10. Post-Traumatic Seizure Disorder

Transcription

1 10. Post-Traumatic Seizure Disorder Robert Teasell MD FRCPC, Jo-Anne Aubut BA, Corbin Lippert MN RN, Shawn Marshall MSc MD FRCPC, Nora Cullen MSc MD FRCPC ERABI Parkwood Hospital 801 Commissioners Rd E, London ON x Module 10-Post Traumatic Seizure Disorder-V9 2013

2 Table of Contents 10.1 Incidence of Post-Traumatic Seizures Classification of Post-Traumatic Seizures and Epilepsy Risk Factors for Post-Traumatic Seizures Identification of the High-Risk Patient Onset Recurrence Clinical Picture of Post-Traumatic Seizures Complications of Post-Traumatic Seizures Cognitive and Behavioral Function Influence on Neurologic Recovery Functional Status Status Epilepticus Mortality Treatment of Post-Traumatic Seizures Seizure Prevention or Prophylaxis Post-Traumatic Seizure Prophylaxis in Children Surgical Treatment of Post-Traumatic Seizures Surgical Excision of the Post-Traumatic Seizure Focus Summary Reference List Module 10-Post Traumatic Seizure Disorder-V9 2013

3 Table Directory Table 10.1 Table 10.2 Table 10.3 Table 10.4 Table 10.5 Table 10.6 Table 10.7 Table 10.8 Table 10.9 Definitions of Post-Traumatic Seizures Studies of Risk Factors for Late Post-traumatic Seizures Studies of Post-Traumatic Seizure Recurrence Summary of Onset and Recurrence of PTS Seizure Prevention or Prophylaxis Summary of RCT Studies of Seizure Prevention or Prophylaxis Summary of non-rct Studies of Seizure Prevention or Prophylaxis Pharmacological Seizure Prevention or Prophylaxis in Children following ABI Surgical Treatment of Post-Traumatic Seizures 3 Module 10-Post Traumatic Seizure Disorder-V9 2013

4 Key Points Increasing age, alcohol abuse, family history, increasing injury severity, and the occurrence of early seizures immediately after the injury all increase the risk of developing late post-traumatic seizure disorder. Later onset seizures have a higher chance of recurrence. Levetiracetam is as effective as phenytoin in treating and preventing seizures in individuals in the intensive care unit post ABI. Anticonvulsants provided immediately post-abi reduce the occurrence of seizures only within the first week. Anticonvulsants provided shortly post-abi do not reduce long term mortality, morbidity or late seizures. Anticonvulsants have negative consequences on motor tasks. Early glucocorticoid exposure may increase seizures. Methylphenidate may not increase the risk of seizures. Intramuscular midazolam may be effective for acute seizure cessation. Phenytoin does not reduce early or late seizures in children post-abi. Surgical excision can reduce seizures if the focus of the seizures can be localized. 4 Module 10-Post Traumatic Seizure Disorder-V9 2013

5 10. Post-Traumatic Seizure Disorder 10.1 Incidence of Post-Traumatic Seizures The Brain Injury Special Interest Group of the American Academy of Physical Medicine and Rehabilitation (1998) has noted that there are approximately 422,000 cases of inpatient traumatic brain injuries (TBI) in the United States every year. Post traumatic seizure disorder, although identified as a serious consequence of TBI, remains an understudied problem (Ferguson et al., 2010). It is believed TBIs result in 20% of symptomatic epilepsy in the general population (Bushnik et al., 2012). Of all TBI patients who are hospitalized 5% to 7% will experience post-traumatic seizures (PTS). However, the incidence of PTS is much higher on rehabilitation units (as high as 17%) which is felt to reflect the increased severity of the injury and a higher number of risk factors in this population (Bontke et al., 1993; Armstrong et al., 1990; Cohen & Groswasser, 1991; Kalisky et al., 1985; Ng et al., 2004; Sazbon & Groswasser, 1990). The incidence of late post traumatic seizures (LPTS) ranges fro 5 to 19% for the general population (Bushnik et al., 2012). Risk factors associated with LPTS included: age, penetrating injury, multiple neurosurgical procedures, severity of injury, location of lesions, depressed skull fractures, intracranial hemorrhage and whether or not the individual experienced PTS (Bushnik et al. 2012; pg 36). For those who sustain a severe non penetrating TBI approximately 11% will experience seizures post injury and for those whose TBI is the result of a penetrating injury the numbers increase to 35 to 50% (Yablon 1993). It has been noted that in young adults TBI is the leading cause of epilepsy (Annegers 1996). Recently in a cohort study conducted by Ferguson and colleagues (2010) they found the incidence of PTE was higher in individuals between 30 and 54 years. They also found those with severe TBIs, or those who had other concomitant injuries, or more comorbid conditions, previous head injuries, stroke or depression were more likely to develop LPTS. These findings mirrored the findings of previous studies (Weiss et al., 1983; Annegers et al., 1998; Andelic et al., 2009). Seizures following a TBI or ABI have been linked to accidental injuries, psychological effects (depression), a loss of independence (driving privileges) and a reduction in employability (Andelic et al., 2009; Brain Injury SIG,1998). It is also believed that by preventing recurrent early seizures, one can prevent the development of chronic epilepsy (Yablon, 1993). This belief has resulted in the use of prophylactic anticonvulsants in those regarded as high risk for developing PTS despite knowing the side effects of these treatments may have a negative impact on cognitive recovery (Yablon, 1993). When looking at children who sustain TBIs study results indicate children are less likely to suffer from LPTS when the injury sustained is a non penetrating injury (Annegers et al., 1980; Hahn et al., 1988; Kollevold 1979); however, as with the adult population, those who experience penetrating TBIs the incidence rates of PTS ranges from 35-50% (Ascroft 1941; Caveness et al., 1979; Caveness and LISS 1961). 5 Module 10-Post Traumatic Seizure Disorder-V9 2013

6 Classification of Post-Traumatic Seizures and Epilepsy Post-traumatic seizure disorders have been defined in the Practice Parameter on the Antiepileptic Drug Treatment of Post-traumatic Seizures by the Brain Injury Special Interest Group of the American Academy of Physical Medicine and Rehabilitation (see Table 10.1). Table 10.1 Definitions of Post-Traumatic Seizures (Asikainen et al., 1999) Seizure Discrete clinical events that reflect a temporary physiologic dysfunction of the brain characterized by excessive and hypersynchronous discharge of cortical neurons. Post-Traumatic Seizure An initial or recurrent seizure episode not attributable to another obvious cause after penetrating or nonpenetrating TBI. The term post-traumatic seizure is preferred over post-traumatic epilepsy because the former encompasses both single and recurrent events. Immediate Post-Traumatic Seizure A seizure due to TBI occurring within the first 24 hours of injury. Early Post-Traumatic Seizure A seizure due to TBI occurring within the first week of injury. Late Post-Traumatic Seizure A seizure due to TBI occurring after the first week of injury. Post-Traumatic Epilepsy A disorder characterized by recurrent late seizure episodes not attributable to another obvious cause in patients following TBI. Although the term post-traumatic epilepsy commonly has designated single or multiple seizures including early seizures, the term should be reserved for recurrent, late PTS. Nonepileptic Seizures Episodic behavioral events that superficially resemble epileptic attacks but are not associated with paroxysmal activity within the brain. Antiepileptic Drug Prophylaxis In the context of PTS, AED treatment administered to prevent seizures in patients who have not manifested seizures. Epilepsy Epilepsy is a condition characterized by recurrent unprovoked seizures. Practice Parameters Results, in the form of one or more specific recommendations, from a scientifically based analysis of a specific clinical problem. 6 Module 10-Post Traumatic Seizure Disorder-V9 2013

7 10.2 Risk Factors for Post-Traumatic Seizures The Brain Injury Special Interest Group of the American Academy of Physical Medicine and Rehabilitation (1998) has noted that the risk of epilepsy is highest within the first two years following brain trauma (Yablon, 1993; Dikmen et al.,1991). There are several patient and injury characteristics that increase the likelihood of developing PTS. These include a Glasgow Coma Scale score of <10, cortical contusions, depressed skull fractures, epidural hematomas, intracerebral hematomas, wounds with dural penetration, a seizure within the first week of injury, prolonged length of coma, and prolonged length of posttraumatic amnesia (Yablon1993; Dikmen et al., 199; Brain Injury SIG of AAPM&R 1998) It has been noted that 20% of adults have spontaneous seizures within two years of TBI with penetrating injury of the brain, intracranial hematoma, cortical contusion, depressed skull fracture or a seizure immediately following the onset of TBI (Temkin 2001). To put this is perspective, (Temkin, 2001) notes that this is over 200 times the rate seen in someone who has not suffered a brain injury. Although the risk of developing post-traumatic seizures is highest within months after the injury (Temkin 2001), the risk remains high for individuals with moderate-severe injuries for as long as 5 years while military personnel suffering severe penetrating missile brain injuries show elevated risks for more than 15 years after the injury (Weiss et al., 1983; Feeney and Walker 1979;Annegers et al.,1998; Salazar et al., 1985; Caveness 1963; Caveness et al.,1979). Table 10.2 Studies of Risk Factors for Late Post-traumatic Seizures (Yablon and Dostrow 2001) Risk Factor Reference Patient Characteristics Age (Annegers et al., 1980; Asikainen et al., 1999; Hahn et al., 1988; Kollevold, 1979) Alcohol use (Evans, 1962; Heikkinen et al., 1990; Kollevold, 1978) Family history (Caveness, 1963; Evans, 1962; Heikkinen et al., 1990; Hendrick & Harris, 1968) Injury Characteristics Bone/metal fragments (Ascroft, 1941; Salazar et al., 1985; Walker & Yablon, 1959) Depressed skull fracture (Jennett, 1975b; Hahn et al., 1988; Phillips, 1954; Wiederholt et al., 1989) Focal contusions/injury (da Silva et al., 1992; De Santis et al., 1992; Eide & Tysnes, 1992; Glotzner et al., 1983; Heikkinen et al., 1990) Focal neurologic deficits (da Silva et al., 1992; Jennett, 1975b; Salazar et al., 1985) Lesion location (da Silva et al., 1992; Evans, 1962; Grafman et al., 1992) Dural penetration (Caveness & LISS, 1961; Evans, 1962; Salazar et al., 1985) Intracranial hemorrhage (Hahn et al., 1988; Glotzner et al., 1983) Injury severity (Evans, 1962; Jennett, 1975b; Salazar et al., 1985; Walker & Yablon, 1961) Other 7 Module 10-Post Traumatic Seizure Disorder-V9 2013

8 Early post-traumatic seizures (Jennett 1975; Jennett and Van De Sande 1975; Salazar et al., 1985; Heikkinen et al. 1990) Conclusions There are several patient and injury characteristics that increase the likelihood for the development of late PTS. Some important patient characteristics include: increasing age, premorbid alcohol abuse, and family history. In terms of injury characteristics, markers of increasing injury severity such as penetrating injuries and depressed skull fracture increase the risks of late PTS. Seizures occurring within the first week post ABI (early seizures) increase the risk of late PTS. As the severity of the brain injury increases, the period of time for which a survivor is at risk of developing PTS also increases. Increasing age, alcohol abuse, family history, increasing injury severity and the occurrence of seizures immediately after the injury all increase the risk of developing late posttraumatic seizure disorder Identification of the High-Risk Patient It is important to identify patients who are at high-risk of developing post-traumatic seizures (PTS) since these patients would benefit greatly from its prevention or from the impediment of its possible reoccurrence. According to Yablon and Dostrow (2001) the clinical characteristics of the patient, the injury, and information obtained from neuroimaging and electrophysiologic assessment techniques can be used to identify those high risk for developing seizure disorders post injury. In addition to this, there are several key risk factors which can aid in the identification of these patients (Yablon and Dostrow 2001) Onset Yablon and Dostrow (2001) have noted that one-half to two-thirds of individuals who suffer PTS will experience seizure onset within the first 12 months, and 75-80% will have seizures by the end of the second year following a TBI (Caveness et al., 1979; da Silva et al., 1992; da Silva et al., 1990; Pohlmann-Eden and Bruckmeir 1997; Walker and Yablon 1959; Walker and Yablon 1961). After 5 years, adults with mild TBI no longer have a significantly increased risk relative to the general population (Annegers et al., 1998), whereas those with moderate or severe TBI or penetrating TBI remain at increased risk for more than 5 years post-injury (Annegers et al., 1998; da Silva et al.,1992; Pagni 1990; Salazar et al., 1985). 8 Module 10-Post Traumatic Seizure Disorder-V9 2013

9 Recurrence Seizure recurrence is an important factor in the determination of disability, employment likelihood, quality of life, and increased health care costs (Baker et al., 1997; van Hout et al., 1997; Yablon and Dostrow 2001). Some studies have reported that following early seizures post-tbi, only one-half of the study participants experienced a recurrence (De Santis et al., 1979; Kollevold1979) while another quarter experienced a total of only 2-3 seizures. Table 10.3 Studies of Post-Traumatic Seizure Recurrence Authors Recurrence Incidence Haltiner et al., (1997) Pohlmann-Eden and Bruckmeir (1997) 63 adults with moderate or severe TBI and later PTS cumulative incidence of recurrent late seizures 86% over 2 yrs; 52% > 4 late seizures, 37% > 9 late seizures. 57 post-traumatic epilepsy patients compared to 50 age and sex-matched severe TBI controls Of PTE patients 35% became seizure-free over 3 years; 21% had > 1 seizure per week; risk factors poor seizure frequency, missile injuries, combined seizure patterns, med noncompliance and alcohol abuse. The following table provides a summary of the natural history for the onset and recurrence of post-traumatic seizures (Yablon and Dostrow (2001) Table 10.4 Summary of Onset and Recurrence of PTS Feature % of cases of patients with early PTS experience a late seizure. Seizure onset after the first week shows a much higher likelihood of seizure recurrence. Seizure frequency within the first year postinjury may predict future recurrence. Persistent PTS may be more common in partial seizures and less common in generalized seizures. A seizure occurring within a few minutes of a head injury has not been found to increase the risk of recurrence in individuals who sustain a mild TBI. A small number of patients experience frequent seizure recurrences, apparently refractory to conventional anti-seizure therapy. Some patients may benefit from surgical intervention. Reference (Yablon & Dostrow, 2001) (Walker & Yablon, 1961; Haltiner et al., 1997) (Salazar et al., 1985) (Salazar et al., 1985) (Jennett and Van De Sande1975; McCrory et al. 1997) (Haltiner et al., 1997; Ng et al., 2004; Pohlmann- Eden & Bruckmeir, 1997) (Marks et al., 1995; Diaz-Arrastia et al., 2000) Conclusions 9 Module 10-Post Traumatic Seizure Disorder-V9 2013

10 Individuals who develop seizures after the first week of the injury have an increased chance of seizure recurrence. Later onset seizures have a higher chance of recurrence Clinical Picture of Post-Traumatic Seizures Wiedemayer et al. (2002) retrospectively analyzed a consecutive series of 1868 adult patients with head injury. Brain injury severity was mild in 568 patients (30.4%), moderate in 681 patients (36.5%), and severe in 619 patients (33.1%). In 109 patients (5.8%), an epileptic seizure occurred in the early post-traumatic period; 39 of the 109 patients who experienced an early post-traumatic seizure, suffered from severe head injury, 58 from a moderate head injury, and 12 from a mild head injury. Sixty-six early seizures (60.5%) occurred within the first day following trauma, with a steep decline over the ensuing days. The first epileptic seizure was generalized in 69 patients (63.3%) and focal in 40 patients (36.7%). Fifty-eight patients (53.2%) experienced a second early seizure during the follow-up period Complications of Post-Traumatic Seizures Seizures following TBI may themselves be a source of significant complications and morbidity and it has been noted that the recurrence of seizures is an important cause of non-elective hospitalization in patients with severe TBI (Cifu et al., 1999). Potential complications include deterioration in cognitive and behavioral functioning and overall functional status, influence on neurological recovery, status epilepticus and death Cognitive and Behavioral Function Post-traumatic seizure disorders may lead to cognitive and behavioural disorders (Yablon & Dostrow, 2001). Cognitive problems may arise during the interictal state in the absence of active seizures (Aarts et al., 1984; Aldenkamp 1997; Binnie and Marston 1992).Patients with PTS experience persistent behavioral abnormalities and a higher incidence of psychiatric-related hospitalizations even compared with patients with penetrating TBI who do not experience PTS (Swanson et al., 1995) Influence on Neurologic Recovery Post-traumatic seizures can influence neurological recovery (Hernandez & Naritoku, 1997; Yablon & Dostrow, 2001). Yablon and Dostrow (2001) have noted that in rodent models, brief and infrequent PTS occurring early after brain damage do not appear to impact functional recovery. However, more severe and widespread seizures occurring within the first 6 days post brain injury result in permanent impairments of functional recovery; while the same seizures occurring after the 6 day mark result in no change in somatosensory recovery (Hernandez and Naritoku 1997). 10 Module 10-Post Traumatic Seizure Disorder-V9 2013

11 Functional Status Recurrent PTS may exert a negative impact on functional status following TBI, an adverse effect independent of the severity of the injury (Barlow et al., 2000; Schwab et al., 1993). In the case of penetrating traumatic brain injuries, post-traumatic seizures have been reported to be an important and independent factor which affects both employment status and cognitive performance (Schwab et al., 1993) However, in the case of nonpenetrating TBI, the impact of PTS on functional prognosis and cognition is less clear (Armstrong et al., 1990; Asikainen et al., 1999). Haltiner et al. (1997) found no significant differences at one year as a consequence of later PTS in terms of neuropsychological performance and psychosocial functioning when adjusting for injury severity. Asikainen et al. (1999) found that patients with PTS did have poorer outcomes on the Glasgow Outcome Scale, although there were no significant differences in employment outcome associated with the presence of PTS Status Epilepticus Status epilepticus can be defined as either more than 30 minutes of continuous seizure activity (Yablon & Dostrow, 2001) or two or more sequential seizures without full recovery of consciousness between seizures (Hernandez and Naritoku 1997; Yablon1993). Status epilepticus is regarded as the most serious of the complications of PTS and may actually lead to additional neurological damage. Fortunately, clinically apparent status epilepticus is an infrequent complication of PTS (Kollevold, 1979) Mortality In earlier studies mortality among those who sustain a TBI and develop PTS mortality was high (Corkin et al., 1984; Walker and Erculei 1970; Walker and Blumer 1989). Recently Englander and colleagues (2009) found mortality rates for those who had sustained a TBI and were diagnosed with late post traumatic seizures (LPTS) were higher than those who has sustained a TBI but had no recorded history of LPTS (non-lpts group). Those in the LTPS group who died were younger than individuals in the non- LPTS group. Earlier studies found those with penetrating TBIs had a higher risk of dying, however, this may have been due to the initial trauma and not a result of PTS (Rish et al., 1983; Rish and Caveness 1973). Yablon and Dostrow (2001) have noted that the complications associated with single later post-traumatic seizures are no different than those found in any seizure and are generally associated with minimal risk. However, the risks of increased mortality and increased morbidity in the form of worsened cognitive and functional prognosis are associated with increasing frequency and severity of the seizure disorder Treatment of Post-Traumatic Seizures Schierhout and Roberts (2001) have reported that a seizure which occurs soon after a head injury may cause secondary brain damage, due to the increased metabolic 11 Module 10-Post Traumatic Seizure Disorder-V9 2013

12 demands of the brain soon after the injury, increased intracranial pressure, and excessive amounts of neurotransmitter release which may lead to additional brain injury. For this reason, the primary therapeutic objective in the use of anticonvulsant drug has been the prevention of early seizures in an attempt to minimize the extent of brain damage following an injury. Some anticonvulsant drugs have been shown to have neuroprotective properties in animal studies. For example, following hypoxia, phenytoin has been linked with reduced neuronal damage in neonatal rats (Vartanian et al., 1996) and in rat hippocampal cell cultures (Tasker et al., 1992). Experimental evidence suggests that the neuroprotective effects of phenytoin are related to a blockage of voltage dependent sodium channels during hypoxia (Tasker et al., 1992; Vartanian et al., 1996) which would be expected to decrease the spread of calcium induced neurotoxicity following hypoxic brain injury. As noted by Schierhout and Roberts (2001), this suggests that anti-epileptics may have beneficial properties which may be independent of their proposed anti-seizure activity. Conversely, anti-epileptic drugs have shown toxic effects even in stable patients, with impaired mental and motor function being the most common adverse effects, although serious adverse effects including deaths as a result of hematological reactions have been also reported (Reynolds et al., 1998). Schierhout and Roberts (2001) have suggested that the injured brain s response to anticonvulsants may be such that toxic effects could be more pronounced and neurological recovery may be delayed Seizure Prevention or Prophylaxis Initially, retrospective and nonrandomized clinical trials in humans showed favorable results for the efficacy of anti-epileptic drug prophylaxis; however, prospective investigations of chronic prophylaxis for late PTS have shown less impressive results. Below is a discussion of the treatments tried and the results of those efforts. Individual Studies Table 10.5 Seizure Prevention or Prophylaxis Author/Year/ Country/Study design/ Pedro Score Methods Outcome Szaflarski et al., (2010) USA RCT PEDro=8 N=34 Individuals who has sustained a moderate to severe TBI and remained in hospital >24 hours, were selected and randomly assigned to either the levetiracetam (LEV) or phenytoin (PHT) group. Those in PHT group were given 20mg/kg intravenously to a maximum of mg given over 60 minutes. The Duration of treatment for the PHT group was 3 to 7 days, for the LEV group was 7 days. At time of discharge and post discharge there were no differences in the GOS. There were no differences in ICP, stroke, hypotension, arrhythmia, (etc.) between the two groups. Those in the LEV group experienced a worsening 12 Module 10-Post Traumatic Seizure Disorder-V9 2013

13 Author/Year/ Country/Study design/ Pedro Score Formisano et al., (2007) Italy/USA Pre/Post Dikmen et al., (1991) USA RCT PEDro = 6 Methods maintenance dose was 5mg/kg/day IV every 12h. Those in the LEV group were started on 20mg/kg/IV rounded to the nearest 250 mg over 60 min then started on a maintenance dose (1000 mg IV every 12 h over 15 min). Therapeutic dose was 1500 mg/12h. GCS, GOS, ICP, DRS were monitored and evaluated. N=137 Two groups were studied, one retrospectively (n=55) and the other prospective (n=83). Patients were given anti-epileptic medications for the treatment of seizures either before or after they began. N=244 Patients who fulfilled at least one of the following requirements: GCS 10 within 24 hours of injury, CT evidence of cortical contusion, depressed skull fracture, subdural/epdidural/intracerebral hemorrhage, penetrating head wound, seizure within the first 24 hours after injury were randomly assigned to be treated with phenytoin or placebo for 1 year starting within 24 hours of injury. Initial dose was 20 mg/kg intravenously. Serum levels of phenytoin were maintained at 3 6 µmol/l. Outcomes were assessed using a comprehensive battery of neuropsychological and psychosocial measures at 1, 12, and 24 months post-injury including: Haltead Reotan Neuropsychological Test Battery, Wechsler Memory Scale, Trails A/B and Stroop test part 1 and 2, Symptom Outcome of their neurological status less often than the PHT group (p=0.024). Those in the PHT group experienced anemia less often than the LEV group (p=0.076). DRS scores were lower for those in the LEV group at 3 and 6 months post discharge. Their GOSE were also higher at 6 months post intervention. In the retrospective study, the incidence of PTE was lower in the non-treated group, the time interval from TBI to seizure onset varied. In the prospective study the occurrence of late PTE was higher in patients treated with an antiepileptic medication vs those not treated (p=0.004). Those who were not treated with an antiepileptic medication (n=13) did not develop a seizure disorder. Out of those treated with medication (n=69) only 30 show epileptic abnormalities on their EEGs. Evidence was not found to support the use of long-term antiepileptic medications. In the severely injured (GCS 9) phenytoin significantly impaired neuropsychological performance at 1 month as evidenced by the overall ranksum type test (p < 0.05). No significant differences in neuropsychological performance were found between groups in either moderately injured (GCS 9) patients at 1 month or in either severity group at 1 year. Patients who stopped receiving phenytoin between 1 and 2 years improved more than corresponding placebo cases on several measures. Changes in neuropsychological measures from 12 to 24 months post-injury showed that phenytoin had negative widespread cognitive effects compared with placebo on most measures as evidenced by the overall rank-sum type test (p < 0.05). 13 Module 10-Post Traumatic Seizure Disorder-V9 2013

14 Author/Year/ Country/Study design/ Pedro Score Methods Checklist, Katz Adjustment Scale, Sickness Impact Profile. Outcome Dikmen et al., (2000) USA RCT PEDro = 8 Glotzner et al., (1983) Germany RCT No Score (German) Manaka (1992) Japan RCT PEDro = 3 McQueen et al., (1983) United Kingdom RCT PEDro = 7 N=279 Consecutive patients with TBI, admitted within 24 hours post-injury, at risk of post-traumatic seizures were randomized to one of three groups: 1) valproic acid (VPA) for 1 month followed by 5 months of placebo 2) Valrproic acid for 6 months or 3) phenytoin (PHT) for 1 week followed by placebo until 6 months post-injury. N=139 Individuals who had sustained a head injury and were over the age of 15 were randomized to treatment with carbamazepine on day 1 after accident for years or placebo. Phenytoin used if carbamazepine was not tolerated. N=126 Severe head injury patients (7-88 years of age) who showed either a disturbance of consciousness, bloody cerebrospinal fluid, focal neurological sign, depressed fractures or basal fracture, abnormal CT, dural tears, early convulsion, and/or linear fracture were randomized to receive phenobarbital (10 25 µg/ml) or no treatment started 4 weeks after head injury and continuing for 2 years. Partipants were gradually tapered off the medication during the third year. Follow up continued for 5 years. N=164 Individuals who had sustained a head injury (HI) between the ages of 5-65 years admitted within < 8 to > 30 days post injury and whose HI was complicated by at least one of the following: dural penetration, intracranial hameatoma, depressed skull fracture, persistent neurological deficit or PTA > 24 Trend towards higher mortality rate in the valproate groups compared with phenytoin group (p=0.07). There were no significant differences at 1, 6 or 12 months on the composite measures based on all the neurospsychological measures or only on the cognitive measures. Adjustment for differential confounders did not change the lack of significance of the results. No individual measure showed a significant difference among the treatment groups at 1, 6 or 12 months post-injury. Those on carbamazepine showed a significant decrease in post-traumatic seizures compared with placebo (p < 0.05). This difference was statically significant for early seizures (within first week) and for the follow up time in total, but not for late seizures. After a 5-year follow, there were no differences in the incidence of late seizures between those with severe TBI in the phenobarbital (16%) and those in the untreated (11%) group. No apparent differences between groups in the incidence of post-traumatic seizures for up to 24 months post-injury; 6 patients in the phenytoin group and 5 patients in the placebo group developed post-traumatic epilepsy within one year. A further 4 patient developed seizures between 1 and 2 years after injury. 14 Module 10-Post Traumatic Seizure Disorder-V9 2013

15 Author/Year/ Country/Study design/ Pedro Score Pechadre et al., (1991) France RCT (French) Methods hours were randomized to receive either phenytoin for 1 year or placebo. Patients with early seizures were excluded. N=86 Individuals between the ages of 3 60 years, and had sustained a TBI were randomly assigned to treatment with phenytoin (for 3 months or one year) or placebo. Phenytoin treatment started within the first 24 hours following brain injury. Follow-up continued for 2 years. Outcome After a 2-year follow-up, there was a significant difference (p< 0.001) in the incidence of late PTS between groups, with the treated patients showing an incidence rate of only 6% compared to an incidence rate of 42% in the untreated group. Smith, Jr. et al., (1994) USA RCT PEDro = 6 Temkin et al., (1990) USA RCT PEDro = 6 N=82 Those who had sustained an ABI and had previously been treated with phenytoin (n = 40) or carbamazepine (n = 42) after their head injury and who had been continuously maintained on their anticonvulsant drug for at least 6 months prior to this study were randomly assigned to continue their anticonvulsant therapy or switch to a placebo for a period of 4 weeks. A battery of neuropsychological tests was administered during a 4-week baseline period (treatment with stable doses of the patient s anticonvulsant continued), at the end of a 4 to 5 week period of continued drug treatment or placebo, and after 4 weeks of not receiving medication. N=404 Individuals who had sustained a severe ABI and met at least one of the following criteria: cortical contusion visible on a CT scan, subdural, epidural or intracerebral hematoma, a depressed skull fracture, a penetrating head wound, a seizures within 24 hours of injury, or a GCS 10 were randomly assigned to treatment with phenytoin (n=208) or placebo (n=196) Treatment was started within 24 hours of brain injury and continued for 1 year with a 2-year followup. Initial dose was 20 mg/kg intravenously within 24 hours of injury. Serum levels of phenytoin were maintained at 3 6 µmol/l. The incidence No significant differences were found between patients in the medication and placebo groups for either drug (phenytoin or carbamazepine) on any of the outcome measures at the end of the placebo phase. Patients in both drug groups (phenytoin and carbamazepine) showed significant improvements (p < 0.01) on several measures of motor and speeded performance following cessation of anticonvulsant drug treatment. During the first week of treatment, there was a significant reduction in the number of seizures in the phenytoin group compared with the placebo group (3.6% vs. 14.2%, p < 0.001). Between day 8 and the end of year 1 and at the 2 year follow up, there were no significant differences in the occurrence of seizures between groups (p > 0.2 for each comparison). This lack of late effect could not be attributed to differential mortality, low phenytoin levels, or treatments of some early seizure in patients assigned go the placebo group. 15 Module 10-Post Traumatic Seizure Disorder-V9 2013

16 Author/Year/ Country/Study design/ Pedro Score Methods of early (within 1 week) or late (> 8 days after initial drug loading) seizures was compared between groups. Outcome Temkin et al., (1999) USA RCT PEDro = 7 Young et al., (1983a) USA RCT PEDro = 6 Young et al., (1983b) USA RCT PEDro = 6 Servit and Musil (1981) Czechoslovakia Non-RCT N=379 Participants were initially randomly assigned to one of three groups: phenytoin for 1 week followed by placebo until 6 months post-injury, Valproate for 1 month followed by placebo for 5 months, or valproate for 6 months. 283 patients were followed up through 2 years. Occurrence of late seizures (> 7 days after injury) and mortality rates followed up for 2 years. N=244 Individuals with severe brain injury patients with either a penetrating missile wound or a blunt head injury were randomly assigned to receive either Phenytoin or placebo within first 24 hours post-injury. The incidence of early seizures (within one week of injury) between groups was assessed. N=179 Same method of randomization as that employed in Young et al. (1983a). Initially 214 patients were randomized, but only 179 patients completed the intervention. The incidence of late seizures between groups was compared. N=167 Moderate to severe (duration of loss of consciousness > 1 hour) brain injury patients were divided into two similar groups in terms of severity and localization of brain injury. The control group (n=24) was treated by conventional methods whereas the other group (n=143) was prophylactically treated with phenytoin ( mg/day) and phenobarbital (30-50 mg/day) orally for at least 2 years for the prevention of The rates of early seizures were low and not significantly different when using either valproate or phenytoin (1.5% in the phenytoin group and 4.5% in the combined valproate groups, p = 0.14). The rates of late seizures did not significantly differ between groups (15% in the 1-week phenytoin group, 16% in the 1-month valproate group, and 24% in the 6-month valproate group, p = 0.19). Mortality rates were not significantly different between groups, but there was a trend toward higher mortality rates in the valproate groups (7.2% in the phenytoin groups and 13.4% in the combined valproate groups, p = 0.07). There were no significant differences between groups in the percentage of early (within 7 days post-injury) seizures. There was no significant difference in the interval from injury to first seizure between the treated and placebo groups There were no significant differences between groups in the percentage of late (> 7 days post-injury) seizures. The incidence of late seizures was 25% in the control group (most occurring during the first year after injury) and 2.1% in the prophylactically treated group following the discontinuance of drug therapy (p<0.001). 16 Module 10-Post Traumatic Seizure Disorder-V9 2013

17 Author/Year/ Country/Study design/ Pedro Score Methods seizures. The incidence of late seizures was compared between groups. Outcome Watson et al., (2004) USA Cohort Wroblewski et al., (1992) USA Case Series Wroblewski & Joseph (1992) USA Case Study No Score N=404 Individuals with severe TBI (GCS 10) who were exposed to glucocorticoid medications (n = 125) within 1 week of TBI were compared with those without any exposure (n = 279) to glucocorticoids. Any type, dose or route of administration of glucacorticoids was considered significant. Patients were followed up for 2 years to compare seizure development and mortality rates. N=30 Severe (PTA > 24 hours) brain injury patients with documented late posttraumatic seizures (> 7 days after injury) who were being treated with methylphenidate for a variety of cognitive and behavioral problems participated in this retrospective chart review. All but 2 of the 30 patients were receiving concurrent anticonvulsant drug therapy (carbamazepine n = 26, valproic acid n = 2). Any seizure occurrence, anticonvulsant use and other comedication use was recorded for (1) 3 months prior to methylphenidate use (baseline); (2) the duration of methylphenidate treatment and (3) 3 months proceeding methylphenidate use (follow-up). N=10 Patients who had suffered a TBI resulting from a motor vehicle accident, fall or anoxic encephalopathy received intramuscular midazolam after other benzodiazepine drugs proved not useful for acute seizure cessation. Seizure occurrence and behavioural changes were assessed. PEDro = Physiotherapy Evidence Database rating scale score (Moseley et al., 2002). Patients dosed with glucocorticoids within 1 day of TBI were more likely to develop first late seizures than were those without the medication (p = 0.04). This early glucocorticoid exposure was associated with an increase of 74% in the risk of a first seizure. Receiving glucocorticoids 2 days after TBI was not associated with first late seizure development (p = 0.66). Receiving glucocorticoids within 1 day (p = 0.28), or 2 days (p = 0.54) after TBI was not associated with the development of a second late seizure. No difference was noted in cumulative mortality between groups (p = 0.57). There was no statistically significant difference in seizure occurrence while patients were received methylphenidate compared with the period when patients were off the drug on measures of seizure occurrence, however there was a trend toward a lesser incidence of seizures in patients during methylphenidate treatment (p = 0.063). All patients experienced seizure cessation within minutes of midazolam administration. Slight to moderate sedation were the only reported side effects. In those individual treated for behavioural problems, midzolan alleviated agitation and violence with no obvious effects on cognition. 17 Module 10-Post Traumatic Seizure Disorder-V9 2013

18 Table 10.6 Summary of RCT Studies of Seizure Prevention or Prophylaxis Authors/Year Methods Results Szaflarski et al., 2010 Dikmen et al., (1991) Dikmen et al., (2000) Glotzner et al., (1983) Manaka (1992) McQueen et al., (1983) Pechadre et al., (1991) Smith, Jr. et al., (1994) Temkin et al., (1990) Temkin et al., (1999) Young et al., (1983a) N = 34 Patients admitted to ICU post ABI. Phenytoin and levetiracetam were given to patients within 24 hours of admission. N = 244 high risk patients. Phenytoin vs. placebo for 1 year starting within 24 hours of injury. N = 279 of high risk patients. Valproate x 1 mos + placebo x 5 mos; valproate x 6 mos; phenytoin x 1 wk; + placebo x 6 mos. N = 139 Carbamazepine at day 1 x years or placebo N = 126 Phenobarbital x 2 years vs. no treatment at all with 5 year follow-up. N =164 Phenytoin x 1 year. Excluded early seizures. N = 91 Phenytoin x 3 mos vs. 1 year with 2 year follow-up. N = 82 patients previously treated with phenytoin or carbamazepine randomly assigned to continue on anticonvulsant or switch to placebo N = 404 Phenytoin x 1 year vs. placebo with 2 year follow-up. N = 379 high-risk patients. Phenytoin x 1 week; valrproate x 1 mos; valproate x 6 mos. Follow up for up to 2 years N = 244 Severe brain injury patients. Phenytoin vs. placebo within 24 hours post-injury. N = 214 Phenytoin or placebo within first 24 hours post-injury. -Levetiracetam was shown to be as safe as phenytoin in reducing seizures post ABI. Levetiracetam was found to have fewer side effects and better GOSE outcomes 3 months post discharge. - for neuropsychological performance at 1- month in severely injured (GCS 9) ND for neuropsychological performance in moderately injured (GCS > 9) at 1-month or either severity group at 1 year. Trend towards higher mortality in valproate vs. phenytoin. Otherwise ND on any measure. + for early seizures ND for later seizures ND for late seizures ND for post-traumatic epilepsy up to 24 months post-injury. + for late seizures + for performance on several measures of motor and speeded performance following cessation of anticonvulsant drug treatment. + for reduction in seizures in 1 st week ND in seizure occurrence after 1 st week. ND between drugs, both Phenytoin and valproate + for early seizures. ND between drugs, both negative for late seizures. Trend toward higher mortality in valproate vs. phenytoin. ND for incidence of early seizures (within 7 days post-injury) Young et al., ND for incidence of late seizures (> 7 days (1983b) post-injury). ND = No difference between groups; + = Improvement compared with control; - = Impairments compared with control 18 Module 10-Post Traumatic Seizure Disorder-V9 2013

19 Table 10.7 Summary of non-rct Studies of Seizure Prevention or Prophylaxis Authors/Year Methods Results Formisano et al., (2007) Watson et al., (2004) Wroblewski et al., (1989) Wroblewski et al., (1992) Wroblewski and Joseph (1992) N=137 Two groups of patients were evaluated (one retrospectively and prospectively) after sustaining a severe. Seizure activity was monitored in all patients even if an anti-epileptic medication was not prescribed N=404 Comparison of TBI patients exposed or not exposed to glucacorticoids within 1 week of injury. Follow up for 2 years. N=27 Patients with late occurrence seizures or at high risk for PTS. Taken off phenytoin, phenobarbital or primidone and treated with carbamazepine instead. N=30 ABI patients at risk of developing seizures who were receiving anticonvulsant drug therapy and concurrently receiving methylphenidate for a variety of cognitive and behavioural problems. N=10 TBI patients. Treated with intramuscular midazolan ( mg) for acute seizure cessation. ND = No difference between groups; + = Improvement compared with control In both groups PTE was found to be lower in the non-treatment group. In the prospective study the occurrence of late PTE was higher in patients treated with an antiepileptic medication than those not treated. ND in mortality or late seizures. Early (< 2 days after TBI) exposure to glucocorticoids is associated with higher risks of seizures. ND in seizure occurrence between pre- and pos carbamazepine periods. Trend towards lower frequency of seizures while on methylphenidate. + for seizure cessation within minutes of midazolan administration Discussion Szaflarski et al. (2010), looked at the effects of levetiracetam (LEV) or phenytoin (PHT) on individuals (n=34) who has sustained a moderate to severe TBI. All subjects were randomly assigned to one of two groups the LEV or PHT. Subjects in the PHT group were administered 20mg/kg PE IV up to a maximum of mg given over 60 minutes. The PHT s maintenance dose was 5mg/kg/day IV every 12h and was administered to patients for the next 3-7 days. The LEV group was started on 20mg/kg/IV which was rounded to the nearest 250 mg over the first 60 minutes. The group was then started on a maintenance dose (1000 mg IV every 12 h over 15 min) and remained on this for 1-7 days (Szaflarski et al., 2010). The Glasgow Coma Scale (GCS), the Glasgow Outcome Scale (GOS), Intracranial Pressure (ICP) and the Disability Rating Scale were all used to evaluate the groups pre- and post treatment. Results indicate that those on LEV performed significantly better on the DRS at 3 months (p=0.042) and the GOS at 6 months (p=0.039) post intervention compared to the PHT group. There were no differences in the number of seizures experienced by study participants regardless of the group they were assigned to (Szaflarski et al., 2010). 19 Module 10-Post Traumatic Seizure Disorder-V9 2013

20 Overall, most published randomized control trials have failed to find any favorable evidence for prophylaxis of late PTS (Young et al., 1983b; Glotzner et al., 1983; McQueen et al., 1983; Temkin et al., 1990). In some of these reports, the occurrence of PTS was actually higher in patients treated with anticonvulsant medications (Young et al., 1983b; McQueen et al., 1983; Temkin et al., 1990; Formisano et al., 2007). In contrast to late PTS, Yablon and Dostrow (2001) have reported that anticonvulsant drug prophylaxis consistently reduces the incidence of early PTS (Glotzner et al., 1983; Temkin et al., 1990; Temkin et al., 1999; Young et al., 1983a). Similar to most of the studies on late PTS, there is no evidence that anticonvulsant drug prophylaxis of early seizures reduces the occurrence of late PTS or has any effect on mortality or neurologic disability (Schierhout & Roberts, 1998). As mentioned earlier, patients who suffer severe penetrating brain injuries experience higher risks of PTS. However, due to the small number of patients with penetrating TBI in most reports, it is not clear if anticonvulsant drug prophylaxis has any effects, either positive or negative, on the occurrence of PTS is this subgroup of brain injury patients. Schierhout and Roberts (2001) conducted a Cochrane review of anti-epileptic drugs for preventing seizures following acute traumatic brain injury. This review of six trials including 1218 randomized patients. The authors noted that all of the included trials were restricted to patients considered to be at high risk for PTS. Most of the trials excluded patients who already had one seizure with the intervention started within 24 hours of admission in four trials; the remaining two trials initiated the intervention between 4 and 8 weeks post head injury. Four trials used phenytoin, one trial phenobarbital and one trial carbamazepine. Drug was provided between 1-2 years in all the trials duration of follow-up was two years in most cases. Loss to follow-up ranged from 5-72%. Four trials reported early and late seizure; two trials reported late seizures only (Schierhout & Roberts, 2001). The number of patients with early seizures (within the first week after injury) was available from 4 trials representing 890 randomized subjects. The pooled relative risk for early seizure prevention was 0.34 (95% CI 0.21, 0.54) with the number needed to treat to keep one patient seizure-free in the acute phase was 10. However, there was no evidence that this resulted in a reduction in mortality or a reduction in death or persistent vegetative state. The occurrence of late seizure was not reduced by antiepileptic prophylaxis at any time. There was insufficient data to examine non-fatal adverse effect, with the exception of a trend towards an increased risk of skin rashes (Schierhout & Roberts, 2001). The authors Schierhout and Roberts (2001) note that, There is no evidence that prophylactic anti-epileptics used at any time after head injury, reduce death and disability. There is evidence that prophylactic anti-epileptics reduce early seizures, but 20 Module 10-Post Traumatic Seizure Disorder-V9 2013

21 this is not supported by a reduction in late seizures. Insufficient evidence is available to establish the net benefit of treatment at any time after head injury. (pg 1) There appears to be very little research to evaluate the efficacy of anticonvulsants given to treat seizures after they have occurred. We identified only one such study in this review. Wroblewski and Joseph (1992) reported on a collection of ten case studies of TBI patients treated with intramuscular midazolam for acute seizure cessation after other benzodiazepine drugs had failed. The authors reported that in all patients, seizures ceased within minutes of midazolam administration. Midazolam also prevented the onset of prolonged seizures or status epilepticus. Slight to moderate sedation were the only reported side effects. Conclusions: There is Level 1b evidence to suggest levetiracetam is as safe and effective as phenytoin in treatment and prevention of seizures in individuals in the intensive care unit post ABI. Based on meta-analysis and the findings of this review there is Level 1a evidence that anticonvulsants given during the first 24 hours post-abi reduce the occurrence of early seizures (within the first week post-injury). There is Level 1a evidence that anticonvulsants given shortly after the onset of injury do not reduce mortality or persistent vegetative state or the occurrence of late seizures (> one week post-injury). There is Level 1a evidence that seizure prophylactic treatment with either phenytoin or valproic acid results in similar incidences of early or late seizures and similar mortality rates. There is Level 1b evidence that both phenytoin and carbamazepine have negative effects on cognitive performance, particularly on tasks with motor and speed components, which theoretically may have a negative impact upon learning during rehabilitation. There is Level 2 evidence that glucocorticoid exposure after brain injury is not associated with a decrease in late seizures, and early exposure (< 2 days after injury) is associated with increased seizure activity. There is Level 4 evidence that methylphenidate for the treatment of cognitive and behavioral problems can be safely used in brain injured patients at risk for posttraumatic seizures as it is not associated with an increase in seizure frequency. 21 Module 10-Post Traumatic Seizure Disorder-V9 2013