Introduction

Traumatic brain injury (TBI) is a clinically heterogeneous condition, ranging from mild to severe forms, and represents one of the most common causes of acquired epilepsy, particularly among adults and elderly patients. The overall incidence of post-traumatic seizures in developed countries ranges from 4 to 53%¹; this rather imprecise estimate is due to the high heterogeneity of studies, which differ for a variety of variables, including age of patients, severity of trauma, length of follow-up and often do not differentiate between early and late post-traumatic seizures.

Post-traumatic epilepsy accounts for approximately 20% of symptomatic epilepsy in the general population² and should be classified as a structural epilepsy.³ Epileptic seizures may develop even several years after the head injury, and their occurrence depends mainly on the severity of the trauma. An epidemiological study has shown that the cumulative risk of developing post-traumatic seizures within 5 years from the event is 0.5% among patients with mild, 1.2% in those with moderate and 10% in those with severe TBI.⁴

After briefly discussing the differences between early (acute symptomatic) and late (unprovoked) post-traumatic seizures, with a focus on the underlying pathophysiology and long-term risk of seizure recurrence, this chapter will discuss the methodological issues and challenges of antiepileptogenesis trials and the role of antiepileptic drugs (AEDs) and neuroprotective agents in preventing post-traumatic seizures. Although animal studies have reported successful attempts to halt the process of epileptogenesis, translation into clinical trials has so far proved inconclusive. This chapter will not deal with preclinical studies, which have been already critically reviewed.^5–7

Early Acute Symptomatic and Late Unprovoked Post-traumatic Seizures

In daily practice, post-traumatic epilepsy can be diagnosed by any of the following conditions: “(1) At least two unprovoked post-traumatic seizures occurring >24 hours apart; (2) one unprovoked post-traumatic seizure and a probability of further seizures similar to the general recurrence risk (at least 60%) after two unprovoked seizures, occurring over the next 10 years”.⁸ The electro-clinical assessment together with the imaging findings showing the TBI should “lead to a reasonable inference that the imaging abnormality is the likely cause of the patient’s seizures”.³

By definition, unprovoked post-traumatic seizures occur more than 7 days after the acute TBI.⁹ They are also termed “late post-traumatic seizures” as opposed to “early post-traumatic seizures”, which occur within the first 7 days after the head trauma. Although somewhat arbitrary, this straightforward temporal cut-off reflects a different underlying pathophysiology. Early post-traumatic seizures are acute symptomatic seizures occurring at the time of, or in close temporal association with, the brain injury.⁹ They are due to transient and acute cellular biochemical dysfunctions leading to alterations in cortical excitability and should be regarded as a “reaction” of the brain to the lesion itself.⁹ The incidence of early post-traumatic seizures ranges from 2.1% to 16.9%.¹ They carry a low risk of long-term seizure recurrence over the following 10 years (13.4% after a first seizure; 95% confidence intervals, CI = 7.0–24.8%)¹⁰ and, hence, do not entail a diagnosis of post-traumatic epilepsy. Conversely, unprovoked seizures occurring more than 7 days after the head trauma are associated with a structural disruption of neuronal networks, increasing the excitability of the brain and leading to an enduring predisposition to generate epileptic seizures (i.e., post-traumatic epilepsy).¹¹ The estimated incidence of late post-traumatic seizures ranges from 1.9% to 30%.¹ A first unprovoked post-traumatic seizure has a risk of seizure recurrence over the next 10 years of 46.6% (95% CI = 30.4–66.3%).¹⁰ Hence, the risk of long-term seizure recurrence after a first unprovoked post-traumatic seizure is significantly higher than after a first acute symptomatic post-traumatic seizure (p < 0.001).¹⁰ However, in this study the risk of seizure recurrence over the next 10 years after a first unprovoked post-traumatic seizure was lower than the threshold of 60% required for the diagnosis of epilepsy after a single epileptic crisis,¹ and the upper confidence interval was only slightly higher. It should be noticed that this study was conducted in patients with TBI, but no details on severity or characteristics of head trauma were provided.¹⁰ Of note, a previous longitudinal cohort study conducted in 63 patients with moderate to severe head trauma developing late post-traumatic seizures showed a cumulative incidence of recurrent late seizures of 86% by approximately 2 years.¹² Factors increasing the risk of long-term seizure recurrence should be, hence, considered on an individual basis: they include cerebral contusion and intracerebral hematoma, skull fracture, bony skull defect or penetrating injuries (e.g., bone or metal fragments, craniectomy or cranial surgery), loss of consciousness or amnesia lasting more than one day, post-traumatic seizures within the first week, severity of head trauma and age ≥ 65 years at the time of TBI.^{1, 4} Adequate knowledge of these variables is important to clinically evaluate the risk of seizure recurrence and epilepsy after a single unprovoked seizure.

Epileptogenesis after Traumatic Brain Injury

In patients developing post-traumatic epilepsy there is usually a period of time of different duration, during which no epileptic seizures occur. During this clinically silent period, the brain undergoes progressive neuronal changes, which increase its excitability and eventually lead to the occurrence of recurrent spontaneous seizures or post-traumatic epilepsy.¹³ This dynamic and chronic process is known as epileptogenesis and occurs in any patient developing structural epilepsy after an initial brain injury, irrespective of its etiology (e.g., stroke, tumor, trauma) (Figure 12.1).

Figure 12.1 Factors influencing epileptogenesis after status epilepticus, TBI or stroke. The primary brain insult leads to “parallel and sequential molecular and cellular events that lead to various functional impairments, including sensorimotor, memory and emotional decline as well as epileptogenesis”.¹³ During the process of epileptogenesis, the neuronal circuitry reorganization leads to increased hyperexcitability and the occurrence of recurrent spontaneous seizures.

Reproduced from Pitkänen et al (2007),¹³ with permission. Blackwell Publishing, Inc. ©International League Against Epilepsy.

Molecular and cellular alterations occurring in the brain after a brain-damaging insult include: selective neuronal cell death and apoptosis, changes in membrane properties, mitochondrial changes, receptor changes (e.g. loss of GABAergic receptors), deafferentation and collateral sprouting.^{7, 13} These changes lead to circuitry reorganization and, hence, to permanent hyperexcitability and development of recurrent spontaneous seizures (epilepsy).¹³

Recent data have shown that neuroinflammation following a TBI plays a relevant role in increasing the propensity to epileptic seizures through a variety of mechanisms, such as the activation of microglia and astrocytes, release of cytokines, disruption of the brain-blood barrier and progressive brain edema.¹⁴ Intriguingly, a polymorphism of the interleukin-1β (IL-1β rs1143634), which is a pro-inflammatory cytokine released in the brain by astrocytes and microglia, was found to be associated with post-traumatic epilepsy in 47.7% of the cases.¹⁵ This finding suggests that genetics may also influence the neuroinflammatory cascade and epileptogenesis after a neurotrauma.

Clinical Challenges in Antiepileptogenesis Trials

Antiepileptogenesis treatments are intended to modify the process of epileptogenesis, preventing or minimizing the risk of developing structural epilepsy after an insult to the brain. Clinical trials assessing the neuroprotective properties of compounds used to prevent epilepsy face many methodological issues. These challenges apply also to antiepileptogenesis trials of post-traumatic seizures, which are acknowledged to be more complex and expensive than conventional pharmacological studies in epilepsy.¹⁶

Antiepileptogenesis trials require longer duration and follow-up than trials assessing drugs for the treatment of epilepsy as seizures and epilepsy may develop even several years after the primary brain lesion.¹⁶ More specifically, most patients (86%) with a first late post-traumatic seizures develop a second seizure within 2 years¹²; the risk remains remarkable in the following years, and it is higher following severe than moderate TBI (10 versus 30 years).⁴ A long trial is, hence, needed to adequately evaluate the efficacy of the tested drug in preventing the occurrence of spontaneous seizures and structural epilepsy. It is, however, worth considering that a longer duration of the study translates into higher rates of patients lost to follow-up, which would affect the power of the study and make it less informative. At the same time, a large number of subjects recruited to compensate for the expected high proportion of patients lost to follow-up would inevitably lead to high costs. Conducting the study in an unselected population (i.e., in patients with low baseline risk of developing structural epilepsy after a brain injury) would also require the inclusion of a large number of patients to obtain informative results, again resulting in high costs.¹⁶ Conversely, carrying out an antiepileptogenesis trial in a selected population (i.e., adopting a population enrichment design) would increase the informative value of the study and its feasibility by including a lower number of patients and reducing costs.^16–18 The identification of patient at high risk of developing post-TBI epilepsy on the basis of reliable predictors is required to design an effective antiepileptogenesis trial.^{16, 19} Using a cohort from the US Traumatic Brain Injury Model Systems National Database, a model to predict the risk of post-traumatic seizures was recently developed. The model identified subdural hematoma, contusion load, craniotomy, craniectomy, early post-traumatic seizures and preinjury incarceration as predictors for post-traumatic seizures at 1 and 2 years after the injury.²⁰ This model showed fair to good predictive accuracy and could be used to identify and select a study population of patients at high risk of developing late post-traumatic seizures.

Other methodological issues which should be carefully considered in antiepileptogenesis studies after a neurotrauma include the co-occurrence of risk factors (e.g., severe head trauma and associated craniectomy or craniotomy), which could compete for the development of epilepsy in the target population and the need for rapid informed consent (including, whenever required and applicable, surrogate and waiver for informed consent).^16–18

A further crucial issue is the need to disentangle the antiseizure effect of a drug from its antiepileptogenic effect. This can be achieved only if the prevention trial first assesses the antiseizure effect in a randomized controlled phase (usually versus placebo) and then the antiepileptogenic effect after washout.¹⁸

Other aspects to be considered concern the ascertainment of diagnosis among participants. Subclinical early post-traumatic seizures can go unnoticed, particularly in patients with impairment of consciousness or sedation.² Continuous video-EEG monitoring could enable to detect subclinical seizures, but this would be rather unfeasible and expensive in a large sample.²¹ Furthermore, detecting late post-traumatic seizures will be even more difficult, unless patients are under close medical supervision during the entire study duration and investigators have a low threshold for suspecting seizures and asking for EEG monitoring. However, this could be hard to reach in a trial with long duration and not conducted in specialized epilepsy centers.

Finally, patients with TBI allocated to any arm should not use other agents with neuroprotective potential, in order to isolate the antiepileptogenic effect of the tested compound. If study groups receive concomitant interventions, the risk of co-intervention bias is particularly high. Unfortunately, benzodiazepines or barbiturates are frequently used for patients with severe TBI requiring sedation, and they could influence the outcome independently from the medication being studied.

Antiepileptogenic Therapies for Preventing Post-traumatic Seizures: The Evidence from the Literature

On May 28 2019, we performed a systematic search of MEDLINE (accessed through PubMed) to identify any randomized controlled trial (RCT) or systematic review assessing the use of AEDs or other neuroprotective agents for the primary prevention of post-traumatic seizures. The following search strategy was used: “(epilepsy OR seizure*) AND randomi* AND trauma”.

With the aim to focus on post-traumatic seizures and provide specific data about this condition, we included only studies restricted to TBI and excluded those performed in patients undergoing brain surgery (craniectomy or craniotomy) for other indications, like brain tumors or primary intracerebral hemorrhage. Additionally, we included only RCTs and high-quality systematic reviews of the literature to rely on the highest currently available evidence.

In total, 196 results were retrieved; among them, a recent Cochrane systematic review of RCTs compared the efficacy of AEDs and neuroprotective agents with placebo, usual care or other pharmacologic agents for the prevention of post-traumatic epilepsy in people with TBI of any severity.²² The outcomes measured included early (within 1 week of trauma) and late (later than 1 week post-trauma) seizures (Table 12.1).

Table 12.1 Characteristics of included studies

No additional RCTs have been published after the aforementioned Cochrane work, which currently represents the best source of information to get guidance on the choice of antiepileptogenic therapies for preventing post-traumatic seizures. The main results are reported in the following sections.²²

Related posts:

Chapter 5 – Post-traumatic Epilepsy in Children Chapter 6 – Sport-related Concussive Convulsions Chapter 7 – Accidents and Injuries during Seizures Chapter 11 – Post-traumatic Epilepsy and Post-traumatic Stress Disorder Chapter 13 – Effects of Antiepileptic Drugs on Cognition Chapter 14 – Post-traumatic Epilepsy in Low Income Countries

Stay updated, free articles. Join our Telegram channel

Join