Predictors of Acute Symptomatic Seizures After Intracranial Hemorrhage in Infants* Seema Bansal, MD1; Tewodros Kebede, MD2; Nathan P. Dean, MD2; Jessica L. Carpenter, MD2
Objective: To determine the prevalence of acute symptomatic seizures in infants with supratentorial intracranial hemorrhage, to identify potential risk factors, and to determine the effect of acute seizures on long-term morbidity and mortality. Design: Children less than 24 months with intracranial hemorrhage were identified from a neurocritical care database. All patients who received seizure prophylaxis beginning at admission were included in the study. Risk factors studied were gender, etiology, location of hemorrhage, seizure(s) on presentation, and the presence of parenchymal injury. Acute clinical and electrographic seizures were identified from hospital medical records. Subsequent development of late seizures was determined based on clinical information from patients’ latest follow-up. Setting and Patients: Patients with idiopathic neonatal intracranial hemorrhage, premature infants, and those with prior history of seizures were excluded from analysis. Seventy-two infants met inclusion criteria. Interventions: None. Measurements and Main Results: Forty percent of infants had acute symptomatic seizures. The prevalence was similar regardless of whether etiology of hemorrhage was traumatic or nontraumatic. Seizures on presentation and parenchymal injury were independent risk factors of acute seizures (p = 0.001 and p = 0.006, respectively). Younger children and women were also at higher risk (p < 0.05). Twenty percent had electrographic-only seizures, and those with parenchymal injury trended toward an increased risk (p < 0.1). Acute seizures were not predictive of mortality, but nearly twice as many patients with acute seizures developed late seizures when compared with those without. Electrographic sei*See also p. 781. 1 Ann and Robert H. Lurie Children’s Hospital of Chicago, Chicago, IL. 2 Children’s National Medical Center, Washington, DC. This work was performed at Children’s National Medical Center, 111 Michigan Ave. NW, Washington, DC 20010. Supported, in part, by department funds. The authors have disclosed that they do not have any potential conflicts of interest. Address requests for reprints to: Seema Bansal, MD, Epilepsy Center, 225 E. Chicago Ave, Box 29, Chicago, IL 60611. E-mail: [email protected] Copyright © 2014 by the Society of Critical Care Medicine and the World Federation of Pediatric Intensive and Critical Care Societies DOI: 10.1097/PCC.0000000000000221
750
www.pccmjournal.org
zures and parenchymal injury were also predictive of development of late seizures (p < 0.001 and p = 0.013, respectively). Conclusions: Despite seizure prophylaxis, infants with supratentorial intracranial hemorrhage are at high risk for acute symptomatic seizures. This is regardless of the etiology of hemorrhage. Younger patients, women, patients with parenchymal injury, and patients presenting with seizure are most likely to develop acute seizures. Although the benefits of seizure prophylaxis have not been studied in this specific population, these results suggest that it is an important component of acute care following intracranial hemorrhage. (Pediatr Crit Care Med 2014; 15:750–755) Key Words: epilepsy; infants; intracranial hemorrhage; risk factors; seizure; symptomatic
S
ymptomatic seizures, defined as those that occur in response to an identifiable CNS insult, are a known complication of various types of brain injury, including both traumatic and nontraumatic etiologies (1, 2). Seizures are further demarcated as acute (within 7 d of injury) or late (beyond 7 d after injury) (3). It is well established that children are at higher risk of acute symptomatic seizures than adults (4). Many studies have examined the epidemiology of post-traumatic seizures in children and have confirmed that younger age groups are at increased risk (5, 6). Chiaretti et al (5) found that 22.5% of children less than 3 years with head injury developed seizures within 1 week of injury; risk factors included severity of injury and the presence of cerebral edema. Less is known about the epidemiology of nontraumatic etiologies. Recently, Beslow et al (1) studied seizures within 7 days of nontraumatic, spontaneous intracranial hemorrhage (ICH) in children. They estimated a frequency of 13% among all age groups and identified elevated intracranial pressure requiring acute intervention as a risk factor. Acute seizures can exacerbate underlying brain injury (7) and may lead to subsequent development of epilepsy in both adults and children (8, 9). In adults, seizure prophylaxis is a level II recommendation after traumatic brain injury (TBI) (10). Studies in adults have supported its efficacy after subarachnoid hemorrhage and supratentorial neurosurgery as well (11, 12). In children, there are no guidelines for seizure October 2014 • Volume 15 • Number 8
Neurocritical Care
prophylaxis after nontraumatic brain injuries, and even in traumatic injury it is only a level III guideline (3). In 2008, our institution introduced a protocol, which included seizure prophylaxis in any child with supratentorial ICH. This recommendation was made based on clinical experience as the epidemiological features of acute symptomatic seizures after ICH in children, particularly infants, is incomplete. In this article, we aimed to 1) determine the frequency of acute symptomatic seizures in infants with supratentorial ICH receiving seizure prophylaxis, 2) identify clinical features predictive of seizure occurrence, and 3) determine the effect of acute seizures on long-term morbidity and mortality. To accomplish this, we analyzed a retrospective cohort of infants with ICH. To our knowledge, this is one of the largest cohorts of infant ICH at a single tertiary care center.
METHODS This is a retrospective cohort study of infant subjects (> 2 wk and < 24 mo), who were diagnosed with supratentorial ICH between June 2008 and February 2013 at a tertiary care center. This study was approved by the Children’s National Medical Center Institutional Review Board (Pro00001916). Neonates with idiopathic ICH were excluded, as were premature infants and patients with a history of seizures. All patients in our sample were started on seizure prophylaxis with antiepileptic drugs (AEDs) at the time of diagnosis of ICH, per institution protocol. AEDs were chosen according to physician discretion. Levetiracetam was administered as a 20 mg/kg bolus followed by a maintenance dose of 20–30 mg/kg/d. (Fos)phenytoin and phenobarbital were administered as a loading dose of 20 mg/kg followed by maintenance of 5 mg/kg/d, with additional boluses and/or adjustment of maintenance based on drug level monitoring. (Fos)phenytoin levels were corrected for hypoalbuminemia. Supratentorial ICH was defined as the presence of any blood products within the skull above the posterior fossa. ICH was diagnosed by either head CT or MRI. The primary outcome studied was acute symptomatic seizure(s), defined as those that occurred during the first 7 days after ICH and after starting seizure prophylaxis. Seizures on presentation, defined as those that occurred as the presenting symptom or in association with the presenting symptoms, were documented as separate events. All clinical data were collected from a neurocritical care database, as well as inpatient and outpatient electronic medical records. All neuroimaging during the first 7 days of admission was reviewed for all patients. Risk factors studied were age, gender, etiology (traumatic or nontraumatic), location of hemorrhage, seizure(s) on presentation, and the presence of parenchymal injury on imaging. Location of ICH was characterized as extra-axial (including intraventricular, subarachnoid, subdural, and epidural hemorrhages), intra-axial (intraparenchymal), or both. Parenchymal injury was defined as any acute pathology involving supratentorial gray or white matter. Examples of hemorrhagic parenchymal injury include contusions and diffuse axonal injury, which were also considered intra-axial ICH. Nonhemorrhagic parenchymal injury included ischemia/infarction and edema. Pediatric Critical Care Medicine
Both clinical and electrographic-only seizures were included in the analysis. As per our institution’s clinical pathway, patients admitted to the PICU with acute encephalopathy are monitored with continuous video electroencephalogram for 48 hours or until mental status improves. Clinical seizures included any stereotyped motor or autonomic manifestations identified as seizure by medical staff and thus warranting treatment with anticonvulsants. For those patients on electroencephalogram monitoring, only clinical events with corresponding ictal change on electroencephalogram were included. Electrographic-only seizures were defined as rhythmic ictal discharges on electroencephalogram lasting at least 10 seconds without clinical correlate seen on the video. Secondary outcomes studied were mortality before discharge and development of late seizures (beginning > 7 d after ICH). Electronic medical records of all outpatient neurology visits from time of discharge to present were reviewed. At each visit, patients were assessed for seizure occurrence. In those with seizures, date of occurrence and the use of AEDs were recorded. In addition, all emergency department visits, hospital admissions, and telephone encounters were reviewed to ensure complete identification of late seizures. Statistical methods included contingency table analysis to evaluate bivariate relationships. Multiple logistic regression analysis was used to evaluate the effects of multiple risk factors on seizure risk. All tests of statistical significance were based on two-tailed p values.
RESULTS Between June 2008 and February 2013, 72 infants were diagnosed with supratentorial ICH. All patients were started on seizure prophylaxis at the time of diagnosis. Prophylaxis was provided with levetiracetam in 41 patients (57%), (fos)phenytoin in 26 patients (36%), and phenobarbital in five patients (7%). Twenty-nine patients (40%) had acute seizures. The prevalence was similar in both traumatic and nontraumatic etiologies (40% and 42%, respectively). Of the patients on levetiracetam, 12 of 41 (29%) had seizures when compared with 2 of 5 patients (40%) on phenobarbital and 15 of 26 patients (58%) on (fos)phenytoin. These differences were not statistically significant (p = 0.07). On average, seizures occurred on hospital day 1.9 (sd, 1.6). Twenty-four patients (83%) required two or more AEDs for seizure control. Eleven patients (38%) required a continuous infusion of benzodiazepines and/or barbiturates for seizure cessation. Clinical and demographic data of patients with and without acute seizures as well as the entire cohort are represented in Table 1. Mean age of all patients was 7 months (sd, 5.7). Patients who experienced acute seizures tended to be younger, with a mean age of 4.5 months (sd, 4.8) when compared with 8.6 months (sd, 5.7) in patients who did not have seizures. In both groups of patients, the most common etiology of hemorrhage was TBI, and the most common location of hemorrhage was extra-axial. Over twice as many patients with acute seizures had seizures on presentation and/or parenchymal injury when compared with patients who did not have acute seizures. www.pccmjournal.org
751
Bansal et al
Table 1. Cohort Demographics and Association of Hemorrhage Characteristics With Acute Symptomatic Seizures Predictor
All Patients (n = 72)
Acute Seizure(s) (n = 29)
No Acute Seizure (n = 43)
p
4.5 ± 4.8
8.6 ± 5.7
0.001a
Age (mo), n ± sd
7 ± 5.7
Gender: female, n (%)
20 (28)
12 (41)
8 (19)
0.03a
Coagulopathyb
4 (5.6)
2 (6.9)
2 (4.7)
0.68
Traumatic brain injury
60 (83)
36 (84)
0.91
Vasculopathy
7 (9.7)
3 (10.3)
4 (9.3)
0.88
Other
1 (1.4)
0 (0)
1 (2.3)
0.41
21 (72)
34 (79)
0.51
5 (6.9)
1 (3.5)
4 (9.3)
0.34
Both
13 (18)
7 (24)
6 (14)
0.24
Seizure on presentation
31 (43)
20 (69)
11 (26)
0.0003a
Parenchymal injury
40 (56)
24 (83)
16 (37)
0.0001a
Etiology, n (%)
c
24 (83)
Location of intracranial hemorrhage, n (%) Extra-axial Intra-axial
52 (72)
Statistically significant result. Etiologies of coagulopathy included vitamin K deficiency and the use of IV heparin during extracorporeal membrane oxygenation. c Etiologies of vasculopathy included arteriorvenous malformations, aneurysms, and cavernomas. a
b
Several demographic and clinical factors were significantly associated with acute symptomatic seizures in bivariate analyses (Table 1). Relative risk was increased significantly in younger children, women, patients with seizures on presentation, and those with parenchymal injury. Etiology and location of ICH were not predictive of acute seizures. After controlling for age differences, seizures on presentation and parenchymal injury remained statistically significant individual predictors of acute symptomatic seizures (p < 0.001 for both). Multivariate analysis of risk factors with p value of less than 0.05 showed that parenchymal injury (OR, 8.1; p = 0.007) and age (OR, 7.2; p = 0.02) remained strong independent risk factors. Patients with seizures on presentation Table 2.
and women continued to trend toward increased risk (OR, 3.3 and 2.7, respectively) although results were not significant. Over 80% of patients had TBI, and of these, 60% were due to presumed nonaccidental trauma (NAT) (Table 2). Furthermore, 37% of patients with NAT had associated diffuse anoxic or ischemic brain injury, whereas no patients with alternate causes of TBI did; this difference was statistically significant (p < 0.001). NAT and associated anoxic or ischemic injury were independently associated with an increased likelihood of acute seizures (p < 0.01). The patients age with NAT did not differ significantly from those with other causes of TBI. Twenty-five percent of patients with TBI had elevated
Cohort Demographics of Patients With Traumatic Brain Injury All Patients (n = 60)
Predictor
Acute Seizure(s) (n = 24)
No Acute Seizure (n = 36)
p
0.003a
Etiology, n (%) Nonaccidental trauma,
36 (60)
20 (83)
16 (44)
Other traumatic brain injury
24 (40)
4 (17)
20 (56)
Present
13(22)
10 (42)
3 (8.3)
Absent
47 (78)
14 (48)
33 (92)
Elevated
13 (22)
6 (25)
7 (19)
Normal
47 (78)
18 (75)
29 (81)
Diffuse ischemia/anoxic injury, n (%) 0.002a
Intracranial pressure, n (%)
a
0.610
Statistically significant result.
752
www.pccmjournal.org
October 2014 • Volume 15 • Number 8
Neurocritical Care
Table 3.
Bivariate Analysis of Risk Factors for Electrographic-Only Seizures
Predictor
Relative Risk
95% CI
p
Gender: female
1.26
0.69–2.27
0.46
Traumatic brain injury
0.99
0.71–1.4
0.95
Location of intracranial hemorrhage
1.23
0.57–2.67
0.56
Seizure on presentation
1.08
0.55–2.14
0.82
Parenchymal injury
3.33
0.56–19.7
0.054
intracranial pressure requiring intervention; however, this was not associated with acute seizures. Forty-five patients (63%) were encephalopathic at admission and therefore monitored with electroencephalogram; duration of monitoring varied based on clinical indication. This included 28 of the 29 patients with acute seizures. One male patient with a normal neurological examination had an acute clinical seizure; his AED dose was increased and had no further events; therefore, electroencephalogram was not done. Of all patients with acute seizures, 59% had electrographic-only seizures during hospitalization; 21% had electrographic-only seizures exclusively. Thirteen patients (77%) with electrographic-only seizures had seizures as a presenting symptom, and 11 (65%) experienced seizures within the first 24 hours of hospitalization. Bivariate analysis suggested that patients with parenchymal injury remained at increased risk of electrographic-only seizures, with results trending toward significance (Table 3). Seven patients (9.7%) died during hospitalization. Bivariate analyses of potential risk factors for mortality before hospital discharge are shown in Table 4. TBI was associated with increased risk of mortality (p < 0.01). Neither acute seizures nor electrographic-only seizures were associated with increased mortality. Follow-up data were available on 63 of 65 surviving patients (97%). Mean duration of follow-up after ICH was 12.4 months (1 mo to 4.6 yr). Eleven patients (15%) experienced late seizures and thus required ongoing treatment with AEDs at their latest follow-up visit. This included 6 of 29 patients (21%) with acute seizures and 5 of 43 patients (12%) without acute seizures. Bivariate analyses (Table 5) suggested an increased risk of developing late seizures in women and patients with seizures on presentation, although these results did not reach statistical significance, the trend is similar to that seen for acute seizures. Similarly, patients with acute seizures had an increased risk of late seizures. Parenchymal injury and electrographic-only seizures were significantly associated with late seizures. Eleven of 33 patients with parenchymal injury developed late seizures when compared with 0 of 32 patients without parenchymal injury (p < 0.001). Results were similar for electrographic-only seizures; all patients with late seizures had experienced electrographic-only seizures during hospitalization (p = 0.013).
DISCUSSION Acute symptomatic seizures are a serious complication of intracranial hemorrhage and are most often studied in TBI. In Pediatric Critical Care Medicine
pediatric patients with TBI, the frequency is estimated between 8.4% and 35%, with younger children at greatest risk (5, 6, 13). In this article, we specifically studied this high-risk population and sought to identify risk factors associated with acute symptomatic seizures after supratentorial ICH of any etiology. To establish the prevalence of acute symptomatic seizures truly, we would need to compare children who received prophylaxis to those who did not; however, seizure prophylaxis for ICH has been an institutional protocol since 2007. In infants receiving prophylaxis, we found an overall prevalence of 40%. This is regardless of whether hemorrhage is traumatic or nontraumatic. Furthermore, in bivariate analysis, TBI was not an independent risk factor for acute seizures, thus hemorrhage of various etiologies may be associated with acute seizures. Women had a trend toward increased risk for acute symptomatic seizures; gender appeared to be an important risk factor in bivariate analysis, but significance was lost in multivariate analysis. The effect of gender on acute post-traumatic seizures and long-term outcomes has been extensively studied (5, 14–17), although results are conflicting, sex hormones undoubtedly influence the brain. Increased levels of peripheral sex steroids following injury are associated with increased mortality and worse long-term outcomes (18), whereas progesterone has been found to be neuroprotective after a variety of CNS insults (19) and may be a potential therapeutic agent after TBI (20). Thus, the influence of gender on brain injury remains unclear but may affect various aspects of clinical course. Another important risk factor is the presence of parenchymal injury, not only for acute seizures but also for late seizures and possibly electrographic seizures as well. Interestingly, however, the location of hemorrhage is not a risk factor. In fact, most patients with acute seizures had extra-axial hemorrhages, specifically subdural hematomas. This suggests that it is not the hemorrhage itself that is epileptogenic, but rather the associated edema and/or ischemia of adjacent tissue. Previous studies have shown that cerebral edema is a risk factor for acute post-traumatic seizures (5, 13), as well as poor outcome (21). We now find that approximately half of all patients with acute seizures have evidence of hypoxia and/or ischemia on neuroimaging. Cerebral ischemia is known to alter gamma-aminobutyric acid (GABA)ergic and glutamatergic neurotransmission leading to neuronal hyperexcitability (22); thus, the increased epileptogenicity in these patients is not surprising. We found a high prevalence of electrographic-only seizures, which is consistent with previous studies. Hasbani et al www.pccmjournal.org
753
Bansal et al
Table 4.
Bivariate Analysis of Risk Factors for Mortality
Predictor
a
p
Relative Risk
95% CI
Gender: female
0.93
0.76–1.12
0.35
Traumatic brain injury
1.43
0.95–2.13
0.0025a
Seizure on presentation
1.00
0.86–1.16
0.99
Acute symptomatic seizure
0.93
0.78–1.10
0.34
Electrographic seizures
0.90
0.68–1.18
0.47
Statistically significant result.
(23) found that 57% of infants, median age of 4 months, with abusive head trauma had electrographic seizures. A study of children with TBI found that only 16% had electrographic seizures; however, this study included all children up to 18 years (24). Additional studies of continuous electroencephalogram in pediatric patients with varying pathologies/etiologies have found similarly high rates of electrographic seizures (25, 26). Electrographic seizures are significant in that they may herald impending decompensation; in adults, electrographic seizures in the setting of ICH are associated with expansion of hemorrhage (7). Furthermore, although we found that the majority of electrographic seizures occurred in the first day of monitoring, time to seizure onset can vary based on etiology of hemorrhage and evolution of pathology. Thus, continuous electroencephalogram monitoring remains a necessary component of initial ICU care following ICH. Its use, however, goes far beyond seizure detection. It affects various aspects of medical management and can serve as a tool for prognostication as well (27, 28). The choice of prophylactic agent may affect the occurrence of acute seizures. Fewer patients treated with levetiracetam had acute seizures when compared with those treated with (fos) phenytoin or phenobarbital. One possible explanation for this is failure to maintain therapeutic levels with the latter anticonvulsants. Although attempts were made during hospitalization to maintain therapeutic levels based on drug level monitoring, we do not have data to confirm that patients were at therapeutic levels at the exact time when seizures occurred in this study. A second potential cause of the discrepant results may be accounted for by patient age. There was a tendency for younger children to be given (fos)phenytoin as their initial prophylactic agent; therefore, the apparent advantage of levetiracetam may in fact be because of the older patient population on this drug. Studies in adult populations comparing levetiracetam with phenytoin have demonstrated similar efficacies with Table 5.
better long-term outcomes and fewer adverse events in patients treated with levetiracetam (11, 12). In children, levetiracetam was found to be superior to (fos)phenytoin in preventing acute symptomatic seizures following hematopoietic stem cell transplantation (29) and comparable with (fos)phenytoin in children with intracranial hemorrhage (30). Although comparison of AEDs was not the primary outcome of this study, our findings suggest that there may be a difference. additional studies are needed to evaluate optimal prophylactic agents in infants and children, including defining optimal dosing regimens for levetiracetam as well as standardized methods of maintaining therapeutic drug levels. The effect of both seizure prophylaxis and acute seizures on long-term outcomes is also well established. Acute seizures lead to worse outcomes based on Glasgow Outcome Score (5, 6) and predict the development of post-traumatic epilepsy (9, 31–33). In turn, late seizures are associated with worse neurological and developmental outcomes (32). Seizure prophylaxis significantly reduces mortality in children with head trauma (34). We found that acute seizures do not affect mortality but may lead to late seizures. The risk of late seizures in children with acute seizures was nearly twice those without acute seizures, and these results were trending toward significance. Our estimated prevalence of late seizures is 15%, which is similar to overall estimates of children with TBI but lower than previous estimates in infants (9, 32). In summary, acute symptomatic seizures are a common complication of ICH and have both short- and long-term implications. Despite studies showing that younger children are at particularly high risk for acute seizures, data regarding seizure prophylaxis in infants is lacking. AEDs are known to be protective against acute post-traumatic seizures in children (32) and may also have a role in neuroprotection (36). Thus, seizure prophylaxis is becoming an increasingly used component of ICU care following head trauma. We found that even
Bivariate Analysis of Risk Factors for Late Seizures Relative Risk
95% CI
p
Gender: female
1.61
0.54–4.85
0.40
Traumatic brain injury
0.63
0.17–2.42
0.52
Seizure on presentation
2.31
0.75–7.13
0.13
Acute symptomatic seizure
1.92
0.65–5.63
0.23
Predictor
754
www.pccmjournal.org
October 2014 • Volume 15 • Number 8
Neurocritical Care
with prophylaxis, nearly half of all infants develop acute seizures. Given our findings that non-traumatic etiologies of ICH have a similar risk of acute symptomatic seizures as traumatic etiologies, we recommend that seizure prophylaxis be considered after all causes of acute ICH in infants, particularly those in whom seizure is a presenting sign of hemorrhage and where there is evidence of parenchymal injury on imaging.
CONCLUSIONS Infants less than 24 months have a high risk of acute symptomatic seizures, despite seizure prophylaxis and regardless of the etiology of ICH. Younger age, those with seizures as a presenting symptom, and those with parenchymal injury are at highest risk. This may affect long-term outcomes. The optimal agent for seizure prophylaxis in the acute setting has yet to be determined.
ACKNOWLEDGMENT We thank Robert McCarter, ScD, for his assistance with statistical analysis.
REFERENCES
1. Beslow LA, Abend NS, Gindville MC, et al: Pediatric intracerebral hemorrhage: Acute symptomatic seizures and epilepsy. JAMA Neurol 2013; 70:448–454 2. Beghi E, Carpio A, Forsgren L, et al: Recommendation for a definition of acute symptomatic seizure. Epilepsia 2010; 51:671–675 3. Kochanek P, Carney N, Adelson D, et al: Guidelines for the early medical management of severe traumatic brain injury in infants, children, and adolescents—second edition. Pediatr Crit Care Med 2012; 13:S72–S82 4. Da Silva AM, Willmore LJ: Posttraumatic epilepsy. Handb Clin Neurol 2012; 108:585–599 5. Chiaretti A, De Benedictis R, Polidori G, et al: Early post-traumatic seizures in children with head injury. Childs Nerv Syst 2000; 16:862–866 6. Ateş O, Ondül S, Onal C, et al: Post-traumatic early epilepsy in pediatric age group with emphasis on influential factors. Childs Nerv Syst 2006; 22:279–284 7. Claassen J, Jetté N, Chum F, et al: Electrographic seizures and periodic discharges after intracerebral hemorrhage. Neurology 2007; 69:1356–1365 8. Agrawal A, Timothy J, Pandit L, et al: Post-traumatic epilepsy: An overview. Clin Neurol Neurosurg 2006; 108:433–439 9. Appleton RE, Demellweek C: Post-traumatic epilepsy in children requiring inpatient rehabilitation following head injury. J Neurol Neurosurg Psychiatry 2002; 72:669–672 10. Bratton S, Chestnut R, Ghajar J, et al; Brain Trauma Foundation; American Association of Neurological Surgeons; Congress of Neurological Surgeons; Joint Section on Neurotrauma and Critical Care, AANS/CNS. Guidelines for the management of severe traumatic brain injury. XIII. Antiseizure prophylaxis. J Neurotrauma 2007; 24:S83–S86 11. Milligan TA, Hurwitz S, Bromfield EB: Efficacy and tolerability of levetiracetam versus phenytoin after supratentorial neurosurgery. Neurology 2008; 71:665–669 12. Szaflarski JP, Sangha KS, Lindsell CJ, et al: Prospective, randomized, single-blinded comparative trial of intravenous levetiracetam versus phenytoin for seizure prophylaxis. Neurocrit Care 2010; 12:165–172
Pediatric Critical Care Medicine
13. Hahn YS, Fuchs S, Flannery AM, et al: Factors influencing posttraumatic seizures in children. Neurosurgery 1988; 22:864–867 14. Ong LC, Dhillon MK, Selladurai BM, et al: Early post-traumatic seizures in children: Clinical and radiological aspects of injury. J Paediatr Child Health 1996; 32:173–176 15. Schmidt AT, Hanten GR, Li X, et al: Decision making after pediatric traumatic brain injury: Trajectory of recovery and relationship to age and gender. Int J Dev Neurosci 2012; 30:225–230 16. Gao GY, Jiang JY. Chinese Head Trauma Data Bank: Effect of gender on the outcome of patients with severe traumatic brain injury. J Neurotrauma 2012; 29:96–100 17. Renner C, Hummelsheim H, Kopczak A, et al: The influence of gender on the injury severity, course and outcome of traumatic brain injury. Brain Inj 2012; 26:1360–1371 18. Wagner AK, McCullough EH, Niyonkuru C, et al: Acute serum hormone levels: Characterization and prognosis after severe traumatic brain injury. J Neurotrauma 2011; 28:871–888 19. Stein DG, Wright DW, Kellermann AL: Does progesterone have neuroprotective properties? Ann Emerg Med 2008; 51:164–172 20. Stein DG: Is progesterone a worthy candidate as a novel therapy for traumatic brain injury? Dialogues Clin Neurosci 2011; 13:352–359 21. Bahloul M, Chaari A, Chabchoub I, et al: Outcome analysis and outcome predictors of traumatic head injury in childhood: Analysis of 454 observations. J Emerg Trauma Shock 2011; 4:198–206 22. Schwartz-Bloom RD, Sah R: Gamma-Aminobutyric acid(A) neurotransmission and cerebral ischemia. J Neurochem 2001; 77:353–371 23. Hasbani DM, Topjian AA, Friess SH, et al: Nonconvulsive electrographic seizures are common in children with abusive head trauma. Pediatr Crit Care Med 2013; 14:709–715 24. Arndt DH, Lerner JT, Matsumoto JH, et al: Subclinical early posttraumatic seizures detected by continuous EEG monitoring in a consecutive pediatric cohort. Epilepsia 2013; 54:1780–1788 25. Williams K, Jarrar R, Buchhalter J: Continuous video-EEG monitoring in pediatric intensive care units. Epilepsia 2011; 52:1130–1136 26. Schreiber JM, Zelleke T, Gaillard WD, et al: Continuous video EEG for patients with acute encephalopathy in a pediatric intensive care unit. Neurocrit Care 2012; 17:31–38 27. Gallentine WB: Utility of continuous EEG in children with acute traumatic brain injury. J Clin Neurophysiol 2013; 30:126–133 28. Abend NS, Topjian AA, Gutierrez-Colina AM, et al: Urgent continuous EEG (cEEG) monitoring leads to changes in treatment in half of cases impact of continuous EEG monitoring on clinical management in critically ill children. Neurocrit Care 2011; 15(1):70–75 29. Soni S, Skeens M, Termuhlen AM, et al: Levetiracetam for busulfaninduced seizure prophylaxis in children undergoing hematopoietic stem cell transplantation. Pediatr Blood Cancer 2012; 59:762–764 30. Bansal S, Blalock D, Kebede T, et al: Levetiracetam versus (fos)phenytoin for seizure prophylaxis in pediatric patients with intracranial hemorrhage. J Neurosurg Pediatr 2014; 13:209–215 31. Kieslich M, Jacobi G: Incidence and risk factors of post-traumatic epilepsy in childhood. Lancet 1995; 345:187 32. Kieslich M, Marquardt G, Galow G, et al: Neurological and mental outcome after severe head injury in childhood: A long-term follow-up of 318 children. Disabil Rehabil 2001; 23:665–669 33. Englander J, Bushnik T, Duong TT, et al: Analyzing risk factors for late posttraumatic seizures: A prospective, multicenter investigation. Arch Phys Med Rehabil 2003; 84:365–373 34. Tilford JM, Simpson PM, Yeh TS, et al: Variation in therapy and outcome for pediatric head trauma patients. Crit Care Med 2001; 29:1056–1061 35. Liesemer K, Bratton SL, Zebrack CM, et al: Early post-traumatic seizures in moderate to severe pediatric traumatic brain injury: Rates, risk factors, and clinical features. J Neurotrauma 2011; 28:755–762 36. Belcastro V, Pierguidi L, Tambasco N: Levetiracetam in brain ischemia: Clinical implications in neuroprotection and prevention of poststroke epilepsy. Brain Dev 2011; 33:289–293
www.pccmjournal.org
755
Objective: To determine the prevalence of acute symptomatic seizures in infants with supratentorial intracranial hemorrhage, to identify potential risk factors, and to determine the effect of acute seizures on long-term morbidity and mortality. Design: Children less than 24 months with intracranial hemorrhage were identified from a neurocritical care database. All patients who received seizure prophylaxis beginning at admission were included in the study. Risk factors studied were gender, etiology, location of hemorrhage, seizure(s) on presentation, and the presence of parenchymal injury. Acute clinical and electrographic seizures were identified from hospital medical records. Subsequent development of late seizures was determined based on clinical information from patients’ latest follow-up. Setting and Patients: Patients with idiopathic neonatal intracranial hemorrhage, premature infants, and those with prior history of seizures were excluded from analysis. Seventy-two infants met inclusion criteria. Interventions: None. Measurements and Main Results: Forty percent of infants had acute symptomatic seizures. The prevalence was similar regardless of whether etiology of hemorrhage was traumatic or nontraumatic. Seizures on presentation and parenchymal injury were independent risk factors of acute seizures (p = 0.001 and p = 0.006, respectively). Younger children and women were also at higher risk (p < 0.05). Twenty percent had electrographic-only seizures, and those with parenchymal injury trended toward an increased risk (p < 0.1). Acute seizures were not predictive of mortality, but nearly twice as many patients with acute seizures developed late seizures when compared with those without. Electrographic sei*See also p. 781. 1 Ann and Robert H. Lurie Children’s Hospital of Chicago, Chicago, IL. 2 Children’s National Medical Center, Washington, DC. This work was performed at Children’s National Medical Center, 111 Michigan Ave. NW, Washington, DC 20010. Supported, in part, by department funds. The authors have disclosed that they do not have any potential conflicts of interest. Address requests for reprints to: Seema Bansal, MD, Epilepsy Center, 225 E. Chicago Ave, Box 29, Chicago, IL 60611. E-mail: [email protected] Copyright © 2014 by the Society of Critical Care Medicine and the World Federation of Pediatric Intensive and Critical Care Societies DOI: 10.1097/PCC.0000000000000221
750
www.pccmjournal.org
zures and parenchymal injury were also predictive of development of late seizures (p < 0.001 and p = 0.013, respectively). Conclusions: Despite seizure prophylaxis, infants with supratentorial intracranial hemorrhage are at high risk for acute symptomatic seizures. This is regardless of the etiology of hemorrhage. Younger patients, women, patients with parenchymal injury, and patients presenting with seizure are most likely to develop acute seizures. Although the benefits of seizure prophylaxis have not been studied in this specific population, these results suggest that it is an important component of acute care following intracranial hemorrhage. (Pediatr Crit Care Med 2014; 15:750–755) Key Words: epilepsy; infants; intracranial hemorrhage; risk factors; seizure; symptomatic
S
ymptomatic seizures, defined as those that occur in response to an identifiable CNS insult, are a known complication of various types of brain injury, including both traumatic and nontraumatic etiologies (1, 2). Seizures are further demarcated as acute (within 7 d of injury) or late (beyond 7 d after injury) (3). It is well established that children are at higher risk of acute symptomatic seizures than adults (4). Many studies have examined the epidemiology of post-traumatic seizures in children and have confirmed that younger age groups are at increased risk (5, 6). Chiaretti et al (5) found that 22.5% of children less than 3 years with head injury developed seizures within 1 week of injury; risk factors included severity of injury and the presence of cerebral edema. Less is known about the epidemiology of nontraumatic etiologies. Recently, Beslow et al (1) studied seizures within 7 days of nontraumatic, spontaneous intracranial hemorrhage (ICH) in children. They estimated a frequency of 13% among all age groups and identified elevated intracranial pressure requiring acute intervention as a risk factor. Acute seizures can exacerbate underlying brain injury (7) and may lead to subsequent development of epilepsy in both adults and children (8, 9). In adults, seizure prophylaxis is a level II recommendation after traumatic brain injury (TBI) (10). Studies in adults have supported its efficacy after subarachnoid hemorrhage and supratentorial neurosurgery as well (11, 12). In children, there are no guidelines for seizure October 2014 • Volume 15 • Number 8
Neurocritical Care
prophylaxis after nontraumatic brain injuries, and even in traumatic injury it is only a level III guideline (3). In 2008, our institution introduced a protocol, which included seizure prophylaxis in any child with supratentorial ICH. This recommendation was made based on clinical experience as the epidemiological features of acute symptomatic seizures after ICH in children, particularly infants, is incomplete. In this article, we aimed to 1) determine the frequency of acute symptomatic seizures in infants with supratentorial ICH receiving seizure prophylaxis, 2) identify clinical features predictive of seizure occurrence, and 3) determine the effect of acute seizures on long-term morbidity and mortality. To accomplish this, we analyzed a retrospective cohort of infants with ICH. To our knowledge, this is one of the largest cohorts of infant ICH at a single tertiary care center.
METHODS This is a retrospective cohort study of infant subjects (> 2 wk and < 24 mo), who were diagnosed with supratentorial ICH between June 2008 and February 2013 at a tertiary care center. This study was approved by the Children’s National Medical Center Institutional Review Board (Pro00001916). Neonates with idiopathic ICH were excluded, as were premature infants and patients with a history of seizures. All patients in our sample were started on seizure prophylaxis with antiepileptic drugs (AEDs) at the time of diagnosis of ICH, per institution protocol. AEDs were chosen according to physician discretion. Levetiracetam was administered as a 20 mg/kg bolus followed by a maintenance dose of 20–30 mg/kg/d. (Fos)phenytoin and phenobarbital were administered as a loading dose of 20 mg/kg followed by maintenance of 5 mg/kg/d, with additional boluses and/or adjustment of maintenance based on drug level monitoring. (Fos)phenytoin levels were corrected for hypoalbuminemia. Supratentorial ICH was defined as the presence of any blood products within the skull above the posterior fossa. ICH was diagnosed by either head CT or MRI. The primary outcome studied was acute symptomatic seizure(s), defined as those that occurred during the first 7 days after ICH and after starting seizure prophylaxis. Seizures on presentation, defined as those that occurred as the presenting symptom or in association with the presenting symptoms, were documented as separate events. All clinical data were collected from a neurocritical care database, as well as inpatient and outpatient electronic medical records. All neuroimaging during the first 7 days of admission was reviewed for all patients. Risk factors studied were age, gender, etiology (traumatic or nontraumatic), location of hemorrhage, seizure(s) on presentation, and the presence of parenchymal injury on imaging. Location of ICH was characterized as extra-axial (including intraventricular, subarachnoid, subdural, and epidural hemorrhages), intra-axial (intraparenchymal), or both. Parenchymal injury was defined as any acute pathology involving supratentorial gray or white matter. Examples of hemorrhagic parenchymal injury include contusions and diffuse axonal injury, which were also considered intra-axial ICH. Nonhemorrhagic parenchymal injury included ischemia/infarction and edema. Pediatric Critical Care Medicine
Both clinical and electrographic-only seizures were included in the analysis. As per our institution’s clinical pathway, patients admitted to the PICU with acute encephalopathy are monitored with continuous video electroencephalogram for 48 hours or until mental status improves. Clinical seizures included any stereotyped motor or autonomic manifestations identified as seizure by medical staff and thus warranting treatment with anticonvulsants. For those patients on electroencephalogram monitoring, only clinical events with corresponding ictal change on electroencephalogram were included. Electrographic-only seizures were defined as rhythmic ictal discharges on electroencephalogram lasting at least 10 seconds without clinical correlate seen on the video. Secondary outcomes studied were mortality before discharge and development of late seizures (beginning > 7 d after ICH). Electronic medical records of all outpatient neurology visits from time of discharge to present were reviewed. At each visit, patients were assessed for seizure occurrence. In those with seizures, date of occurrence and the use of AEDs were recorded. In addition, all emergency department visits, hospital admissions, and telephone encounters were reviewed to ensure complete identification of late seizures. Statistical methods included contingency table analysis to evaluate bivariate relationships. Multiple logistic regression analysis was used to evaluate the effects of multiple risk factors on seizure risk. All tests of statistical significance were based on two-tailed p values.
RESULTS Between June 2008 and February 2013, 72 infants were diagnosed with supratentorial ICH. All patients were started on seizure prophylaxis at the time of diagnosis. Prophylaxis was provided with levetiracetam in 41 patients (57%), (fos)phenytoin in 26 patients (36%), and phenobarbital in five patients (7%). Twenty-nine patients (40%) had acute seizures. The prevalence was similar in both traumatic and nontraumatic etiologies (40% and 42%, respectively). Of the patients on levetiracetam, 12 of 41 (29%) had seizures when compared with 2 of 5 patients (40%) on phenobarbital and 15 of 26 patients (58%) on (fos)phenytoin. These differences were not statistically significant (p = 0.07). On average, seizures occurred on hospital day 1.9 (sd, 1.6). Twenty-four patients (83%) required two or more AEDs for seizure control. Eleven patients (38%) required a continuous infusion of benzodiazepines and/or barbiturates for seizure cessation. Clinical and demographic data of patients with and without acute seizures as well as the entire cohort are represented in Table 1. Mean age of all patients was 7 months (sd, 5.7). Patients who experienced acute seizures tended to be younger, with a mean age of 4.5 months (sd, 4.8) when compared with 8.6 months (sd, 5.7) in patients who did not have seizures. In both groups of patients, the most common etiology of hemorrhage was TBI, and the most common location of hemorrhage was extra-axial. Over twice as many patients with acute seizures had seizures on presentation and/or parenchymal injury when compared with patients who did not have acute seizures. www.pccmjournal.org
751
Bansal et al
Table 1. Cohort Demographics and Association of Hemorrhage Characteristics With Acute Symptomatic Seizures Predictor
All Patients (n = 72)
Acute Seizure(s) (n = 29)
No Acute Seizure (n = 43)
p
4.5 ± 4.8
8.6 ± 5.7
0.001a
Age (mo), n ± sd
7 ± 5.7
Gender: female, n (%)
20 (28)
12 (41)
8 (19)
0.03a
Coagulopathyb
4 (5.6)
2 (6.9)
2 (4.7)
0.68
Traumatic brain injury
60 (83)
36 (84)
0.91
Vasculopathy
7 (9.7)
3 (10.3)
4 (9.3)
0.88
Other
1 (1.4)
0 (0)
1 (2.3)
0.41
21 (72)
34 (79)
0.51
5 (6.9)
1 (3.5)
4 (9.3)
0.34
Both
13 (18)
7 (24)
6 (14)
0.24
Seizure on presentation
31 (43)
20 (69)
11 (26)
0.0003a
Parenchymal injury
40 (56)
24 (83)
16 (37)
0.0001a
Etiology, n (%)
c
24 (83)
Location of intracranial hemorrhage, n (%) Extra-axial Intra-axial
52 (72)
Statistically significant result. Etiologies of coagulopathy included vitamin K deficiency and the use of IV heparin during extracorporeal membrane oxygenation. c Etiologies of vasculopathy included arteriorvenous malformations, aneurysms, and cavernomas. a
b
Several demographic and clinical factors were significantly associated with acute symptomatic seizures in bivariate analyses (Table 1). Relative risk was increased significantly in younger children, women, patients with seizures on presentation, and those with parenchymal injury. Etiology and location of ICH were not predictive of acute seizures. After controlling for age differences, seizures on presentation and parenchymal injury remained statistically significant individual predictors of acute symptomatic seizures (p < 0.001 for both). Multivariate analysis of risk factors with p value of less than 0.05 showed that parenchymal injury (OR, 8.1; p = 0.007) and age (OR, 7.2; p = 0.02) remained strong independent risk factors. Patients with seizures on presentation Table 2.
and women continued to trend toward increased risk (OR, 3.3 and 2.7, respectively) although results were not significant. Over 80% of patients had TBI, and of these, 60% were due to presumed nonaccidental trauma (NAT) (Table 2). Furthermore, 37% of patients with NAT had associated diffuse anoxic or ischemic brain injury, whereas no patients with alternate causes of TBI did; this difference was statistically significant (p < 0.001). NAT and associated anoxic or ischemic injury were independently associated with an increased likelihood of acute seizures (p < 0.01). The patients age with NAT did not differ significantly from those with other causes of TBI. Twenty-five percent of patients with TBI had elevated
Cohort Demographics of Patients With Traumatic Brain Injury All Patients (n = 60)
Predictor
Acute Seizure(s) (n = 24)
No Acute Seizure (n = 36)
p
0.003a
Etiology, n (%) Nonaccidental trauma,
36 (60)
20 (83)
16 (44)
Other traumatic brain injury
24 (40)
4 (17)
20 (56)
Present
13(22)
10 (42)
3 (8.3)
Absent
47 (78)
14 (48)
33 (92)
Elevated
13 (22)
6 (25)
7 (19)
Normal
47 (78)
18 (75)
29 (81)
Diffuse ischemia/anoxic injury, n (%) 0.002a
Intracranial pressure, n (%)
a
0.610
Statistically significant result.
752
www.pccmjournal.org
October 2014 • Volume 15 • Number 8
Neurocritical Care
Table 3.
Bivariate Analysis of Risk Factors for Electrographic-Only Seizures
Predictor
Relative Risk
95% CI
p
Gender: female
1.26
0.69–2.27
0.46
Traumatic brain injury
0.99
0.71–1.4
0.95
Location of intracranial hemorrhage
1.23
0.57–2.67
0.56
Seizure on presentation
1.08
0.55–2.14
0.82
Parenchymal injury
3.33
0.56–19.7
0.054
intracranial pressure requiring intervention; however, this was not associated with acute seizures. Forty-five patients (63%) were encephalopathic at admission and therefore monitored with electroencephalogram; duration of monitoring varied based on clinical indication. This included 28 of the 29 patients with acute seizures. One male patient with a normal neurological examination had an acute clinical seizure; his AED dose was increased and had no further events; therefore, electroencephalogram was not done. Of all patients with acute seizures, 59% had electrographic-only seizures during hospitalization; 21% had electrographic-only seizures exclusively. Thirteen patients (77%) with electrographic-only seizures had seizures as a presenting symptom, and 11 (65%) experienced seizures within the first 24 hours of hospitalization. Bivariate analysis suggested that patients with parenchymal injury remained at increased risk of electrographic-only seizures, with results trending toward significance (Table 3). Seven patients (9.7%) died during hospitalization. Bivariate analyses of potential risk factors for mortality before hospital discharge are shown in Table 4. TBI was associated with increased risk of mortality (p < 0.01). Neither acute seizures nor electrographic-only seizures were associated with increased mortality. Follow-up data were available on 63 of 65 surviving patients (97%). Mean duration of follow-up after ICH was 12.4 months (1 mo to 4.6 yr). Eleven patients (15%) experienced late seizures and thus required ongoing treatment with AEDs at their latest follow-up visit. This included 6 of 29 patients (21%) with acute seizures and 5 of 43 patients (12%) without acute seizures. Bivariate analyses (Table 5) suggested an increased risk of developing late seizures in women and patients with seizures on presentation, although these results did not reach statistical significance, the trend is similar to that seen for acute seizures. Similarly, patients with acute seizures had an increased risk of late seizures. Parenchymal injury and electrographic-only seizures were significantly associated with late seizures. Eleven of 33 patients with parenchymal injury developed late seizures when compared with 0 of 32 patients without parenchymal injury (p < 0.001). Results were similar for electrographic-only seizures; all patients with late seizures had experienced electrographic-only seizures during hospitalization (p = 0.013).
DISCUSSION Acute symptomatic seizures are a serious complication of intracranial hemorrhage and are most often studied in TBI. In Pediatric Critical Care Medicine
pediatric patients with TBI, the frequency is estimated between 8.4% and 35%, with younger children at greatest risk (5, 6, 13). In this article, we specifically studied this high-risk population and sought to identify risk factors associated with acute symptomatic seizures after supratentorial ICH of any etiology. To establish the prevalence of acute symptomatic seizures truly, we would need to compare children who received prophylaxis to those who did not; however, seizure prophylaxis for ICH has been an institutional protocol since 2007. In infants receiving prophylaxis, we found an overall prevalence of 40%. This is regardless of whether hemorrhage is traumatic or nontraumatic. Furthermore, in bivariate analysis, TBI was not an independent risk factor for acute seizures, thus hemorrhage of various etiologies may be associated with acute seizures. Women had a trend toward increased risk for acute symptomatic seizures; gender appeared to be an important risk factor in bivariate analysis, but significance was lost in multivariate analysis. The effect of gender on acute post-traumatic seizures and long-term outcomes has been extensively studied (5, 14–17), although results are conflicting, sex hormones undoubtedly influence the brain. Increased levels of peripheral sex steroids following injury are associated with increased mortality and worse long-term outcomes (18), whereas progesterone has been found to be neuroprotective after a variety of CNS insults (19) and may be a potential therapeutic agent after TBI (20). Thus, the influence of gender on brain injury remains unclear but may affect various aspects of clinical course. Another important risk factor is the presence of parenchymal injury, not only for acute seizures but also for late seizures and possibly electrographic seizures as well. Interestingly, however, the location of hemorrhage is not a risk factor. In fact, most patients with acute seizures had extra-axial hemorrhages, specifically subdural hematomas. This suggests that it is not the hemorrhage itself that is epileptogenic, but rather the associated edema and/or ischemia of adjacent tissue. Previous studies have shown that cerebral edema is a risk factor for acute post-traumatic seizures (5, 13), as well as poor outcome (21). We now find that approximately half of all patients with acute seizures have evidence of hypoxia and/or ischemia on neuroimaging. Cerebral ischemia is known to alter gamma-aminobutyric acid (GABA)ergic and glutamatergic neurotransmission leading to neuronal hyperexcitability (22); thus, the increased epileptogenicity in these patients is not surprising. We found a high prevalence of electrographic-only seizures, which is consistent with previous studies. Hasbani et al www.pccmjournal.org
753
Bansal et al
Table 4.
Bivariate Analysis of Risk Factors for Mortality
Predictor
a
p
Relative Risk
95% CI
Gender: female
0.93
0.76–1.12
0.35
Traumatic brain injury
1.43
0.95–2.13
0.0025a
Seizure on presentation
1.00
0.86–1.16
0.99
Acute symptomatic seizure
0.93
0.78–1.10
0.34
Electrographic seizures
0.90
0.68–1.18
0.47
Statistically significant result.
(23) found that 57% of infants, median age of 4 months, with abusive head trauma had electrographic seizures. A study of children with TBI found that only 16% had electrographic seizures; however, this study included all children up to 18 years (24). Additional studies of continuous electroencephalogram in pediatric patients with varying pathologies/etiologies have found similarly high rates of electrographic seizures (25, 26). Electrographic seizures are significant in that they may herald impending decompensation; in adults, electrographic seizures in the setting of ICH are associated with expansion of hemorrhage (7). Furthermore, although we found that the majority of electrographic seizures occurred in the first day of monitoring, time to seizure onset can vary based on etiology of hemorrhage and evolution of pathology. Thus, continuous electroencephalogram monitoring remains a necessary component of initial ICU care following ICH. Its use, however, goes far beyond seizure detection. It affects various aspects of medical management and can serve as a tool for prognostication as well (27, 28). The choice of prophylactic agent may affect the occurrence of acute seizures. Fewer patients treated with levetiracetam had acute seizures when compared with those treated with (fos) phenytoin or phenobarbital. One possible explanation for this is failure to maintain therapeutic levels with the latter anticonvulsants. Although attempts were made during hospitalization to maintain therapeutic levels based on drug level monitoring, we do not have data to confirm that patients were at therapeutic levels at the exact time when seizures occurred in this study. A second potential cause of the discrepant results may be accounted for by patient age. There was a tendency for younger children to be given (fos)phenytoin as their initial prophylactic agent; therefore, the apparent advantage of levetiracetam may in fact be because of the older patient population on this drug. Studies in adult populations comparing levetiracetam with phenytoin have demonstrated similar efficacies with Table 5.
better long-term outcomes and fewer adverse events in patients treated with levetiracetam (11, 12). In children, levetiracetam was found to be superior to (fos)phenytoin in preventing acute symptomatic seizures following hematopoietic stem cell transplantation (29) and comparable with (fos)phenytoin in children with intracranial hemorrhage (30). Although comparison of AEDs was not the primary outcome of this study, our findings suggest that there may be a difference. additional studies are needed to evaluate optimal prophylactic agents in infants and children, including defining optimal dosing regimens for levetiracetam as well as standardized methods of maintaining therapeutic drug levels. The effect of both seizure prophylaxis and acute seizures on long-term outcomes is also well established. Acute seizures lead to worse outcomes based on Glasgow Outcome Score (5, 6) and predict the development of post-traumatic epilepsy (9, 31–33). In turn, late seizures are associated with worse neurological and developmental outcomes (32). Seizure prophylaxis significantly reduces mortality in children with head trauma (34). We found that acute seizures do not affect mortality but may lead to late seizures. The risk of late seizures in children with acute seizures was nearly twice those without acute seizures, and these results were trending toward significance. Our estimated prevalence of late seizures is 15%, which is similar to overall estimates of children with TBI but lower than previous estimates in infants (9, 32). In summary, acute symptomatic seizures are a common complication of ICH and have both short- and long-term implications. Despite studies showing that younger children are at particularly high risk for acute seizures, data regarding seizure prophylaxis in infants is lacking. AEDs are known to be protective against acute post-traumatic seizures in children (32) and may also have a role in neuroprotection (36). Thus, seizure prophylaxis is becoming an increasingly used component of ICU care following head trauma. We found that even
Bivariate Analysis of Risk Factors for Late Seizures Relative Risk
95% CI
p
Gender: female
1.61
0.54–4.85
0.40
Traumatic brain injury
0.63
0.17–2.42
0.52
Seizure on presentation
2.31
0.75–7.13
0.13
Acute symptomatic seizure
1.92
0.65–5.63
0.23
Predictor
754
www.pccmjournal.org
October 2014 • Volume 15 • Number 8
Neurocritical Care
with prophylaxis, nearly half of all infants develop acute seizures. Given our findings that non-traumatic etiologies of ICH have a similar risk of acute symptomatic seizures as traumatic etiologies, we recommend that seizure prophylaxis be considered after all causes of acute ICH in infants, particularly those in whom seizure is a presenting sign of hemorrhage and where there is evidence of parenchymal injury on imaging.
CONCLUSIONS Infants less than 24 months have a high risk of acute symptomatic seizures, despite seizure prophylaxis and regardless of the etiology of ICH. Younger age, those with seizures as a presenting symptom, and those with parenchymal injury are at highest risk. This may affect long-term outcomes. The optimal agent for seizure prophylaxis in the acute setting has yet to be determined.
ACKNOWLEDGMENT We thank Robert McCarter, ScD, for his assistance with statistical analysis.
REFERENCES
1. Beslow LA, Abend NS, Gindville MC, et al: Pediatric intracerebral hemorrhage: Acute symptomatic seizures and epilepsy. JAMA Neurol 2013; 70:448–454 2. Beghi E, Carpio A, Forsgren L, et al: Recommendation for a definition of acute symptomatic seizure. Epilepsia 2010; 51:671–675 3. Kochanek P, Carney N, Adelson D, et al: Guidelines for the early medical management of severe traumatic brain injury in infants, children, and adolescents—second edition. Pediatr Crit Care Med 2012; 13:S72–S82 4. Da Silva AM, Willmore LJ: Posttraumatic epilepsy. Handb Clin Neurol 2012; 108:585–599 5. Chiaretti A, De Benedictis R, Polidori G, et al: Early post-traumatic seizures in children with head injury. Childs Nerv Syst 2000; 16:862–866 6. Ateş O, Ondül S, Onal C, et al: Post-traumatic early epilepsy in pediatric age group with emphasis on influential factors. Childs Nerv Syst 2006; 22:279–284 7. Claassen J, Jetté N, Chum F, et al: Electrographic seizures and periodic discharges after intracerebral hemorrhage. Neurology 2007; 69:1356–1365 8. Agrawal A, Timothy J, Pandit L, et al: Post-traumatic epilepsy: An overview. Clin Neurol Neurosurg 2006; 108:433–439 9. Appleton RE, Demellweek C: Post-traumatic epilepsy in children requiring inpatient rehabilitation following head injury. J Neurol Neurosurg Psychiatry 2002; 72:669–672 10. Bratton S, Chestnut R, Ghajar J, et al; Brain Trauma Foundation; American Association of Neurological Surgeons; Congress of Neurological Surgeons; Joint Section on Neurotrauma and Critical Care, AANS/CNS. Guidelines for the management of severe traumatic brain injury. XIII. Antiseizure prophylaxis. J Neurotrauma 2007; 24:S83–S86 11. Milligan TA, Hurwitz S, Bromfield EB: Efficacy and tolerability of levetiracetam versus phenytoin after supratentorial neurosurgery. Neurology 2008; 71:665–669 12. Szaflarski JP, Sangha KS, Lindsell CJ, et al: Prospective, randomized, single-blinded comparative trial of intravenous levetiracetam versus phenytoin for seizure prophylaxis. Neurocrit Care 2010; 12:165–172
Pediatric Critical Care Medicine
13. Hahn YS, Fuchs S, Flannery AM, et al: Factors influencing posttraumatic seizures in children. Neurosurgery 1988; 22:864–867 14. Ong LC, Dhillon MK, Selladurai BM, et al: Early post-traumatic seizures in children: Clinical and radiological aspects of injury. J Paediatr Child Health 1996; 32:173–176 15. Schmidt AT, Hanten GR, Li X, et al: Decision making after pediatric traumatic brain injury: Trajectory of recovery and relationship to age and gender. Int J Dev Neurosci 2012; 30:225–230 16. Gao GY, Jiang JY. Chinese Head Trauma Data Bank: Effect of gender on the outcome of patients with severe traumatic brain injury. J Neurotrauma 2012; 29:96–100 17. Renner C, Hummelsheim H, Kopczak A, et al: The influence of gender on the injury severity, course and outcome of traumatic brain injury. Brain Inj 2012; 26:1360–1371 18. Wagner AK, McCullough EH, Niyonkuru C, et al: Acute serum hormone levels: Characterization and prognosis after severe traumatic brain injury. J Neurotrauma 2011; 28:871–888 19. Stein DG, Wright DW, Kellermann AL: Does progesterone have neuroprotective properties? Ann Emerg Med 2008; 51:164–172 20. Stein DG: Is progesterone a worthy candidate as a novel therapy for traumatic brain injury? Dialogues Clin Neurosci 2011; 13:352–359 21. Bahloul M, Chaari A, Chabchoub I, et al: Outcome analysis and outcome predictors of traumatic head injury in childhood: Analysis of 454 observations. J Emerg Trauma Shock 2011; 4:198–206 22. Schwartz-Bloom RD, Sah R: Gamma-Aminobutyric acid(A) neurotransmission and cerebral ischemia. J Neurochem 2001; 77:353–371 23. Hasbani DM, Topjian AA, Friess SH, et al: Nonconvulsive electrographic seizures are common in children with abusive head trauma. Pediatr Crit Care Med 2013; 14:709–715 24. Arndt DH, Lerner JT, Matsumoto JH, et al: Subclinical early posttraumatic seizures detected by continuous EEG monitoring in a consecutive pediatric cohort. Epilepsia 2013; 54:1780–1788 25. Williams K, Jarrar R, Buchhalter J: Continuous video-EEG monitoring in pediatric intensive care units. Epilepsia 2011; 52:1130–1136 26. Schreiber JM, Zelleke T, Gaillard WD, et al: Continuous video EEG for patients with acute encephalopathy in a pediatric intensive care unit. Neurocrit Care 2012; 17:31–38 27. Gallentine WB: Utility of continuous EEG in children with acute traumatic brain injury. J Clin Neurophysiol 2013; 30:126–133 28. Abend NS, Topjian AA, Gutierrez-Colina AM, et al: Urgent continuous EEG (cEEG) monitoring leads to changes in treatment in half of cases impact of continuous EEG monitoring on clinical management in critically ill children. Neurocrit Care 2011; 15(1):70–75 29. Soni S, Skeens M, Termuhlen AM, et al: Levetiracetam for busulfaninduced seizure prophylaxis in children undergoing hematopoietic stem cell transplantation. Pediatr Blood Cancer 2012; 59:762–764 30. Bansal S, Blalock D, Kebede T, et al: Levetiracetam versus (fos)phenytoin for seizure prophylaxis in pediatric patients with intracranial hemorrhage. J Neurosurg Pediatr 2014; 13:209–215 31. Kieslich M, Jacobi G: Incidence and risk factors of post-traumatic epilepsy in childhood. Lancet 1995; 345:187 32. Kieslich M, Marquardt G, Galow G, et al: Neurological and mental outcome after severe head injury in childhood: A long-term follow-up of 318 children. Disabil Rehabil 2001; 23:665–669 33. Englander J, Bushnik T, Duong TT, et al: Analyzing risk factors for late posttraumatic seizures: A prospective, multicenter investigation. Arch Phys Med Rehabil 2003; 84:365–373 34. Tilford JM, Simpson PM, Yeh TS, et al: Variation in therapy and outcome for pediatric head trauma patients. Crit Care Med 2001; 29:1056–1061 35. Liesemer K, Bratton SL, Zebrack CM, et al: Early post-traumatic seizures in moderate to severe pediatric traumatic brain injury: Rates, risk factors, and clinical features. J Neurotrauma 2011; 28:755–762 36. Belcastro V, Pierguidi L, Tambasco N: Levetiracetam in brain ischemia: Clinical implications in neuroprotection and prevention of poststroke epilepsy. Brain Dev 2011; 33:289–293
www.pccmjournal.org
755
