Recurrent Status Epilepticus in Chddren Shlomo Shinnar, MD, PhD,"?# Joseph Maytal, MD,"?"" Lawrence Krasnoff, PhD,l and Solomon L. Moshe, M D " i f #
Status epilepticus is an uncommon but life-threatening seizure. Little is known about the risk of recurrent status epilepticus in patients who present with an initial episode. To determine the risk of recurrent status epilepticus in children, we prospectively followed 95 children, identified at the time of their first episode of status epilepticus, for a mean of 29.0 months (range, 4-60 months). The patients' ages ranged from 1 month to 18 years (mean, 4.6 years). The cause of the status epilepticus was classified as idiopathic (n = 24), remote symptomatic (n = 18), febrile (n = 29), acute symptomatic (n = IS), or progressive neurological disorder (n = 6). Sixteen children (17%) had at least 2 episodes of status epilepticus. The risk of recurrent status was 4% (n = 1) in the idiopathic group, 44% (n = 8) in the remote symptomatic group, 3% (n = 1) in the febrile group, 11% (n = 2) in the acute symptomatic group, and 67% (n = 4) in those with progressive neurological disease. Recurrent status epilepticus occurred primarily in neurologically abnormal children. While neurologically abnormal children accounted for 34%)(n = 32) of all children with status epilepticus, they comprised 88% (n = 14) of the children with recurrent status epilepricus ( p < 0.001) and all 5 of the children with multiple ( 2 3) episodes of status ( p < 0.001). Fifteen of 16 children with recurrent status epilepticus were being treated with antiepileptic drugs at the time of recurrence. The morbidity and mortality of status epilepticus were low. No child had a measurable deterioration in neurological function following a subsequent episode of status epilepticus. Recurrent status epilepticus occurs primarily in children with an underlying neurological abnormality. Status epilepticus reoccurs in this group despite conventional antiepileptic drug therapy. Shinnar S, Maytal J, Krasnoff L, Moshe SL. Recurrent status epilepticus in children. A n n Neurol 1902;31:598-604
Status epilepticus is one of the more common neurological emergencies requiring immediate aggressive intervention. In contrast to brief seizures that d o not cause damage, status epilepticus may result in neurological sequelae [I-4). It is relatively common in patients with epilepsy and particularly in children [ S , 61. Among adults with epilepsy, 1.3 to 12.3% will have at least 1 episode of status epilepticus [ 5 , 61. In one series, 16% of children with epilepsy experienced at least 1 episode of status [l}. Among patients with a first unprovoked seizure, 9% of children [7} and 11%)of adults [S} presented with status epilepticus. Recent studies have shown that the morbidity and mortality associated with a single episode of status epilepticus per se are low [9-12). The question frequently raised is whether patients with 1 episode of status epilepticus are more prone to experience further episodes. Surprisingly, there are few data that would allow the clinician to identify children ar risk for recurrent status epilepticus. The answer to this question and the identification of the specific subgroups that may be at particularly high risk for recurrent episodes of status are of great importance. These subgroups are presum-
ably the children one would treat most aggressively with antiepileptic drugs. A few studies reported that 10 to 18% of children with status epilepticus experienced at least 1 additional episode { I , 12, 131. However, little information is available on the length of follow-up, the types of patients, the risk factors for recurrence, or the effect of treatment. In a study of status epilepticus in children [9}, we found that the short-term morbidity and mortality of status were low and depended primarily on the etiology of the seizures. This cohort has been followed for several years. In the present study, we report the risk of recurrent status epilepticus in the 95 children in this cohort who were identified prospectively at the rime of the initial episode of status epilepticus. Risk factors for recurrent status epilepticus and the effect of trearment with chronic antiepileptic drug therapy are also analyzed.
From the Departments of "Neurology, tpediatrics, and tNeuroscience, the Albert Einstein College of Medicine, Bronx; SMontefiorei Einstein Epilepsy Management Center, and llMontefiore/Einstein Department Of Biostatistics, Center, Bronx; and **Division of Pediatric Neurology, Schneider Children's Hospital, Long Island Jewish Medical Center, Queens, NY.
Received May 20, 1991, and in revised form Oct 28. Accepted for publication Nov 11, 1991.
598
Materials and Methods PopuIatzon This prospective study consisted of all parients who had their first episode of status epilepticus between January 1985 and
Address correspondence to D r Shinnar, Lhvision ot Pediatric N c w r()iogy, VCp-207, M(,ntcfiore Medical (;cnrer, [ [ 1 E 2 1 Ortt Sr. Bronx, NY 104h7.
Copyright 0 1992 by the American Neurological Association
June 1987 and who were identified at the time of the episode and evaluated by us. Children were seen at the affiliated hospitals of Albert Einstein College of Medicine in the Bronx, New York, including Montefiore Medical Center, Bronx Municipal Hospital Center, and North Central Bronx Hospital. A detailed medical history was obtained, and included type, cause, and duration of seizures, and the treatment given. A neurological examination was performed in all patients. Of the 120 patients in this prospective study, 3 died and 10 were lost to follow-up immediately after discharge and were not included in the analysis. As previously reported, the cause of death in the 3 children was related to the acute underlying insult and not specifically to the episode of status epilepticus [9). An additional 12 children were excluded from this analysis because they had had prior episodes of status epilepticus. The remaining 95 children were followed either at the MontefiorelEinstein Epilepsy Management Center o r the Bronx Municipal Hospital Pediatric Neurology Clinic; the mean duration of follow-up was 29.0 months (range, 4-60 months). In this observational study, no attempt was made to influence antiepileptic drug therapy. As ail the index episodes of status epilepticus occurred between January 1985 and June 1987, at least 33 months had elapsed between the time of study entry and data analysis. The shorter length of follow-up available for some patients was due to losses to follow-up.
Follow-up Follow-up consisted of periodic clinic visits with neurological evaluation by the investigators. Neurological sequelae were determined by formal neurological examination. Cognitive function was assessed by mental status testing as part of the neurological and neurodevelopmental examination. No formal intelligence testing was carried out in the majority of children. However, in addition to screening by the examiner, parents were asked whether there were any differences in the child's functioning o r school performance after the acute illness. Any reported deterioration in preschool performance or function was considered a sign of cognitive impairment. Serial follow-up telephone calls were made to determine the occurrence of subsequent seizures as well as episodes of subsequent status epilepticus. If the child was given emergency care for seizures after entry to the study, the medical records were reviewed. Prior neurological status was determined by the history and from a review of the medical records when available.
Status Epilepticus Status epilepticus was defined as any seizure lasting more than 30 minutes or recurrent seizures lasting a total of more than 30 minutes without the patient fully regaining consciousness [ S , 6, 91. Seizures were classified in accordance with the revised international classification [14}. The etiology of the seizures was classified using our modification 191 of the system of Hauser and colleagues [ S ] : 1. Idiopathic-a seizure occurring in the absence of an acute precipitating central nervous system insult or systemic metabolic dysfunction.
Remote symptomatic-a seizure occurring without acute provocation in a patient with a history of a prior central nervous system insult known to be associated with an increased risk of convulsions (e.g., stroke, significant head trauma, meningitis, presence of a static encephalopathy, and so on). Febrile-a provoked seizure where the sole acute provocation was fever (temperature > 38.4"C) in a patient without a history of afebrile seizures. Children with previously abnormal neurological status were included. This group technically meets the definition of a prolonged febrile seizure as defined by the National Institutes of Health (NIH) consensus conference [l5, 161, though it is at the extreme of that group. Long-term outcomes of this group were recently reported [ 17). Acute symptomatic-a seizure occurring during an acute illness associated with a known neurological insult or systemic metabolic dysfunction. Febrile convulsions in the absence of another insult were classified separately. Acute symptomatic seizures were further subclassified by specific etiology (i.e., meningitis, trauma, hypoxia, and so on). Progressive encephalopathy-a seizure occurring in the context of a progressive neurological disease. Included in this category were neurodegenerative diseases, malignancies not in remission, and neurocutaneous syndromes. Etiology was classified at the time of the first episode of status epilepticus. No subject was subsequently reclassified.
Data Analysis Data were analyzed using standard statistical methods { 181. The descriptive statistics included the relative proportion of different seizure etiologies, neurological status of the patient before the status epilepticus, and age distribution. The recurrence of status epilepticus was analyzed as a function of the neurological status prior to status epilepticus and of treatment. Chi-square statistics were used. Allp values were computed using two-tailed distributions. As length of follow-up influences the probability of observing a recurrence, the statistical methods of analysis took into account the variable length of follow-up for each child. Univariate analyses for dichotomous variables were performed using the Kaplan-Meier survival analysis {I91and the statistical significance of observed recurrence risks was calculated using the method of Mantel (20, 211. Results are displayed as Kaplan-Meier survival curves, with the cumulative probability of seizure recurrence plotted as a function of time from the first seizure. For the total study group, there were sufficient data to also calculate a 48-month recurrence risk. Univariate analyses for all risk factors and multivariable analyses were performed using the Cox proportional hazards model [ZO, 221. A11 statistical analyses were performed using SAS software [23]. Survival analyses were performed and crosschecked using both PROC PHGLM {24) and PROC LIFETEST (SAS Institute Inc, Cary, NC) 1211.
Results Population Characteristics The 95 children included 47 girls and 48 boys. Twenty-seven children were white, 24 were black, and Shinnar et al: Recurrent Status Epilepticus
5 9
Table 2. Recurvent StatuJ EpileptiruJ (SEI
Table I . Reszirvent Status Epileptisus: Study Population
~~
n Idiopathic Rcmote symptomatic Febrile Acute symptomatic Progressive encephalopathy Total
24 18
20 18
6 75
Mean Follow-up (mo)
Range (mo)
33.6 28.7 31.3 23.6 17.0 27.0
20-52 9-45 6-60 4-48 4-24 4-60
44 were Hispanic, which corresponds to the ethnic distribution served by our medical center. The mean age was 4.6 years (range, 1 month-18 years). The study population is described in Table 1. The type of seizure has been previously described C73. Almost all the children had convulsive status epilepticus. The mean length of follow-up was 29.0 months (range, 4-60 months). Ten children (11%) were followed for more than 4 years; 30 (32%), for more than 3 years; 57 (60(2), for more than 2 years; and 82 (86%), for more than 1 year. Sixty-three (669%)of the children were neurologically normal except for seizures and 32 (34%) were neurologically abnormal. By definition, the 24 children in the idiopathic group were otherwise neurologically normal, normal in the sense of having normal intelligence ( I Q > 70) and no motor deficit. The 18 children with remote symptomatic seizures were all neurologically abnormal, including 7 children with moderate mental retardation, 2 with mental retardation and other congenital anomalies, 3 with normal intelligence and a hemiparesis that was secondary to an infarct in 2, and 6 children with mental retardation and cerebral palsy (spastic quadriparesis in 4 , hemiparesis in 1, and ataxia in 1). The febrile status group included 24 neurologically normal children and 5 neurologically abnormal children, including 1 child with normal cognitive function and a hemiparesis; 1 with microcephaly, developmental delay, and multiple congenital anomalies; and 3 with mental retardation and cerebral palsy (spastic quadriparesis in 2 and hemiparesis in 1). The acute symptomatic group included 15 neurologically normal children and 3 neurologically abnormal children with mental retardation of varying degrees. The 6 children with progressive encephalopathy had progressive neurodegenerative disorders, including 2 with urea cycle defects (1 with partial ornithine transcarbamoylase deficiency and 1 with citrullinemia), 1 with mi tochondrial cytopathy, and 3 with undiagnosed neurodegenerative disorders. Rerurrent Status Epilepticus Sixteen children (17%) had more than 1 episode of status epilepticus (Table 2). The overall Kaplan-Meier estimate of recurrence was 11% at 12 months, 189% at 24 months, and 2096 at 36 and 48 months (Fig 1). 600 Annals of Neurology Vol 31 No 6 June 1792
~
~~
No. of Episodes o f SE
Idiopathic Remote symptomatic Febrile Acute symptomatic Progressive encephalopathy Total
1.0 m
i
n
1
24 18 29 18
23 10 28 16 2 79
6 95
2
23
1 7 1 2 0
0 1 0 0 4
11
5
,
-
-
$! 7 0.4
0
6
12
18 TIME (Months)
24
30
36
-
Fig 1 . Probability of recurrent Jtatus epilepticus following a first episode of status. Kaplan-Meier rurne.
The mean time between the first and second episodes of status was 10.4 months (range, 2-28 months). In 11 (69%) of the 16 children, recurrence occurred within 12 months of the initial episode. Age at the time of the initial episode of status epilepticus did not affect recurrence risk. There were marked differences in recurrence risk according to the etiology of the status epilepticus (see Table 2). Among the 24 children with idiopathic status epilepticus, only 1 experienced a second episode of status. The highest rate of recurrent status epilepticus occurred in the remote symptomatic and progressive encephalopathy groups; 12 ( 5 0 % ) of the 24 children in these two groups experienced recurrent status epilepticus and 5 (2 1%) had 3 or more episodes of status. These children accounted for all 5 patients with multiple (23 ) episodes of status epilepticus ( p < 0.001). The Kaplan-Meier estimates of recurrence for selected groups are shown in Figure 2. The idiopathic group had a cumulative risk of recurrence of 4% at 36 months. The remote symptomatic group had a cumulative recurrence risk of 17% at 12 months, 4573 at 24 months, and 599% at 36 months. T h e relative risk of these two groups was 18.4, with 75% confidence limits of 2.9 and 115.3. The differences were statistically significant in terms of both crude proportions (x2 = 9.7, p < 0.002) and relative risks ( p < 0.005). Of particular interest is that the small group of children with progressive encephalopathy (see Table 2, Fig
1.0
1 0 7 (I)
3
7
09Progressive Encephalopathy(n = 6)
08v)
5
07Remote Symptomatic (n = 18)
06-
Neurologically Abnormal (n = 32)
rT--i--
3
2a 8
05-
04-
-
Neurologically Normal (n = 63)
0
l " " ' 1 ' " " 6 12
I " " " " " ' " ' " ' 7 24 30
18 TIME (Months)
36
F i g 2. Probability of recurrent status epilepticus following a first episode of status in idiopathic, remote symptomatic, and progressive encephalopatby groups. Kapkzn-M eier curve.
F i g 3. Probability of recurrent status epilepticus following an initial episode of status, as a function of neurological status l n o m l versus abnormal). Kaplan-Meier curve.
Table 3. Recurrent Status Epilepticus (SE) and Neurological Status
months in the neurologically normal group, compared with 27%, 48%, and 58% at 12, 24, and 36 months, respectively, in the neurologically abnormal group. The relative risk of recurrence for the neurologically abnormal children was 23.7 with 9595 confidence limits of 6.8 and 82.8 ( p < 0.001). The risk of recurrence in the neurologically abnormal group was related to the degree of neurological impairment though the numbers were too small to permit detailed subgroup analysis. Seventeen children had severe involvement affecting both cognitive and motor development, including 6 with progressive encephalopathy, 7 with both mental retardation and cerebral palsy, and 4 with known genetic syndromes or multiple congenital malformations. Twelve (7 1%) of these children experienced recurrent status epilepticus. In contrast, 15 children had only mental retardation (n = 8) or cerebral palsy (n = 7 ) of varying degrees of severity and only 2 (13%) of them had a second episode of status ( p < 0.005).
Idiopathic Remote symptomatic
Normal
Abnormal
2 2
2 2 n Episodes
n
n
Episodes
24
24
1
18
0 0
0 0 1
8
8 Febrile
29
24
0
1
5 Acute symptomatic
18
15
1
1
3 Progressive encephalopathy Total
6 95
0
4
0
63 2
6 3
14
2
2 ) appeared to have a very high recurrence risk. The cumulative risk of recurrence of this group was already 78% at 12 months. Furthermore, they accounted for 4 of the 5 children with 3 or more episodes of status epilepticus.
Recuwence Risk and Netlvological Status When the recurrence risk was analyzed by neurological status (Table 3), abnormal neurological status was a powerful predictor of recurrence in all groups. Neurologically abnormal children who constituted 34% of the srudy population accounted for 88% of the recurrent status group ( p < 0.001) and for all 5 children with multiple (23) episodes of status epilepticus ( p < 0.001). The Kaplan-Meier estimates of recurrence in these two groups are shown in Figure 3. The estimated risk of recurrent status epilepticus is 3% at 12, 24, and 36
Treatment Eight (SO%) of the 16 children with recurrent status epilepticus, including S of the 8 children in the remote symptomatic group, had prior unprovoked seizures. Ten (63%) of the 16 children with recurrent status epilepticus were already receiving antiepileptic drug therapy at the time of their first episode of status epilepticus. At the time of their second episode, 15 (94%) of the 16 children, including all 14 neurologically abnormal children, were receiving antiepileptic medications. The 1 child who was not receiving medication at the time of the second episode of status epilepticus was an otherwise normal child with acute symptomatic status. It should be noted that drug levels were not being monitored as part of this observational study and the levels may well have been subtherapeutic in some children. Shinnar et al: Recurrent Status Epilepticus
601
Morbidity and Mortality The 3 deaths within 3 months of the occurrence of the index episode of status epilepticus, all of which occurred in the acute symptomatic and the progressive encephalopathy groups, were reported previously 191. They included a 2-month-old with status epilepticus secondary to hypoxic-ischemic encephalopathy associated with cardiac surgery who died 3 days after the episode of status, a 5-month-old with status epilepticus secondary to an intracranial hemorrhage associated with disseminated intravascular coagulation who died 3 days after admission, and a 3-year-old with status associated with a malignant retinoblastoma with central nervous system spread who died 10 days after the episode of status. None of the deaths could be clearly attributed to the status epilepticus per se. These children were not included in this study. Of the 95 children in this study, l child with a progressive neurodegenerative disorder died 13 months after the first and only episode of status epilepticus. She had intractable seizures. None of the 16 children with a history of 2 or more episodes of status epilepticus died during the follow-up period. Moreover, no child had a measurable deterioration in neurological function following a subsequent episode of status epilepticus.
Discussion The striking finding in this study was the high risk of recurrent status epilepticus in children who were neurologically abnormal. The neurologically abnormal group, which constituted 34% of the children with status epilepticus, accounted for 8895 of the children with recurrent status epilepticus and all 5 children with multiple episodes of status. The risk was related to the severity of the neurological abnormality, with multiply handicapped children being at highest risk. The size of this cohort precluded determination of the precise degree of disability necessary to substantially increase the recurrence risk. These recurrent episodes of status epilepticus occurred despite the fact that all 14 neurologically abnormal children were on antiepileptic drug therapy at the time of recurrence. In contrast, the risk of recurrent status epilepticus in neurologically normal children was only 3%) and none of these children experienced more than 2 episodes of status epilepticus during the study. While the increased risk of recurrence in neurologically abnormal children was expected, the magnitude of the increase, with relative risks being greater than 23, was surprising. Few data regarding the risk of recurrent status epilepticus are available in the literature. Four published studies of status epilepticus in children [ l , 4, 12, 131 addressed the issue briefly or not at all. Dunn [ l Z ) reported that 12 (12%) of 97 children with at least 1 episode of status epilepticus experienced recurrent sta-
602
Annals of Neurology Vol 31 No 6 June 1992
tus epilepticus. Cavazzuti and associates 1131 reported that 12 (18%) of 66 children had recurrent status epilepticus. In an older study, Aicardi and Chevrie I l l reported that of 143 children whose long-term outcome was known, 15 (10.55%)had at least 1 subsequent episode of status epilepticus. These studies provided no data regarding the neurological status of the children with recurrent status epilepticus nor how they were different from the children who did not have a recurrence. No data on treatment with antiepileptic drugs were provided. The exception is one recent abstract describing a retrospective chart review of children with status epilepticus, which also concluded that neurologically abnormal children are at high risk for recurrent status epilepticus {25}. The literature regarding adults with status epilepticus is similarly devoid of data regarding the risk of multiple episodes of status epilepticus [ 5 , 63. Support for the low risk of recurrent status epilepticus in neurologically normal children can be deduced from studies of children and adults who present with a first unprovoked seizure [7, 8, 26, 271. In a recent study of children with a first unprovoked seizure [7}, we reported that the risk of any subsequent seizures in neurologically normal children who presented with status epilepticus as their first unprovoked seizure was no different than the risk of subsequent seizures in children who presented with a brief initial seizure. None of the children who presented with status epilepticus as their first unprovoked seizure had a subsequent episode of convulsive status epilepticus. In another study of children and adults who presented with a first unprovoked seizure, Hauser and coworkers [ 8 ) also found that the recurrence risk in subjects with an idiopathic first seizure was no different in those who presented with status epilepticus as their first unprovoked seizure than in those who presented with a briefer initial seizure. In contrast, Hauser and associates [26] found that in subjects with remote symptomatic seizures, the occurrence of status epilepticus as the first seizure is a significant predictor of future seizures. They did not report on the risk of recurrent status epilepticus. Additional evidence for the increased susceptibility of neurologically abnormal children to experience status epilepticus can be deduced from the high proportion of children in the original study who were neurologically abnormal. In a general epilepsy population, 30% of children would be expected to be neurologically abnormal [28, 291. In studies of children with a first unprovoked seizure, the proportion of neurologically abnormal children is even lower 17, 8, 271. In contrast, more than 40% of all children with status epilepticus are neurologically abnormal [ 1, 6, 12, 2 5 1. The question of whether to initiate long-term antiepileptic drug therapy following an initial episode of
status epilepticus is of great importance. In making long-term treatment decisions, the decision is based in part on the risk of future seizures {7, 26, 27, 301. The major concern, however, would be regarding the probability of subsequent life-threatening seizures such as status epilepticus. In our study of children who presented with a first unprovoked seizure {71, we argued that the majority of children who present with a first unprovoked seizure do not require long-term therapy. In fact, we have not treated children with an idiopathic first seizure with long-term antiepileptic drug therapy even when the initial episode was status epilepticus 173. The recurrence risk in that cohort was no different than in children with a brief initial seizure. More importantly, none experienced another episode of convulsive status epilepticus. The results of this study of children with status epilepticus confirm that the risk of subsequent episodes of status epilepticus in neurologically normal children is extremely low. The low morbidity and mortality of status epilepticus in this population, reported in our previous article [9], are also emphasized in this long-term follow-up study. The neurologically abnormal children had a high rate of recurrent status epilepticus despite the fact that all these children were taking antiepileptic drugs at the time of the second episode of status epilepticus. At the time of the first episode of status epilepticus, many were already receiving antiepileptic drug therapy due to prior seizures. In this observational study, we did not attempt to influence the course of treatment and drug levels were not routinely measured. Thus, this is not a study of the efficacy of antiepileptic drug therapy in this population but an intent-to-treat analysis. However, it should be noted that these were neurologically abnormal children who had already experienced an episode of status epilepticus. Therefore, these are the children in whom the neurologists were most aggressive in their therapy. Presumably, these would also be the children for whom the parents would be most concerned about the chances of further seizures in general and of status epilepticus in particular and would therefore be more likely to be compliant. This high-risk subgroup of children with recurrent status epilepticus underscores the need for better treatment strategies in neurologically abnormal children with seizure disorders.
Conclusions Approximately 1 in 6 children with an episode of status epilepticus will experience at least 1 more episode within a few years of the initial episode. The morbidity and mortality of status epilepticus in this population are low. The risk of recurrent status epilepticus in neurologically normal children is very low. In contrast, the risk of recurrent status epilepticus in neurologically abnormal children approaches 50%, even during the limited follow-up period of this study. Neurologically
abnormal children account for almost all children with multiple episodes of status epilepticus. The data suggest that status epilepticus in neurologically normal children is, by and large, an isolated event that should not unduly influence decisions regarding further therapy. The data also underscore the need for effective treatment strategies for neurologically abnormal children with status epilepticus. Larger-scale populationbased studies are needed to better define the precise recurrence risks in this group of patients. This study was supported in part by a teacher investigator development award (1 KO7 NS00930) and a grant (1 R01 NS26151) (ro Dr Shinnar) from the National Institute of Neurological Disorders and Stroke and a Merrit-Putnam fellowship from the Epilepsy Foundation of America (to Dr Maytal). We thank all of the fellows and faculty in the Division of Pediatric Neurology, and the Directors of the Pediatric Emergency Rooms and Intensive Care Units at Montefiore Medical Center, Bronx Municipal Hospital Center, and North Central Bronx Hospital for their help in identifying and recruiting the subjects and for providing their acute medical care. We also acknowledge the cooperation of the New York City Health and Hospital Corporation and Bronx Municipal Hospital Center and North Central Bronx Hospital. We thank Drs Anne T. Berg and Katherine D. Freeman for statistical consultation and Ms Yael Ptachewich for editorial assistance. Lisa Pistorino and Brenda Colon provided computer support. Presented in part at the annual meeting of the American Epilepsy Society, Boston, MA, December 2-7, 1989.
References 1. AicardiJ, ChevrieJJ. Convulsive status epilepticus in infants and in children: a study of 239 cases. Epilepsia 1970;11:187-197 2. Aicardi J, Chevrie JJ. Consequences of status epilepticus in infants and children. Adv Neurol 1983;34:115-125 3. Oxbury JM, Whitty CW. Causes and consequences of status epilepticus in adults: a study of 86 cases. Brain 1971;94: 7 33- 744 4. Fujiwara T, Ishida S, Mlyakoshi M. Stams epilepticus in childhood: a retrospective study of initial convulsive status and subsequent epilepsies. Folia Psychiatr Neurol Jpn 1979;33:337-344 5. Hauser WA. Status epilepticus, frequency, etiology and neurological sequelae. Adv Neurol 1983;34:3-14 6. Hauser WA. Status epilepticus: epidemiologic considerations. Neurology 1990;4O(suppl 2):9- 13 7. Shinnar S, Berg AT, Moshe SL, et al. The risk of seizure recurrence following a first unprovoked seizure in childhood: a prospective study. Pediatrics 1990;85:1076-1085 8. Hauser WA, Anderson VE, Loewenson RB.Seizure recurrence after a first unprovoked seizure. N Engl J Med 1982;307: 522-528 9. Maytal J, Shinnar S, Moshe SL, Alvarez LA. Low morbidity and mortality of status epilepticus in children. Pediatrics 1989;83: 323-331 10. Dulac 0,Aubourg P, Chercoury A, et al. Infantile status epilepticus: clinical, etiological and prognostic aspects. Rev Electroencephalogr Neurophysiol 1985;14:255-262 11. Goulon M, L'evy-Alcover MA, Nouaihat F. Status epilepticus in the adult: epidemiologic and clinical study in an intensive care unit. Rev Electroencephalogr Neurophysiol 1985;14:277-285 12. Dunn WD. Status epilepticus in children: etiology, clinical features and outcome. J Child Neurol 1988;3:167-173
Shinnar et al: Recurrent Status Epilepticus
603
13. Cavazzuti GB, Ferrari P, Lalla M. Follow-up study of 482 cases with convulsive disorders in the first year of life. Dev Med Child Neurol 1984;26:425-437 14. Commission on Classification and Terminology of the International League against Epilepsy. Proposal for revised clinical and electroencephalographic classification of epileptic seizures. Epilepsia 1981;22:489-50 1 15. Nelson KB, Ellenberg JH. Prognosis in children with febrile seizures. Pediatrics 1978;61:720-727 16. National Institutes of Health. Febrile seizures: consensus development conference summary, vol 3, no 2. Bethesda, MD: National Institutes of Health, 1980 17. Maytal J, Shinnar S. Febrile status epilepticus. Pediatrics 1990;86:61 1-6 16 18. Zar JH. Biostatistical analysis. 2nd ed. Englewood Cliffs, NJ: Prentice-Hall, 1984 19. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958;53:457-481 20. Kalbfleisch JD, Prentice RL. The statistical analysis of failure time data. New York: John Wiley, 1980:70-142 2 1. Mantel N. Evaluation of survival data and two new rank order statistics arising in its consideration. Cancer Chemother Rep 196630: 163-1 70
604 Annals of Neurology Vol 31 No 6 June 1992
22. Cox DR. Regressive models and life-tables. J R Stat Sot: [Bl 1972;34:187-220 23. SAS user’s guide: statistics. Version 5 ed. Cary, NC: SAS Institute, 1985 24. SUGI supplemental library user’s guide. Version 5. Cary, NC: SAS Institute, 1986 25. Driscoll SM, Towne AR, Pellock JM, DeLorenzo RJ. Recurrent status epilepticus in children. Neurology 1990;40(suppl I ):I97 (Abstract) 26. Hauser WA, RJch SS, Annegers JF, Anderson VE. Seizure recurrence following a first unprovoked seizure: an extended follow-up. Neurology 1990;40:1163-1 170 27. Berg AT, Shinnar S. The risk of seizure recurrence following a first unprovoked seizure: a quantitative review. Neurology 1991;41965-972 28. Hauser WA, Kurland LT. The epidemiology of epilepsy in Rochester, Minnesota, I935 through 1967. Epilepsia 1975; 16:l-66 29. Hauser WA, Hesdorffer DC. Epilepsy: frequency, causes and consequences. New York: Demos, 1990 30. Shinnar S. Treatment decisions in childhood seirures. In: Dodson WE, Pellock JM, eds. Pediatric epilepsy: diagnosis and thcrapy. New York: Demos, 1991:215-221
Status epilepticus is an uncommon but life-threatening seizure. Little is known about the risk of recurrent status epilepticus in patients who present with an initial episode. To determine the risk of recurrent status epilepticus in children, we prospectively followed 95 children, identified at the time of their first episode of status epilepticus, for a mean of 29.0 months (range, 4-60 months). The patients' ages ranged from 1 month to 18 years (mean, 4.6 years). The cause of the status epilepticus was classified as idiopathic (n = 24), remote symptomatic (n = 18), febrile (n = 29), acute symptomatic (n = IS), or progressive neurological disorder (n = 6). Sixteen children (17%) had at least 2 episodes of status epilepticus. The risk of recurrent status was 4% (n = 1) in the idiopathic group, 44% (n = 8) in the remote symptomatic group, 3% (n = 1) in the febrile group, 11% (n = 2) in the acute symptomatic group, and 67% (n = 4) in those with progressive neurological disease. Recurrent status epilepticus occurred primarily in neurologically abnormal children. While neurologically abnormal children accounted for 34%)(n = 32) of all children with status epilepticus, they comprised 88% (n = 14) of the children with recurrent status epilepricus ( p < 0.001) and all 5 of the children with multiple ( 2 3) episodes of status ( p < 0.001). Fifteen of 16 children with recurrent status epilepticus were being treated with antiepileptic drugs at the time of recurrence. The morbidity and mortality of status epilepticus were low. No child had a measurable deterioration in neurological function following a subsequent episode of status epilepticus. Recurrent status epilepticus occurs primarily in children with an underlying neurological abnormality. Status epilepticus reoccurs in this group despite conventional antiepileptic drug therapy. Shinnar S, Maytal J, Krasnoff L, Moshe SL. Recurrent status epilepticus in children. A n n Neurol 1902;31:598-604
Status epilepticus is one of the more common neurological emergencies requiring immediate aggressive intervention. In contrast to brief seizures that d o not cause damage, status epilepticus may result in neurological sequelae [I-4). It is relatively common in patients with epilepsy and particularly in children [ S , 61. Among adults with epilepsy, 1.3 to 12.3% will have at least 1 episode of status epilepticus [ 5 , 61. In one series, 16% of children with epilepsy experienced at least 1 episode of status [l}. Among patients with a first unprovoked seizure, 9% of children [7} and 11%)of adults [S} presented with status epilepticus. Recent studies have shown that the morbidity and mortality associated with a single episode of status epilepticus per se are low [9-12). The question frequently raised is whether patients with 1 episode of status epilepticus are more prone to experience further episodes. Surprisingly, there are few data that would allow the clinician to identify children ar risk for recurrent status epilepticus. The answer to this question and the identification of the specific subgroups that may be at particularly high risk for recurrent episodes of status are of great importance. These subgroups are presum-
ably the children one would treat most aggressively with antiepileptic drugs. A few studies reported that 10 to 18% of children with status epilepticus experienced at least 1 additional episode { I , 12, 131. However, little information is available on the length of follow-up, the types of patients, the risk factors for recurrence, or the effect of treatment. In a study of status epilepticus in children [9}, we found that the short-term morbidity and mortality of status were low and depended primarily on the etiology of the seizures. This cohort has been followed for several years. In the present study, we report the risk of recurrent status epilepticus in the 95 children in this cohort who were identified prospectively at the rime of the initial episode of status epilepticus. Risk factors for recurrent status epilepticus and the effect of trearment with chronic antiepileptic drug therapy are also analyzed.
From the Departments of "Neurology, tpediatrics, and tNeuroscience, the Albert Einstein College of Medicine, Bronx; SMontefiorei Einstein Epilepsy Management Center, and llMontefiore/Einstein Department Of Biostatistics, Center, Bronx; and **Division of Pediatric Neurology, Schneider Children's Hospital, Long Island Jewish Medical Center, Queens, NY.
Received May 20, 1991, and in revised form Oct 28. Accepted for publication Nov 11, 1991.
598
Materials and Methods PopuIatzon This prospective study consisted of all parients who had their first episode of status epilepticus between January 1985 and
Address correspondence to D r Shinnar, Lhvision ot Pediatric N c w r()iogy, VCp-207, M(,ntcfiore Medical (;cnrer, [ [ 1 E 2 1 Ortt Sr. Bronx, NY 104h7.
Copyright 0 1992 by the American Neurological Association
June 1987 and who were identified at the time of the episode and evaluated by us. Children were seen at the affiliated hospitals of Albert Einstein College of Medicine in the Bronx, New York, including Montefiore Medical Center, Bronx Municipal Hospital Center, and North Central Bronx Hospital. A detailed medical history was obtained, and included type, cause, and duration of seizures, and the treatment given. A neurological examination was performed in all patients. Of the 120 patients in this prospective study, 3 died and 10 were lost to follow-up immediately after discharge and were not included in the analysis. As previously reported, the cause of death in the 3 children was related to the acute underlying insult and not specifically to the episode of status epilepticus [9). An additional 12 children were excluded from this analysis because they had had prior episodes of status epilepticus. The remaining 95 children were followed either at the MontefiorelEinstein Epilepsy Management Center o r the Bronx Municipal Hospital Pediatric Neurology Clinic; the mean duration of follow-up was 29.0 months (range, 4-60 months). In this observational study, no attempt was made to influence antiepileptic drug therapy. As ail the index episodes of status epilepticus occurred between January 1985 and June 1987, at least 33 months had elapsed between the time of study entry and data analysis. The shorter length of follow-up available for some patients was due to losses to follow-up.
Follow-up Follow-up consisted of periodic clinic visits with neurological evaluation by the investigators. Neurological sequelae were determined by formal neurological examination. Cognitive function was assessed by mental status testing as part of the neurological and neurodevelopmental examination. No formal intelligence testing was carried out in the majority of children. However, in addition to screening by the examiner, parents were asked whether there were any differences in the child's functioning o r school performance after the acute illness. Any reported deterioration in preschool performance or function was considered a sign of cognitive impairment. Serial follow-up telephone calls were made to determine the occurrence of subsequent seizures as well as episodes of subsequent status epilepticus. If the child was given emergency care for seizures after entry to the study, the medical records were reviewed. Prior neurological status was determined by the history and from a review of the medical records when available.
Status Epilepticus Status epilepticus was defined as any seizure lasting more than 30 minutes or recurrent seizures lasting a total of more than 30 minutes without the patient fully regaining consciousness [ S , 6, 91. Seizures were classified in accordance with the revised international classification [14}. The etiology of the seizures was classified using our modification 191 of the system of Hauser and colleagues [ S ] : 1. Idiopathic-a seizure occurring in the absence of an acute precipitating central nervous system insult or systemic metabolic dysfunction.
Remote symptomatic-a seizure occurring without acute provocation in a patient with a history of a prior central nervous system insult known to be associated with an increased risk of convulsions (e.g., stroke, significant head trauma, meningitis, presence of a static encephalopathy, and so on). Febrile-a provoked seizure where the sole acute provocation was fever (temperature > 38.4"C) in a patient without a history of afebrile seizures. Children with previously abnormal neurological status were included. This group technically meets the definition of a prolonged febrile seizure as defined by the National Institutes of Health (NIH) consensus conference [l5, 161, though it is at the extreme of that group. Long-term outcomes of this group were recently reported [ 17). Acute symptomatic-a seizure occurring during an acute illness associated with a known neurological insult or systemic metabolic dysfunction. Febrile convulsions in the absence of another insult were classified separately. Acute symptomatic seizures were further subclassified by specific etiology (i.e., meningitis, trauma, hypoxia, and so on). Progressive encephalopathy-a seizure occurring in the context of a progressive neurological disease. Included in this category were neurodegenerative diseases, malignancies not in remission, and neurocutaneous syndromes. Etiology was classified at the time of the first episode of status epilepticus. No subject was subsequently reclassified.
Data Analysis Data were analyzed using standard statistical methods { 181. The descriptive statistics included the relative proportion of different seizure etiologies, neurological status of the patient before the status epilepticus, and age distribution. The recurrence of status epilepticus was analyzed as a function of the neurological status prior to status epilepticus and of treatment. Chi-square statistics were used. Allp values were computed using two-tailed distributions. As length of follow-up influences the probability of observing a recurrence, the statistical methods of analysis took into account the variable length of follow-up for each child. Univariate analyses for dichotomous variables were performed using the Kaplan-Meier survival analysis {I91and the statistical significance of observed recurrence risks was calculated using the method of Mantel (20, 211. Results are displayed as Kaplan-Meier survival curves, with the cumulative probability of seizure recurrence plotted as a function of time from the first seizure. For the total study group, there were sufficient data to also calculate a 48-month recurrence risk. Univariate analyses for all risk factors and multivariable analyses were performed using the Cox proportional hazards model [ZO, 221. A11 statistical analyses were performed using SAS software [23]. Survival analyses were performed and crosschecked using both PROC PHGLM {24) and PROC LIFETEST (SAS Institute Inc, Cary, NC) 1211.
Results Population Characteristics The 95 children included 47 girls and 48 boys. Twenty-seven children were white, 24 were black, and Shinnar et al: Recurrent Status Epilepticus
5 9
Table 2. Recurvent StatuJ EpileptiruJ (SEI
Table I . Reszirvent Status Epileptisus: Study Population
~~
n Idiopathic Rcmote symptomatic Febrile Acute symptomatic Progressive encephalopathy Total
24 18
20 18
6 75
Mean Follow-up (mo)
Range (mo)
33.6 28.7 31.3 23.6 17.0 27.0
20-52 9-45 6-60 4-48 4-24 4-60
44 were Hispanic, which corresponds to the ethnic distribution served by our medical center. The mean age was 4.6 years (range, 1 month-18 years). The study population is described in Table 1. The type of seizure has been previously described C73. Almost all the children had convulsive status epilepticus. The mean length of follow-up was 29.0 months (range, 4-60 months). Ten children (11%) were followed for more than 4 years; 30 (32%), for more than 3 years; 57 (60(2), for more than 2 years; and 82 (86%), for more than 1 year. Sixty-three (669%)of the children were neurologically normal except for seizures and 32 (34%) were neurologically abnormal. By definition, the 24 children in the idiopathic group were otherwise neurologically normal, normal in the sense of having normal intelligence ( I Q > 70) and no motor deficit. The 18 children with remote symptomatic seizures were all neurologically abnormal, including 7 children with moderate mental retardation, 2 with mental retardation and other congenital anomalies, 3 with normal intelligence and a hemiparesis that was secondary to an infarct in 2, and 6 children with mental retardation and cerebral palsy (spastic quadriparesis in 4 , hemiparesis in 1, and ataxia in 1). The febrile status group included 24 neurologically normal children and 5 neurologically abnormal children, including 1 child with normal cognitive function and a hemiparesis; 1 with microcephaly, developmental delay, and multiple congenital anomalies; and 3 with mental retardation and cerebral palsy (spastic quadriparesis in 2 and hemiparesis in 1). The acute symptomatic group included 15 neurologically normal children and 3 neurologically abnormal children with mental retardation of varying degrees. The 6 children with progressive encephalopathy had progressive neurodegenerative disorders, including 2 with urea cycle defects (1 with partial ornithine transcarbamoylase deficiency and 1 with citrullinemia), 1 with mi tochondrial cytopathy, and 3 with undiagnosed neurodegenerative disorders. Rerurrent Status Epilepticus Sixteen children (17%) had more than 1 episode of status epilepticus (Table 2). The overall Kaplan-Meier estimate of recurrence was 11% at 12 months, 189% at 24 months, and 2096 at 36 and 48 months (Fig 1). 600 Annals of Neurology Vol 31 No 6 June 1792
~
~~
No. of Episodes o f SE
Idiopathic Remote symptomatic Febrile Acute symptomatic Progressive encephalopathy Total
1.0 m
i
n
1
24 18 29 18
23 10 28 16 2 79
6 95
2
23
1 7 1 2 0
0 1 0 0 4
11
5
,
-
-
$! 7 0.4
0
6
12
18 TIME (Months)
24
30
36
-
Fig 1 . Probability of recurrent Jtatus epilepticus following a first episode of status. Kaplan-Meier rurne.
The mean time between the first and second episodes of status was 10.4 months (range, 2-28 months). In 11 (69%) of the 16 children, recurrence occurred within 12 months of the initial episode. Age at the time of the initial episode of status epilepticus did not affect recurrence risk. There were marked differences in recurrence risk according to the etiology of the status epilepticus (see Table 2). Among the 24 children with idiopathic status epilepticus, only 1 experienced a second episode of status. The highest rate of recurrent status epilepticus occurred in the remote symptomatic and progressive encephalopathy groups; 12 ( 5 0 % ) of the 24 children in these two groups experienced recurrent status epilepticus and 5 (2 1%) had 3 or more episodes of status. These children accounted for all 5 patients with multiple (23 ) episodes of status epilepticus ( p < 0.001). The Kaplan-Meier estimates of recurrence for selected groups are shown in Figure 2. The idiopathic group had a cumulative risk of recurrence of 4% at 36 months. The remote symptomatic group had a cumulative recurrence risk of 17% at 12 months, 4573 at 24 months, and 599% at 36 months. T h e relative risk of these two groups was 18.4, with 75% confidence limits of 2.9 and 115.3. The differences were statistically significant in terms of both crude proportions (x2 = 9.7, p < 0.002) and relative risks ( p < 0.005). Of particular interest is that the small group of children with progressive encephalopathy (see Table 2, Fig
1.0
1 0 7 (I)
3
7
09Progressive Encephalopathy(n = 6)
08v)
5
07Remote Symptomatic (n = 18)
06-
Neurologically Abnormal (n = 32)
rT--i--
3
2a 8
05-
04-
-
Neurologically Normal (n = 63)
0
l " " ' 1 ' " " 6 12
I " " " " " ' " ' " ' 7 24 30
18 TIME (Months)
36
F i g 2. Probability of recurrent status epilepticus following a first episode of status in idiopathic, remote symptomatic, and progressive encephalopatby groups. Kapkzn-M eier curve.
F i g 3. Probability of recurrent status epilepticus following an initial episode of status, as a function of neurological status l n o m l versus abnormal). Kaplan-Meier curve.
Table 3. Recurrent Status Epilepticus (SE) and Neurological Status
months in the neurologically normal group, compared with 27%, 48%, and 58% at 12, 24, and 36 months, respectively, in the neurologically abnormal group. The relative risk of recurrence for the neurologically abnormal children was 23.7 with 9595 confidence limits of 6.8 and 82.8 ( p < 0.001). The risk of recurrence in the neurologically abnormal group was related to the degree of neurological impairment though the numbers were too small to permit detailed subgroup analysis. Seventeen children had severe involvement affecting both cognitive and motor development, including 6 with progressive encephalopathy, 7 with both mental retardation and cerebral palsy, and 4 with known genetic syndromes or multiple congenital malformations. Twelve (7 1%) of these children experienced recurrent status epilepticus. In contrast, 15 children had only mental retardation (n = 8) or cerebral palsy (n = 7 ) of varying degrees of severity and only 2 (13%) of them had a second episode of status ( p < 0.005).
Idiopathic Remote symptomatic
Normal
Abnormal
2 2
2 2 n Episodes
n
n
Episodes
24
24
1
18
0 0
0 0 1
8
8 Febrile
29
24
0
1
5 Acute symptomatic
18
15
1
1
3 Progressive encephalopathy Total
6 95
0
4
0
63 2
6 3
14
2
2 ) appeared to have a very high recurrence risk. The cumulative risk of recurrence of this group was already 78% at 12 months. Furthermore, they accounted for 4 of the 5 children with 3 or more episodes of status epilepticus.
Recuwence Risk and Netlvological Status When the recurrence risk was analyzed by neurological status (Table 3), abnormal neurological status was a powerful predictor of recurrence in all groups. Neurologically abnormal children who constituted 34% of the srudy population accounted for 88% of the recurrent status group ( p < 0.001) and for all 5 children with multiple (23) episodes of status epilepticus ( p < 0.001). The Kaplan-Meier estimates of recurrence in these two groups are shown in Figure 3. The estimated risk of recurrent status epilepticus is 3% at 12, 24, and 36
Treatment Eight (SO%) of the 16 children with recurrent status epilepticus, including S of the 8 children in the remote symptomatic group, had prior unprovoked seizures. Ten (63%) of the 16 children with recurrent status epilepticus were already receiving antiepileptic drug therapy at the time of their first episode of status epilepticus. At the time of their second episode, 15 (94%) of the 16 children, including all 14 neurologically abnormal children, were receiving antiepileptic medications. The 1 child who was not receiving medication at the time of the second episode of status epilepticus was an otherwise normal child with acute symptomatic status. It should be noted that drug levels were not being monitored as part of this observational study and the levels may well have been subtherapeutic in some children. Shinnar et al: Recurrent Status Epilepticus
601
Morbidity and Mortality The 3 deaths within 3 months of the occurrence of the index episode of status epilepticus, all of which occurred in the acute symptomatic and the progressive encephalopathy groups, were reported previously 191. They included a 2-month-old with status epilepticus secondary to hypoxic-ischemic encephalopathy associated with cardiac surgery who died 3 days after the episode of status, a 5-month-old with status epilepticus secondary to an intracranial hemorrhage associated with disseminated intravascular coagulation who died 3 days after admission, and a 3-year-old with status associated with a malignant retinoblastoma with central nervous system spread who died 10 days after the episode of status. None of the deaths could be clearly attributed to the status epilepticus per se. These children were not included in this study. Of the 95 children in this study, l child with a progressive neurodegenerative disorder died 13 months after the first and only episode of status epilepticus. She had intractable seizures. None of the 16 children with a history of 2 or more episodes of status epilepticus died during the follow-up period. Moreover, no child had a measurable deterioration in neurological function following a subsequent episode of status epilepticus.
Discussion The striking finding in this study was the high risk of recurrent status epilepticus in children who were neurologically abnormal. The neurologically abnormal group, which constituted 34% of the children with status epilepticus, accounted for 8895 of the children with recurrent status epilepticus and all 5 children with multiple episodes of status. The risk was related to the severity of the neurological abnormality, with multiply handicapped children being at highest risk. The size of this cohort precluded determination of the precise degree of disability necessary to substantially increase the recurrence risk. These recurrent episodes of status epilepticus occurred despite the fact that all 14 neurologically abnormal children were on antiepileptic drug therapy at the time of recurrence. In contrast, the risk of recurrent status epilepticus in neurologically normal children was only 3%) and none of these children experienced more than 2 episodes of status epilepticus during the study. While the increased risk of recurrence in neurologically abnormal children was expected, the magnitude of the increase, with relative risks being greater than 23, was surprising. Few data regarding the risk of recurrent status epilepticus are available in the literature. Four published studies of status epilepticus in children [ l , 4, 12, 131 addressed the issue briefly or not at all. Dunn [ l Z ) reported that 12 (12%) of 97 children with at least 1 episode of status epilepticus experienced recurrent sta-
602
Annals of Neurology Vol 31 No 6 June 1992
tus epilepticus. Cavazzuti and associates 1131 reported that 12 (18%) of 66 children had recurrent status epilepticus. In an older study, Aicardi and Chevrie I l l reported that of 143 children whose long-term outcome was known, 15 (10.55%)had at least 1 subsequent episode of status epilepticus. These studies provided no data regarding the neurological status of the children with recurrent status epilepticus nor how they were different from the children who did not have a recurrence. No data on treatment with antiepileptic drugs were provided. The exception is one recent abstract describing a retrospective chart review of children with status epilepticus, which also concluded that neurologically abnormal children are at high risk for recurrent status epilepticus {25}. The literature regarding adults with status epilepticus is similarly devoid of data regarding the risk of multiple episodes of status epilepticus [ 5 , 63. Support for the low risk of recurrent status epilepticus in neurologically normal children can be deduced from studies of children and adults who present with a first unprovoked seizure [7, 8, 26, 271. In a recent study of children with a first unprovoked seizure [7}, we reported that the risk of any subsequent seizures in neurologically normal children who presented with status epilepticus as their first unprovoked seizure was no different than the risk of subsequent seizures in children who presented with a brief initial seizure. None of the children who presented with status epilepticus as their first unprovoked seizure had a subsequent episode of convulsive status epilepticus. In another study of children and adults who presented with a first unprovoked seizure, Hauser and coworkers [ 8 ) also found that the recurrence risk in subjects with an idiopathic first seizure was no different in those who presented with status epilepticus as their first unprovoked seizure than in those who presented with a briefer initial seizure. In contrast, Hauser and associates [26] found that in subjects with remote symptomatic seizures, the occurrence of status epilepticus as the first seizure is a significant predictor of future seizures. They did not report on the risk of recurrent status epilepticus. Additional evidence for the increased susceptibility of neurologically abnormal children to experience status epilepticus can be deduced from the high proportion of children in the original study who were neurologically abnormal. In a general epilepsy population, 30% of children would be expected to be neurologically abnormal [28, 291. In studies of children with a first unprovoked seizure, the proportion of neurologically abnormal children is even lower 17, 8, 271. In contrast, more than 40% of all children with status epilepticus are neurologically abnormal [ 1, 6, 12, 2 5 1. The question of whether to initiate long-term antiepileptic drug therapy following an initial episode of
status epilepticus is of great importance. In making long-term treatment decisions, the decision is based in part on the risk of future seizures {7, 26, 27, 301. The major concern, however, would be regarding the probability of subsequent life-threatening seizures such as status epilepticus. In our study of children who presented with a first unprovoked seizure {71, we argued that the majority of children who present with a first unprovoked seizure do not require long-term therapy. In fact, we have not treated children with an idiopathic first seizure with long-term antiepileptic drug therapy even when the initial episode was status epilepticus 173. The recurrence risk in that cohort was no different than in children with a brief initial seizure. More importantly, none experienced another episode of convulsive status epilepticus. The results of this study of children with status epilepticus confirm that the risk of subsequent episodes of status epilepticus in neurologically normal children is extremely low. The low morbidity and mortality of status epilepticus in this population, reported in our previous article [9], are also emphasized in this long-term follow-up study. The neurologically abnormal children had a high rate of recurrent status epilepticus despite the fact that all these children were taking antiepileptic drugs at the time of the second episode of status epilepticus. At the time of the first episode of status epilepticus, many were already receiving antiepileptic drug therapy due to prior seizures. In this observational study, we did not attempt to influence the course of treatment and drug levels were not routinely measured. Thus, this is not a study of the efficacy of antiepileptic drug therapy in this population but an intent-to-treat analysis. However, it should be noted that these were neurologically abnormal children who had already experienced an episode of status epilepticus. Therefore, these are the children in whom the neurologists were most aggressive in their therapy. Presumably, these would also be the children for whom the parents would be most concerned about the chances of further seizures in general and of status epilepticus in particular and would therefore be more likely to be compliant. This high-risk subgroup of children with recurrent status epilepticus underscores the need for better treatment strategies in neurologically abnormal children with seizure disorders.
Conclusions Approximately 1 in 6 children with an episode of status epilepticus will experience at least 1 more episode within a few years of the initial episode. The morbidity and mortality of status epilepticus in this population are low. The risk of recurrent status epilepticus in neurologically normal children is very low. In contrast, the risk of recurrent status epilepticus in neurologically abnormal children approaches 50%, even during the limited follow-up period of this study. Neurologically
abnormal children account for almost all children with multiple episodes of status epilepticus. The data suggest that status epilepticus in neurologically normal children is, by and large, an isolated event that should not unduly influence decisions regarding further therapy. The data also underscore the need for effective treatment strategies for neurologically abnormal children with status epilepticus. Larger-scale populationbased studies are needed to better define the precise recurrence risks in this group of patients. This study was supported in part by a teacher investigator development award (1 KO7 NS00930) and a grant (1 R01 NS26151) (ro Dr Shinnar) from the National Institute of Neurological Disorders and Stroke and a Merrit-Putnam fellowship from the Epilepsy Foundation of America (to Dr Maytal). We thank all of the fellows and faculty in the Division of Pediatric Neurology, and the Directors of the Pediatric Emergency Rooms and Intensive Care Units at Montefiore Medical Center, Bronx Municipal Hospital Center, and North Central Bronx Hospital for their help in identifying and recruiting the subjects and for providing their acute medical care. We also acknowledge the cooperation of the New York City Health and Hospital Corporation and Bronx Municipal Hospital Center and North Central Bronx Hospital. We thank Drs Anne T. Berg and Katherine D. Freeman for statistical consultation and Ms Yael Ptachewich for editorial assistance. Lisa Pistorino and Brenda Colon provided computer support. Presented in part at the annual meeting of the American Epilepsy Society, Boston, MA, December 2-7, 1989.
References 1. AicardiJ, ChevrieJJ. Convulsive status epilepticus in infants and in children: a study of 239 cases. Epilepsia 1970;11:187-197 2. Aicardi J, Chevrie JJ. Consequences of status epilepticus in infants and children. Adv Neurol 1983;34:115-125 3. Oxbury JM, Whitty CW. Causes and consequences of status epilepticus in adults: a study of 86 cases. Brain 1971;94: 7 33- 744 4. Fujiwara T, Ishida S, Mlyakoshi M. Stams epilepticus in childhood: a retrospective study of initial convulsive status and subsequent epilepsies. Folia Psychiatr Neurol Jpn 1979;33:337-344 5. Hauser WA. Status epilepticus, frequency, etiology and neurological sequelae. Adv Neurol 1983;34:3-14 6. Hauser WA. Status epilepticus: epidemiologic considerations. Neurology 1990;4O(suppl 2):9- 13 7. Shinnar S, Berg AT, Moshe SL, et al. The risk of seizure recurrence following a first unprovoked seizure in childhood: a prospective study. Pediatrics 1990;85:1076-1085 8. Hauser WA, Anderson VE, Loewenson RB.Seizure recurrence after a first unprovoked seizure. N Engl J Med 1982;307: 522-528 9. Maytal J, Shinnar S, Moshe SL, Alvarez LA. Low morbidity and mortality of status epilepticus in children. Pediatrics 1989;83: 323-331 10. Dulac 0,Aubourg P, Chercoury A, et al. Infantile status epilepticus: clinical, etiological and prognostic aspects. Rev Electroencephalogr Neurophysiol 1985;14:255-262 11. Goulon M, L'evy-Alcover MA, Nouaihat F. Status epilepticus in the adult: epidemiologic and clinical study in an intensive care unit. Rev Electroencephalogr Neurophysiol 1985;14:277-285 12. Dunn WD. Status epilepticus in children: etiology, clinical features and outcome. J Child Neurol 1988;3:167-173
Shinnar et al: Recurrent Status Epilepticus
603
13. Cavazzuti GB, Ferrari P, Lalla M. Follow-up study of 482 cases with convulsive disorders in the first year of life. Dev Med Child Neurol 1984;26:425-437 14. Commission on Classification and Terminology of the International League against Epilepsy. Proposal for revised clinical and electroencephalographic classification of epileptic seizures. Epilepsia 1981;22:489-50 1 15. Nelson KB, Ellenberg JH. Prognosis in children with febrile seizures. Pediatrics 1978;61:720-727 16. National Institutes of Health. Febrile seizures: consensus development conference summary, vol 3, no 2. Bethesda, MD: National Institutes of Health, 1980 17. Maytal J, Shinnar S. Febrile status epilepticus. Pediatrics 1990;86:61 1-6 16 18. Zar JH. Biostatistical analysis. 2nd ed. Englewood Cliffs, NJ: Prentice-Hall, 1984 19. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958;53:457-481 20. Kalbfleisch JD, Prentice RL. The statistical analysis of failure time data. New York: John Wiley, 1980:70-142 2 1. Mantel N. Evaluation of survival data and two new rank order statistics arising in its consideration. Cancer Chemother Rep 196630: 163-1 70
604 Annals of Neurology Vol 31 No 6 June 1992
22. Cox DR. Regressive models and life-tables. J R Stat Sot: [Bl 1972;34:187-220 23. SAS user’s guide: statistics. Version 5 ed. Cary, NC: SAS Institute, 1985 24. SUGI supplemental library user’s guide. Version 5. Cary, NC: SAS Institute, 1986 25. Driscoll SM, Towne AR, Pellock JM, DeLorenzo RJ. Recurrent status epilepticus in children. Neurology 1990;40(suppl I ):I97 (Abstract) 26. Hauser WA, RJch SS, Annegers JF, Anderson VE. Seizure recurrence following a first unprovoked seizure: an extended follow-up. Neurology 1990;40:1163-1 170 27. Berg AT, Shinnar S. The risk of seizure recurrence following a first unprovoked seizure: a quantitative review. Neurology 1991;41965-972 28. Hauser WA, Kurland LT. The epidemiology of epilepsy in Rochester, Minnesota, I935 through 1967. Epilepsia 1975; 16:l-66 29. Hauser WA, Hesdorffer DC. Epilepsy: frequency, causes and consequences. New York: Demos, 1990 30. Shinnar S. Treatment decisions in childhood seirures. In: Dodson WE, Pellock JM, eds. Pediatric epilepsy: diagnosis and thcrapy. New York: Demos, 1991:215-221
