There is insufficient evidence to support an evidence-based recommendation related to the benefits and harms of seizure prophylaxis in the critical care setting.

In the absence of direct scientific evidence, EXPERT CONSENSUS concluded that:

• Anti-seizure prophylaxis is indicated for a minimum of 7 days in patients with pTBI to reduce early post-traumatic seizures. (100% consensus)
• No specific anti-epileptic drug can be recommended as being superior for the prevention of early post-traumatic seizures. (100% consensus)

No evidence or expert opinion supported distinct recommendations based on patient gender, age, wounding mechanism, or military vs. civilian context.

The association between seizures and pTBI has been described since antiquity and subsequently reported in the medical literature associated with most major conflicts. While understood to be more prevalent than in blunt TBI, the rate of seizures after pTBI is, however, difficult to elucidate given the vast heterogeneity of the study populations. Indeed, the risk of readmission for seizure in nearly 3 times higher in pTBI patients than blunt TBI patients.6

Post-traumatic seizures following pTBI are relatively common, though the exact incidence is unknown. The extent and location of injury, presence, and type of retained material, and follow up time are often not fully described or disparate between studied populations, creating difficulty when trying to make any comparative deductions. Transcranial bihemispheric or intraventricular injury especially increases the risk of seizures considerably1. Much more is known about post-traumatic seizures in blunt TBI. Seizures following blunt severe TBI, focal or generalized, occur in up to 20% of patients.2 They cause increased sympathetic drive, increased serebral blood flow and can precipitate increase in intracranial pressure - or even hernation - as well as metabolic crises leading to secondary injury.3 Notably, prophyaxis against early seizures does not impact development of post-traumatic epilepsy.4 which occurs in 1.06% of combat veterans.4 Other risk factors in pTBI are dural penetration with metal and bone fragments, multiple cranial surgeries, multiple subcortical contusions, subdural hematoma, midline shift greater than 5mm, and GCS score.5

The prior penetrating head injury guidelines addressed "Antiseizure Prophylaxis for Penetrating Brain Injury."7 Recommendations created in this edition were based on 9 articles with 4 defined as Class I and 5 defined as Class III. No Level I or Level II recommendations on this subject were provided. Level III/IV recommendations stated antiseizure medication in the first week following pTBI was recommended, and subsequent prophylaxis could not be recommended. Direct evidence supporting this recommendation was lacking, however. None of the Class I studies met inclusion criteria for the updated pTBI guidelines because most of the patients studied had blunt TBI, while a small number had penetrating TBI. Moreover, results were not presented based on TBI type.

Amongst these studies, some explored onset of early seizures (<7days), late seizures (>7 days) and post-traumatic epilepsy as a syndrome. This temporal distinction established the convention for classification of early and late seizures.8 Most studies showed little benefit of prophylactic anti-epileptics in reducing rates of post-traumatic epilepsy, while carbamazepine9 and phenytoin8 reduced risk of seizures in wthe first week in the mixed TBI cohort. Since it is not possible to extricate data on pTBI patients from these publications, none of them were included for the specific evidence-base for pTBI.

In the last iteration of the guidelines for management of severe blunt TBI, a Level IIA recommendation was made for the use of phenytoin to decrease the incidence of early post-traumatic seizures when the overall benefit is felt to outweigh the risks of this treatment.10 It must be borne in mind that admisnitering phenytoin in the intensive care unit to critically ill patients can have side effects. With enteral administration, feeding needs to be paused for 1 hour prior and 2 hours after administration for optimal absorption, and blood levels need to be checked to ensure high therapeutic levels and to avoid toxicity or underdosing. Severe allergic reactions can be seen including rare incidence of toxic epidermal necrolysis. Phenytoin is metabolized by cytochrome P450 and interacts with a long list of medications, which can alter levels if either drug unpredictably. Lastly, when administered intravenously, cardiac, and hemodynamic risks exist. It is acknowledged that levetiracetam is now commonly used as seizure prophylaxis however literature support for this practice is lacking despite an apparent favorable side-effect profile and neuroprotective properties [cite OBTT]. Throughout this document, we have made efforts to compare evidence specific to pTBI management.

Early vs Late Posttraumatic Seizures

While there is consensus that a posttraumatic seizure occurring in the first week after injury is considered "early" and subsequent episodes are "late" seizures, this is largely based on non-penetrating TBI literature.8 Of the 13 articles used to support this update, early seizures were defined by two studies as occurring within the first week and the remaining 11 failed to stratify events by timing of occurrence during hospital stay.

Prophylactic Use of Anticonvulsants

We identified no studies that compared seizure prophylaxis with no seizure prophylaxis or a different seizure prophylaxis in patients with pTBI.

We identified 11 noncomparative studies (N=1227) that reported the rate of seizures after seizure prophylaxis.11-21 Sample sizes ranged from 21 to 520, with most studies having fewer than 100 patients. Six studies were in civilians,13-18 while the other five were in military or war-affected populations. Seizure prophylaxis was mainly phenytoin (or diphenylhydantoin) 300 to 500 mg/day for adults and 4 to 7 mg/kg/day in children, when it was reported. One study administered levetiracetam12 and one study reported administering barbiturates (drug not specified) for seizure prophylaxis,21 while three studies did not report the type of seizure prophylaxis used.15,18,19

Four studies (N=207) reported no deaths occurred,14,17,19,20 one study (N=99) reported 5% of patients died within 10 and 15 days of injury11, one study (N=36) reported that five patients died but it was unclear the proportion of patients that died with penetrating injuries,12 four studies (N=709) reported between 13% and 23% of patients died within 0 to 3 months,13,15,16,18 and one study (N=176) did not report deaths.21

Four studies (N=354) reported that no seizures occurred with diphenylhydantoin or barbiturates given prophylactically,11,14,20,21 while other studies reported low rates of seizures (6 studies [N=795] reported 5% to 14% of patients developed a seizure after prophylaxis,12,13,15-17,19 1 study [N=78] reported a 28% posttraumatic seizure rate18). Three of the studies indicated seizure prophylaxis was continued from the emergency department,11,18,20 while one study initiated seizure prophylaxis in the emergency room only in ventilated individuals.15 Three studies reported initiating seizure prophylaxis in the intensive care unit (ICU),13,17,19 while the remaining four were unclear when the medications were initiated.12,14,16,21

Only one small study from Petridis et al reported 27 patients with gunshot wounds to the head reported not giving seizure prophylaxis; one patient had symptomatic convulsions after cranial reconstructive surgery, which occurred 6 months after the initial injury, with no seizures experienced in the remaining patients.22 The majority of patients had a Glasgow Coma Scale score of 3 to 8 (70%), and the overall mortality was 47%. The authors conclude from this small study that only patients with retained fragments and evidence of epileptiform activity on EEG should receive anticonvulsants.

Atwood et al (N=36) administered levetiracetam for seizure prophylaxis in Iraq and Afghanistan war veterans, reported two patients with pTBI (5.5%) had a posttraumatic seizure.12 Both patients suffered a transcranial bihemispheric gunshot wound, whereas most patients who did not have a seizure suffered a blast injury. Both patients who had a posttraumatic seizure eventually died. Although this study did include patients who did not receive seizure prophylaxis, these patients were cleared by neurosurgery for no prophylaxis or had isolated subarachnoid hemorrhage and likely did not have a penetrating injury.

Safety of Anticonvulsants

Cosar et al provided a large study (N=400) which reported harms after prophylactic phenytoin which was administered using an 11 mg/kg intravenous (IV) and a 13 mg/kg intramuscular (IM) parenteral loading dose, and oral maintenance dose (therapeutic blood concentrations of 10-20 mcg/mL) and were followed for 12 to 36 months after surgery.23 A skin reaction occurred in 11% of patients between 5 and 70 days after surgery, and of those, two patients developed exfoliative dermatitis.

The benefit of anticonvulsant medications in the early post injury period with pTBI patients is widely accepted though it must be remembered that risks of antiepileptic drugs must be considered. However, goiven the risks inherent to phenytoin use in intensive care many now administer levetiracetam given its availability, and low cost and improved side-effect profile. However, encephalopathy in geriatric patients and those with renal impairment must still be reason to deliberate.

The evidence presented in the previous guidelines has received further support from subsequently published literature. While this consisted entirely of case studies, the data continue to substantiate administering anticonvulsants to pTBI patients in the first seven days post injury. No new evidence was reviewed to recommend anticonvulsants after one week. In addition, only one study utilized levetiracetam for most patients and had a low incidence of early post-injury seizures; therefore, while there isn't evidence to favor utilizing one medication over another, there is evidence underlying the use of levetiracetam as initial seizure prophylaxis.

Penetrating TBI patients are at high risk of seizures and at higher risk than patients with blunt TBI. Seizures can precipitate intracranial hypertension or its worsening and cerebral herniation events. Seizure prophylaxis is well established in severe TBI care and is generally administered to patients with blood on more than one CT head slice. Seizure prophylaxis is likely even more important in pTBI patients. Injuries to the frontal and temporal lobes are likely more epileptogenic than injuries to other brain regions.

Although phenytoin has traditionally been administered as prophylaxis for TBI patients and has the most evidenciary support, levetiracetam has emerged as the preferred agent in modern practice. Though supportive literature is deficient, levetiracetam seems to be both efficacious and safe - perhaps safer than phenytoin which is associated with risk of very serious complications such as drug interactions, hemodynamic instability and rarely toxic epidermal necrolysis. Interestingly, Operation Brain Trauma Therapy which systematically studied promising pre-clinical neuroprotective agents, found that levetiracetam was the most promising pharmaceutical studied though relevant human studies are awaited.

Our group did not reach consensus related to the duration of anticonvulsant therapy in pTBI nor criteria for the cessation of prophylaxis. The notion of obtaining a neurology consultation and potentially EEG testing prior to cessation of seizure prophylaxis was discussed but did not receive consensus support from the group.

Further high-quality studies are required pertaining to the use of anti-seizure prophylaxis based on injury mechanism and neurological injury, as well as duration of use and procedures for seizure prophylaxis in pTBI patients. Such studies could be readily conducted in environments treating a high volume of pTBI patients. Studies comparing different agents are desirable as well as those exploring differences in best care between blunt and penetrating TBI patients. Some pTBI patients are likely higher risk for seizures such as those with frontal or temporal lobe injuries, bihemspheric or intraventricular injury. It is possible that ideal treatment protocols for such patients may differ from other pTBI or TBI patients. Given that seizure prophylaxis is an entrenched aspect of TBI care there may be ethical challenges inherent to some investigations. Not withstanding this, with the access to continuous video electroencephalograms with computer tomography and magnetic resonance imaging compatible electrodes, another option may be explored is to use antiseizure treatment when certain electrographic markers of cortical irritability are seen, or actual non-convulsive seizures.

Though clinical trials are needed, comparative effectiveness studies may also be undertaken with propensity matching between cohorts who are rountiely prescribed antizseisure medications and those who are not. For this and several other topics in this guideline, the establishment of a penetrating TBI research consortium is essential.