. The sick newborn often displays a variety of common paroxysmal behaviors, referred to as subtle seizures, that probably are not epileptic and may represent brainstem release phenomena that result from severe diffuse brain injury [12]. Experiences from the neonatal clinical research centers suggest that clinical observation of seizure semiology unaccompanied by electroencephalogram (EEG) confirmation has severe limitations [13]. Accordingly, the EEG plays an important role in understanding the nature of paroxysmal movements in neonates and in differentiating epileptic seizures from nonepileptic events. The presence of epileptiform discharges on the EEG is valuable in confirming a clinical suspicion of epileptic seizures. About half of all adults with seizures will have epileptiform discharges on the initial EEG [14]. This confirmatory rate increases to 75% when more than one EEG study is performed [15]. Similar confirmatory rates have been reported in children with seizures [16]. The confirmatory rate in both children and adults appears to be influenced by the duration of the EEG record, with higher confirmatory rates in EEG records of longer duration. Outside the neonatal period the etiology for seizures may be symptomatic or idiopathic; neonatal seizures are almost exclusively symptomatic. As in the older age groups, interictal epileptiform discharges can occur in the absence of documented seizures. The influences of gestation and etiology on the EEG confirmatory rate in neonates treated for seizures in the contemporary intensive care unit were examined in this study. Patients and Methods
Introduction The highest incidence of seizures occurs during the neonatal period [1-9]. Seizures in neonates can be variable
From the Department of Neurology; University of Wisconsin; Madison, Wisconsin.
© 1999 by Elsevier Science Inc. All rights reserved. PII S0887-8994(98)00078-2 ● 0887-8994/99/$20.00
Neonates admitted to a neonatal intensive care unit at a children’s hospital between January 1, 1986 and December 31, 1996 were studied. The neonatal unit (442 6 36 admissions per year) served as a regional tertiary referral center for a rural population that remained stable for the duration of the study period. The medical records of each neonate
Communications should be addressed to: Dr. Sheth; Department of Neurology; University of Wisconsin; 600 Highland Avenue H6/574 CSC; Madison, WI 53792-5132. Received April 10, 1998; accepted June 17, 1998.
Sheth: EEG Confirmatory Rate in Neonatal Seizures 27
admitted to the neonatal unit were abstracted daily to a standardized data entry form by a neonatologist or neonatal nurse practitioner and entered into a comprehensive prospectively maintained neonatal database. Data integrity was validated by a designated database supervisor at least three times weekly. In instances in which neonates were transferred from the neonatal intensive care unit, the final disposition was tracked until the neonate died or was discharged from the hospital. All neonatal EEGs performed over the 11-year period were reviewed for the presence of epileptiform discharges or electrographic seizures and background abnormalities by one of three electroencephalographers [11]. EEG recordings were performed at the bedside using a 21-channel portable system and were, typically, at least 1 hour long. Surface electrodes were applied according to the 10-20 International System, and tracings were recorded at 15 mm/s, with a recording sensitivity of 7 mV/mm. Extraocular movements, electrocardiogram, and thoracic respiratory movements were recorded simultaneously with the EEG. EEG background was evaluated with respect to gestation and sleep state. The EEG background was examined interictally and classified as normal or mildly, moderately, or severely abnormal, as previously described [17]. An electrographic seizure was defined as ictal by the appearance of sudden, repetitive, evolving stereotypical waveforms with a minimum duration of 10 seconds [18]. Rhythmic or semirhythmic sharp waves occurring repeatedly in a localized area were used to identify interictal epileptiform EEG discharges. Visual analysis of sharp transients, with respect to gestation and temporal or frontal location, was used to differentiate normal sharp transients from epileptiform discharges. An EEG was categorized as confirmatory if it demonstrated electrographic seizures or interictal epileptiform discharges. Absence of epileptiform discharges or electrographic seizures in the face of an abnormal EEG background was considered supportive of the diagnosis. Clinical seizures were identified by direct observation at the bedside by a member of the neonatal team. Neonates with clinical seizures that were treated with an antiepileptic medication were included. Postmenstrual age at birth was determined using the best obstetric estimate available (prenatal ultrasound, menstrual dates, positive pregnancy test) and corroborated by postnatal clinical assessment. Seizure etiology was approached from the perspective of neuropathologic findings and was classified into four broad neurologic categories. Neurologic disorders examined included germinal matrix-periventricular hemorrhage, hypoxic-ischemic encephalopathy, central nervous system (CNS) infections (including TORCH [toxoplasmosis, other (virus), rubella, cytomegalovirus, herpes] infections, encephalitis, and culturepositive meningitis), and cerebral dysgenesis (dysgenesis included congenital brain malformations and those chromosomal aberrations with CNS manifestations). Bivariate analysis using the Student t test for continuous variables was used to examine the relationship between the EEG findings and gestation, seizure etiology, and mortality rate.
Results One hundred eighty-three neonates (3.9% of all neonates in the intensive care unit) with seizures underwent a total of 352 EEGs. The EEG was abnormal in 144 neonates (79%) and normal in 39 neonates. Epileptiform discharges on the initial EEG were present in 84 neonates (45%). A further 29 neonates without epileptiform discharges on initial EEG demonstrated epileptiform discharges on subsequent EEG studies. Thus, of the total cohort of 183 neonates, the EEG was directly confirmatory in 113 neonates, yielding an overall EEG confirmatory rate for neonatal seizures of 60%. Of these 113 neonates, 34 had frank electrographic seizures recorded (with or without associated clinical manifestation). In addition, 84 of the 113 neonates with epileptiform
28
PEDIATRIC NEUROLOGY
Vol. 20 No. 1
Table 1. Electroencephalogram confirmatory rates by etiology in 183 neonates with seizures
Etiology
Neonates (n)
EEF Confirmatory Rate (%)
Hypoxic-ischemic encephalopathy Germinal matrix hemorrhage CNS dysgenesis CNS infection Miscellaneous
71 28 23 15 68
80 65 65 95 66
Abbreviation: CNS 5 Central nervous system
discharges had moderate or severe EEG background abnormalities, 20 had mild background abnormalities, and the remaining nine neonates had a normal EEG background for gestation. Nonepileptiform EEG background abnormalities were present in 31 neonates (17%): 14 had burst suppression, 11 had dysmaturity of the background, and six had electrocerebral silence. Gestation exerted a powerful influence on the overall EEG confirmatory rate. The confirmatory rate increased from 63% at a gestation of 28 weeks to 77% at term (P , 0.02). EEG background abnormalities also depended on gestation, such that proportionately, marked background abnormalities were more likely to be present in younger gestation neonates, and moderate and mild background abnormalities were more likely as the neonate approached term (P 5 NS). Etiology for the neonatal seizures also influenced the overall EEG confirmatory rate (Table 1). Etiology could be categorized into one of the four neurologic groups in 115 neonates (63%): 71 (50%) had hypoxic-ischemic encephalopathy, 28 (20%) had germinal matrix-intraventricular hemorrhage, 23 (16%) had cerebral dysgenesis, and 15 (10%) had a CNS infection. Twenty-two neonates had two or more neurologic insults underlying their seizures. Etiology for the remaining 68 neonates (37%) formed a miscellaneous group that included metabolic derangement, stroke, shock, and seizures of undetermined origin. The EEG confirmatory rate for neonates with seizures resulting from hypoxic-ischemic encephalopathy was 80% (P , 0.05) and 95% (P , 0.05) from CNS infection. The confirmatory rate for seizures resulting from germinal matrix-intraventricular hemorrhage or congenital malformations was 65% each (P 5 NS). Not unexpectedly, the distribution of seizure etiology was influenced heavily by gestation. Seizures were more likely to result from CNS infection or germinal matrix-intraventricular hemorrhage in preterm neonates; neonates closer to term were more likely to have seizures as a result of hypoxic-ischemic encephalopathy or CNS malformations. To define the relative contribution of etiology to the EEG confirmatory rate beyond that exerted by gestation alone, multivariate regression analysis was performed. Hypoxicischemic encephalopathy and CNS infections influenced
the EEG confirmatory rate beyond that expected by the influence of gestation alone (P , 0.05 for each). The influence of germinal matrix-intraventricular hemorrhage and CNS dysgenesis on the EEG confirmatory rate, however, appeared to result almost entirely from the influence of gestation. An etiology other than one of the four neurologic groups occurred in 68 neonates. For these neonates the overall EEG confirmatory rate was 66% (45 had an abnormal EEG and 23 a normal EEG). Trends in the overall EEG confirmatory rate over time also were examined. During the 11-year period, there was a statistically significant trend toward lower EEG confirmatory rates with time. Generally, neonates born before 1990 had an EEG confirmatory rate of 82%, and neonates born during or after 1990 had an EEG confirmatory rate of 68% (P , 0.05). A total of 29 neonates (15%) of this cohort died during their intensive care unit stay. A confirmatory EEG was predictive of mortality (19% vs 6%, P , 0.03). Discussion The overall 3.9% incidence of seizures of this cohort of 4,654 neonates during the 11-year period was similar to the incidence reported for other contemporary studies [19]. Of the 183 neonates with seizures, 77% had either epileptiform discharges or nonepileptiform background abnormalities on the EEG. The initial EEG revealed epileptiform discharges in 45% of the cohort, subsequent recordings were confirmatory in an additional 15%. Thus the overall 60% direct confirmatory rate observed in neonates was approximately 15% lower than that reported for seizures occurring in patients beyond the neonatal period [14]. Gestation exerted a strong influence on the EEG confirmatory rate. The confirmatory rate increased from approximately 63% at 28 weeks gestation to 79% at term. The lower rate observed in preterm neonates may be a reflection of behavioral movements resembling seizures but not associated with an EEG signature [13]. CNS maturation is determined by gestation such that rapid changes in the degree of neuronal organization, bioelectrical connectivity, and ictal propagation may underlie this observation [20-26]. Accordingly, the ability of a neonate who has sustained neurologic insult to manifest seizures may depend on gestation. The distribution of neurologic disorders in neonates is influenced heavily by gestation [27]. In this study, term neonates were more likely to have hypoxic-ischemic encephalopathy and CNS dysgenesis as etiologies for seizures; more preterm neonates were more likely to have germinal matrix-intraventricular hemorrhage and CNS infections as etiologies. However, etiology exerted an influence on the EEG confirmatory rate above and beyond that expected for gestation alone. The EEG confirmatory rate was 80% for hypoxic-ischemic encephalopathy and 95% for CNS infection; whereas the confirmatory rate was
65% each for germinal matrix-intraventricular hemorrhage and CNS dysgenesis. This divergence between gestation and etiology for seizures on the EEG confirmatory rate might be a reflection of the contribution of different underlying pathogenic mechanism. For those neonates in whom an etiology for seizures could not be categorized into one of the four neurologic groups the confirmatory rate was similar to the lower EEG confirmatory rate observed with germinal matrix-intraventricular hemorrhage and CNS dysgenesis. Seventeen percent of neonates without epileptiform discharges had isolated nonepileptiform background abnormalities. The absence of epileptiform discharges does not allow confidence in ascribing an epileptic nature to the events being treated. However, because the vast majority of seizures occurring in the intensive care unit are likely to be acute symptomatic rather than benign idiopathic, background abnormalities may help provide some indirect supportive evidence for a diagnosis of neonatal seizures. In the late 1970s and early 1980s a number of innocent neonatal episodic behaviors were characterized as epileptic. Prolonged video-EEG monitoring of neonates with suspected seizures between 1982 and 1986 helped identify those clinical seizure manifestations with a high degree of EEG confirmation [13]. The EEG confirmatory rates for various neonatal seizure types has been well defined [22], although it was not possible to reliably ascertain the EEG confirmatory rates by clinical seizure type in this study. The retrospective nature of this study imposes limitations. For instance, sharp transients were separated from interictal activity by their morphology, distribution, and EEG background accompaniments. However, interictal epileptiform EEG discharges in the neonate are not easily defined and may be difficult to separate from normal sharp transients, especially in preterm neonates. Furthermore, the lower confirmatory rate for seizures in neonates compared with older individuals may be a result of some neonates with seizures having a normal interictal EEG. Duration of EEG sampling also may be a factor in determining EEG confirmatory rates. Almost all neonates in this cohort had EEG recordings that lasted for at least 1 hour, although continuous EEG monitoring may have yielded higher EEG confirmatory rates. Monitoring seizures in neonates with simultaneous video-EEG in the neonatal intensive care unit is highly desirable, although not widely available [22,28]. The clinical circumstance reported in this study may thus offer some generally applicable estimates of neonatal EEG confirmatory rates. Interestingly, confirmatory rates for EEGs performed on or before 1990 were 14% higher than the rate from 1990 to the end of the study period, suggesting an increased use of EEG in the management of neonatal seizures. Routine neonatal video-EEG monitoring and a higher EEG repetition rate in neonates with a normal initial EEG perhaps may have increased the sensitive of the observed EEG confirmatory rates in neonatal seizures.
Sheth: EEG Confirmatory Rate in Neonatal Seizures 29
References [1] Rose AL, Lombroso CT. Neonatal seizure states: A study of clinical, pathological, and electroencephalographic features in 137 fullterm babies with a long-term follow-up. Pediatrics 1970;45:404-25. [2] Hauser WA, Annegers JF, Kurland LT. Prevalence of epilepsy in Rochester, Minnesota: 1940-1980. Epilepsia 1991;32:429-45. [3] Camfield CS, Camfield PR, Gordon K, Wirrell E, Dooley J. Incidence of epilepsy in childhood and adolescence: A populationbased study in Nova Scotia from 1977 to 1985. Epilepsia 1996;37:19-23. [4] Lanska MJ, Lanska DJ. Neonatal seizures in the United States: Results of the National Hospital Discharge Survey, 1980-1991. Neuroepidemiology 1996;15:117-25. [5] Holmes GL. Do seizures cause brain damage? Epilepsia 1991; 32(Suppl 5):S14-28. [6] Scher MS, Painter MJ, Bergman I, Barmada MA, Brunberg J. EEG diagnoses of neonatal seizures: Clinical correlation and outcome. Pediatr Neurol 1989;5:17-24. [7] Legido A, Clancy RR, Berman PH. Neurologic outcome after electroencephalographically proven neonatal seizures. Pediatrics 1991; 88:583-96. [8] Temple CM, Dennis J, Carney R, Sharich J. Neonatal seizures: Long-term outcome and cognitive development among “normal” survivors. Dev Med Child Neurol 1995;37:109-18. [9] Ortibus EL, Sum JM, Hahn JS. Predictive value of EEG for outcome and epilepsy following neonatal seizures. Electroencephalogr Clin Neurophysiol 1996;98:175-85. [10] Volpe JJ. Neonatal seizures: Current concepts and revised classification. Pediatrics 1989;84:422-8. [11] Volpe JJ. Neonatal seizures. In: Neurology of the newborn, 3rd ed. Philadelphia: WB Saunders, 1995:172-207. [12] Mizrahi EM, Kellaway P. Characterization and classification of neonatal seizures. Neurology 1987;37:1837-44. [13] Clancy RR. The contribution of EEG to the understanding of neonatal seizures. Epilepsia 1996;37(Suppl 1):S52-9. [14] Salinsky M, Kanter R, Dasheiff RM. Effectiveness of multiple EEGs in supporting the diagnosis of epilepsy: An operational curve. Epilepsia 1987;28:331-4.
30
PEDIATRIC NEUROLOGY
Vol. 20 No. 1
[15] Cascino GD, Trenerry MR, So EL, et al. Routine EEG and temporal lobe epilepsy: Relation to long-term EEG monitoring, quantitative MRI, and operative outcome. Epilepsia 1996;37:651-6. [16] Theodore WH, Sato S, Porter RJ. Serial EEG in intractable epilepsy. Neurology 1984;34:863-7. [17] Clancy RR, Legido A. Postnatal epilepsy after EEG-confirmed neonatal seizures. Epilepsia 1991;32:69-76. [18] Clancy RR, Legido A. The exact ictal and interictal duration of electroencephalographic neonatal seizures. Epilepsia 1987;28:537-41. [19] Scher MS, Aso K, Beggarly ME, Hamid MY, Steppe DA, Painter MJ. Electrographic seizures in preterm and full-term neonates: Clinical correlates, associated brain lesions, and risk for neurologic sequelae. Pediatrics 1993;91:128-34. [20] Johnston MV. Developmental aspects of epileptogenesis. Epilepsia 1996;(Suppl 1):S2-9. [21] Holmes GL. The long-term effects of seizures on the developing brain: Clinical and laboratory issues. Brain Dev 1991;13:393-409. [22] Mizrahi EM. Consensus and controversy in the clinical management of neonatal seizures. Clin Perinatol 1989;16:485-500. [23] Tharp BR, Scher MS, Clancy RR. Serial EEGs in normal and abnormal neonates with birth weights less than 1200 grams: A prospective study with long term follow-up. Neuropediatrics 1989;20:64-72. [24] Jensen FE, Applegate CD, Holtzman D, Belin TR, Burchfiel JL. Epileptogenic effect of hypoxia in the immature rodent brain. Ann Neurol 1991;29:629-37. [25] Lacey DJ, Topper WH, Buckwald S, Zorn WA, Berger PE. Preterm very-low-birth-weight infants: Relationship of EEG to intracranial hemorrhage, perinatal complications, and developmental outcome. Neurology 1986;36:1084-7. [26] Moshe SL. Epileptogenesis and the immature brain. Epilepsia 1987;28(Suppl 1):S3-15. [27] Volpe JJ. Hypoxic-ischemic encephalopathy: Clinical aspects. In: Neurology of the newborn, 2nd ed. Philadelphia, WB Saunders, 1995:236-79. [28] Lombroso CT. Neonatal seizures: Historic note and present controversies. Epilepsia 1996;37(Suppl 3):5-13.
and often are difficult to classify [10,11]. The sick newborn often displays a variety of common paroxysmal behaviors, referred to as subtle seizures, that probably are not epileptic and may represent brainstem release phenomena that result from severe diffuse brain injury [12]. Experiences from the neonatal clinical research centers suggest that clinical observation of seizure semiology unaccompanied by electroencephalogram (EEG) confirmation has severe limitations [13]. Accordingly, the EEG plays an important role in understanding the nature of paroxysmal movements in neonates and in differentiating epileptic seizures from nonepileptic events. The presence of epileptiform discharges on the EEG is valuable in confirming a clinical suspicion of epileptic seizures. About half of all adults with seizures will have epileptiform discharges on the initial EEG [14]. This confirmatory rate increases to 75% when more than one EEG study is performed [15]. Similar confirmatory rates have been reported in children with seizures [16]. The confirmatory rate in both children and adults appears to be influenced by the duration of the EEG record, with higher confirmatory rates in EEG records of longer duration. Outside the neonatal period the etiology for seizures may be symptomatic or idiopathic; neonatal seizures are almost exclusively symptomatic. As in the older age groups, interictal epileptiform discharges can occur in the absence of documented seizures. The influences of gestation and etiology on the EEG confirmatory rate in neonates treated for seizures in the contemporary intensive care unit were examined in this study. Patients and Methods
Introduction The highest incidence of seizures occurs during the neonatal period [1-9]. Seizures in neonates can be variable
From the Department of Neurology; University of Wisconsin; Madison, Wisconsin.
© 1999 by Elsevier Science Inc. All rights reserved. PII S0887-8994(98)00078-2 ● 0887-8994/99/$20.00
Neonates admitted to a neonatal intensive care unit at a children’s hospital between January 1, 1986 and December 31, 1996 were studied. The neonatal unit (442 6 36 admissions per year) served as a regional tertiary referral center for a rural population that remained stable for the duration of the study period. The medical records of each neonate
Communications should be addressed to: Dr. Sheth; Department of Neurology; University of Wisconsin; 600 Highland Avenue H6/574 CSC; Madison, WI 53792-5132. Received April 10, 1998; accepted June 17, 1998.
Sheth: EEG Confirmatory Rate in Neonatal Seizures 27
admitted to the neonatal unit were abstracted daily to a standardized data entry form by a neonatologist or neonatal nurse practitioner and entered into a comprehensive prospectively maintained neonatal database. Data integrity was validated by a designated database supervisor at least three times weekly. In instances in which neonates were transferred from the neonatal intensive care unit, the final disposition was tracked until the neonate died or was discharged from the hospital. All neonatal EEGs performed over the 11-year period were reviewed for the presence of epileptiform discharges or electrographic seizures and background abnormalities by one of three electroencephalographers [11]. EEG recordings were performed at the bedside using a 21-channel portable system and were, typically, at least 1 hour long. Surface electrodes were applied according to the 10-20 International System, and tracings were recorded at 15 mm/s, with a recording sensitivity of 7 mV/mm. Extraocular movements, electrocardiogram, and thoracic respiratory movements were recorded simultaneously with the EEG. EEG background was evaluated with respect to gestation and sleep state. The EEG background was examined interictally and classified as normal or mildly, moderately, or severely abnormal, as previously described [17]. An electrographic seizure was defined as ictal by the appearance of sudden, repetitive, evolving stereotypical waveforms with a minimum duration of 10 seconds [18]. Rhythmic or semirhythmic sharp waves occurring repeatedly in a localized area were used to identify interictal epileptiform EEG discharges. Visual analysis of sharp transients, with respect to gestation and temporal or frontal location, was used to differentiate normal sharp transients from epileptiform discharges. An EEG was categorized as confirmatory if it demonstrated electrographic seizures or interictal epileptiform discharges. Absence of epileptiform discharges or electrographic seizures in the face of an abnormal EEG background was considered supportive of the diagnosis. Clinical seizures were identified by direct observation at the bedside by a member of the neonatal team. Neonates with clinical seizures that were treated with an antiepileptic medication were included. Postmenstrual age at birth was determined using the best obstetric estimate available (prenatal ultrasound, menstrual dates, positive pregnancy test) and corroborated by postnatal clinical assessment. Seizure etiology was approached from the perspective of neuropathologic findings and was classified into four broad neurologic categories. Neurologic disorders examined included germinal matrix-periventricular hemorrhage, hypoxic-ischemic encephalopathy, central nervous system (CNS) infections (including TORCH [toxoplasmosis, other (virus), rubella, cytomegalovirus, herpes] infections, encephalitis, and culturepositive meningitis), and cerebral dysgenesis (dysgenesis included congenital brain malformations and those chromosomal aberrations with CNS manifestations). Bivariate analysis using the Student t test for continuous variables was used to examine the relationship between the EEG findings and gestation, seizure etiology, and mortality rate.
Results One hundred eighty-three neonates (3.9% of all neonates in the intensive care unit) with seizures underwent a total of 352 EEGs. The EEG was abnormal in 144 neonates (79%) and normal in 39 neonates. Epileptiform discharges on the initial EEG were present in 84 neonates (45%). A further 29 neonates without epileptiform discharges on initial EEG demonstrated epileptiform discharges on subsequent EEG studies. Thus, of the total cohort of 183 neonates, the EEG was directly confirmatory in 113 neonates, yielding an overall EEG confirmatory rate for neonatal seizures of 60%. Of these 113 neonates, 34 had frank electrographic seizures recorded (with or without associated clinical manifestation). In addition, 84 of the 113 neonates with epileptiform
28
PEDIATRIC NEUROLOGY
Vol. 20 No. 1
Table 1. Electroencephalogram confirmatory rates by etiology in 183 neonates with seizures
Etiology
Neonates (n)
EEF Confirmatory Rate (%)
Hypoxic-ischemic encephalopathy Germinal matrix hemorrhage CNS dysgenesis CNS infection Miscellaneous
71 28 23 15 68
80 65 65 95 66
Abbreviation: CNS 5 Central nervous system
discharges had moderate or severe EEG background abnormalities, 20 had mild background abnormalities, and the remaining nine neonates had a normal EEG background for gestation. Nonepileptiform EEG background abnormalities were present in 31 neonates (17%): 14 had burst suppression, 11 had dysmaturity of the background, and six had electrocerebral silence. Gestation exerted a powerful influence on the overall EEG confirmatory rate. The confirmatory rate increased from 63% at a gestation of 28 weeks to 77% at term (P , 0.02). EEG background abnormalities also depended on gestation, such that proportionately, marked background abnormalities were more likely to be present in younger gestation neonates, and moderate and mild background abnormalities were more likely as the neonate approached term (P 5 NS). Etiology for the neonatal seizures also influenced the overall EEG confirmatory rate (Table 1). Etiology could be categorized into one of the four neurologic groups in 115 neonates (63%): 71 (50%) had hypoxic-ischemic encephalopathy, 28 (20%) had germinal matrix-intraventricular hemorrhage, 23 (16%) had cerebral dysgenesis, and 15 (10%) had a CNS infection. Twenty-two neonates had two or more neurologic insults underlying their seizures. Etiology for the remaining 68 neonates (37%) formed a miscellaneous group that included metabolic derangement, stroke, shock, and seizures of undetermined origin. The EEG confirmatory rate for neonates with seizures resulting from hypoxic-ischemic encephalopathy was 80% (P , 0.05) and 95% (P , 0.05) from CNS infection. The confirmatory rate for seizures resulting from germinal matrix-intraventricular hemorrhage or congenital malformations was 65% each (P 5 NS). Not unexpectedly, the distribution of seizure etiology was influenced heavily by gestation. Seizures were more likely to result from CNS infection or germinal matrix-intraventricular hemorrhage in preterm neonates; neonates closer to term were more likely to have seizures as a result of hypoxic-ischemic encephalopathy or CNS malformations. To define the relative contribution of etiology to the EEG confirmatory rate beyond that exerted by gestation alone, multivariate regression analysis was performed. Hypoxicischemic encephalopathy and CNS infections influenced
the EEG confirmatory rate beyond that expected by the influence of gestation alone (P , 0.05 for each). The influence of germinal matrix-intraventricular hemorrhage and CNS dysgenesis on the EEG confirmatory rate, however, appeared to result almost entirely from the influence of gestation. An etiology other than one of the four neurologic groups occurred in 68 neonates. For these neonates the overall EEG confirmatory rate was 66% (45 had an abnormal EEG and 23 a normal EEG). Trends in the overall EEG confirmatory rate over time also were examined. During the 11-year period, there was a statistically significant trend toward lower EEG confirmatory rates with time. Generally, neonates born before 1990 had an EEG confirmatory rate of 82%, and neonates born during or after 1990 had an EEG confirmatory rate of 68% (P , 0.05). A total of 29 neonates (15%) of this cohort died during their intensive care unit stay. A confirmatory EEG was predictive of mortality (19% vs 6%, P , 0.03). Discussion The overall 3.9% incidence of seizures of this cohort of 4,654 neonates during the 11-year period was similar to the incidence reported for other contemporary studies [19]. Of the 183 neonates with seizures, 77% had either epileptiform discharges or nonepileptiform background abnormalities on the EEG. The initial EEG revealed epileptiform discharges in 45% of the cohort, subsequent recordings were confirmatory in an additional 15%. Thus the overall 60% direct confirmatory rate observed in neonates was approximately 15% lower than that reported for seizures occurring in patients beyond the neonatal period [14]. Gestation exerted a strong influence on the EEG confirmatory rate. The confirmatory rate increased from approximately 63% at 28 weeks gestation to 79% at term. The lower rate observed in preterm neonates may be a reflection of behavioral movements resembling seizures but not associated with an EEG signature [13]. CNS maturation is determined by gestation such that rapid changes in the degree of neuronal organization, bioelectrical connectivity, and ictal propagation may underlie this observation [20-26]. Accordingly, the ability of a neonate who has sustained neurologic insult to manifest seizures may depend on gestation. The distribution of neurologic disorders in neonates is influenced heavily by gestation [27]. In this study, term neonates were more likely to have hypoxic-ischemic encephalopathy and CNS dysgenesis as etiologies for seizures; more preterm neonates were more likely to have germinal matrix-intraventricular hemorrhage and CNS infections as etiologies. However, etiology exerted an influence on the EEG confirmatory rate above and beyond that expected for gestation alone. The EEG confirmatory rate was 80% for hypoxic-ischemic encephalopathy and 95% for CNS infection; whereas the confirmatory rate was
65% each for germinal matrix-intraventricular hemorrhage and CNS dysgenesis. This divergence between gestation and etiology for seizures on the EEG confirmatory rate might be a reflection of the contribution of different underlying pathogenic mechanism. For those neonates in whom an etiology for seizures could not be categorized into one of the four neurologic groups the confirmatory rate was similar to the lower EEG confirmatory rate observed with germinal matrix-intraventricular hemorrhage and CNS dysgenesis. Seventeen percent of neonates without epileptiform discharges had isolated nonepileptiform background abnormalities. The absence of epileptiform discharges does not allow confidence in ascribing an epileptic nature to the events being treated. However, because the vast majority of seizures occurring in the intensive care unit are likely to be acute symptomatic rather than benign idiopathic, background abnormalities may help provide some indirect supportive evidence for a diagnosis of neonatal seizures. In the late 1970s and early 1980s a number of innocent neonatal episodic behaviors were characterized as epileptic. Prolonged video-EEG monitoring of neonates with suspected seizures between 1982 and 1986 helped identify those clinical seizure manifestations with a high degree of EEG confirmation [13]. The EEG confirmatory rates for various neonatal seizure types has been well defined [22], although it was not possible to reliably ascertain the EEG confirmatory rates by clinical seizure type in this study. The retrospective nature of this study imposes limitations. For instance, sharp transients were separated from interictal activity by their morphology, distribution, and EEG background accompaniments. However, interictal epileptiform EEG discharges in the neonate are not easily defined and may be difficult to separate from normal sharp transients, especially in preterm neonates. Furthermore, the lower confirmatory rate for seizures in neonates compared with older individuals may be a result of some neonates with seizures having a normal interictal EEG. Duration of EEG sampling also may be a factor in determining EEG confirmatory rates. Almost all neonates in this cohort had EEG recordings that lasted for at least 1 hour, although continuous EEG monitoring may have yielded higher EEG confirmatory rates. Monitoring seizures in neonates with simultaneous video-EEG in the neonatal intensive care unit is highly desirable, although not widely available [22,28]. The clinical circumstance reported in this study may thus offer some generally applicable estimates of neonatal EEG confirmatory rates. Interestingly, confirmatory rates for EEGs performed on or before 1990 were 14% higher than the rate from 1990 to the end of the study period, suggesting an increased use of EEG in the management of neonatal seizures. Routine neonatal video-EEG monitoring and a higher EEG repetition rate in neonates with a normal initial EEG perhaps may have increased the sensitive of the observed EEG confirmatory rates in neonatal seizures.
Sheth: EEG Confirmatory Rate in Neonatal Seizures 29
References [1] Rose AL, Lombroso CT. Neonatal seizure states: A study of clinical, pathological, and electroencephalographic features in 137 fullterm babies with a long-term follow-up. Pediatrics 1970;45:404-25. [2] Hauser WA, Annegers JF, Kurland LT. Prevalence of epilepsy in Rochester, Minnesota: 1940-1980. Epilepsia 1991;32:429-45. [3] Camfield CS, Camfield PR, Gordon K, Wirrell E, Dooley J. Incidence of epilepsy in childhood and adolescence: A populationbased study in Nova Scotia from 1977 to 1985. Epilepsia 1996;37:19-23. [4] Lanska MJ, Lanska DJ. Neonatal seizures in the United States: Results of the National Hospital Discharge Survey, 1980-1991. Neuroepidemiology 1996;15:117-25. [5] Holmes GL. Do seizures cause brain damage? Epilepsia 1991; 32(Suppl 5):S14-28. [6] Scher MS, Painter MJ, Bergman I, Barmada MA, Brunberg J. EEG diagnoses of neonatal seizures: Clinical correlation and outcome. Pediatr Neurol 1989;5:17-24. [7] Legido A, Clancy RR, Berman PH. Neurologic outcome after electroencephalographically proven neonatal seizures. Pediatrics 1991; 88:583-96. [8] Temple CM, Dennis J, Carney R, Sharich J. Neonatal seizures: Long-term outcome and cognitive development among “normal” survivors. Dev Med Child Neurol 1995;37:109-18. [9] Ortibus EL, Sum JM, Hahn JS. Predictive value of EEG for outcome and epilepsy following neonatal seizures. Electroencephalogr Clin Neurophysiol 1996;98:175-85. [10] Volpe JJ. Neonatal seizures: Current concepts and revised classification. Pediatrics 1989;84:422-8. [11] Volpe JJ. Neonatal seizures. In: Neurology of the newborn, 3rd ed. Philadelphia: WB Saunders, 1995:172-207. [12] Mizrahi EM, Kellaway P. Characterization and classification of neonatal seizures. Neurology 1987;37:1837-44. [13] Clancy RR. The contribution of EEG to the understanding of neonatal seizures. Epilepsia 1996;37(Suppl 1):S52-9. [14] Salinsky M, Kanter R, Dasheiff RM. Effectiveness of multiple EEGs in supporting the diagnosis of epilepsy: An operational curve. Epilepsia 1987;28:331-4.
30
PEDIATRIC NEUROLOGY
Vol. 20 No. 1
[15] Cascino GD, Trenerry MR, So EL, et al. Routine EEG and temporal lobe epilepsy: Relation to long-term EEG monitoring, quantitative MRI, and operative outcome. Epilepsia 1996;37:651-6. [16] Theodore WH, Sato S, Porter RJ. Serial EEG in intractable epilepsy. Neurology 1984;34:863-7. [17] Clancy RR, Legido A. Postnatal epilepsy after EEG-confirmed neonatal seizures. Epilepsia 1991;32:69-76. [18] Clancy RR, Legido A. The exact ictal and interictal duration of electroencephalographic neonatal seizures. Epilepsia 1987;28:537-41. [19] Scher MS, Aso K, Beggarly ME, Hamid MY, Steppe DA, Painter MJ. Electrographic seizures in preterm and full-term neonates: Clinical correlates, associated brain lesions, and risk for neurologic sequelae. Pediatrics 1993;91:128-34. [20] Johnston MV. Developmental aspects of epileptogenesis. Epilepsia 1996;(Suppl 1):S2-9. [21] Holmes GL. The long-term effects of seizures on the developing brain: Clinical and laboratory issues. Brain Dev 1991;13:393-409. [22] Mizrahi EM. Consensus and controversy in the clinical management of neonatal seizures. Clin Perinatol 1989;16:485-500. [23] Tharp BR, Scher MS, Clancy RR. Serial EEGs in normal and abnormal neonates with birth weights less than 1200 grams: A prospective study with long term follow-up. Neuropediatrics 1989;20:64-72. [24] Jensen FE, Applegate CD, Holtzman D, Belin TR, Burchfiel JL. Epileptogenic effect of hypoxia in the immature rodent brain. Ann Neurol 1991;29:629-37. [25] Lacey DJ, Topper WH, Buckwald S, Zorn WA, Berger PE. Preterm very-low-birth-weight infants: Relationship of EEG to intracranial hemorrhage, perinatal complications, and developmental outcome. Neurology 1986;36:1084-7. [26] Moshe SL. Epileptogenesis and the immature brain. Epilepsia 1987;28(Suppl 1):S3-15. [27] Volpe JJ. Hypoxic-ischemic encephalopathy: Clinical aspects. In: Neurology of the newborn, 2nd ed. Philadelphia, WB Saunders, 1995:236-79. [28] Lombroso CT. Neonatal seizures: Historic note and present controversies. Epilepsia 1996;37(Suppl 3):5-13.