Post–Acquired Brain Injury Epilepsy
ROBERT J. KOTLOSKI
Epilepsy resulting from an acquired brain injury (ABI), that is acquired epilepsy, is a significant clinical problem that is likely to increase in the future owing to medical advances that will allow for improved survival following ABI and to shifts in demographics resulting in the growth of elderly populations at high risk for ABI and epilepsy. Although the capacity of ABI to cause epilepsy has been known for millennia, our understanding of the pathophysiology of acquired epilepsy and our abilities to diagnose and to treat remain severely limited. However, with increased awareness of ABI and its sequelae, including epilepsy, it can be reasonably hoped that improvements will be forthcoming.
In this chapter, we will overview acquired epilepsy with a focus on posttraumatic epilepsy (PTE), including epidemiology, pathophysiology, and clinical care. In addition, we will briefly discuss other common etiologies of post-ABI epilepsy, including poststroke epilepsy and tumor-associated epilepsy (TAE). A case study will be provided at the end of the chapter to provide further consideration of disability and life-altering consequences of ABI and epilepsy.
To address the issue of epilepsy resulting from an ABI, it is important to first review the definitions of relevant terms. As noted so far in this book, ABI refers to a wide spectrum of brain injuries resulting from nongenetic and noncongenital causes, typically divided into traumatic or nontraumatic etiologies. ABIs include traumatic brain injury (TBI) or cerebral concussion, stroke, anoxia, tumor, intracranial and intracerebral hemorrhage, encephalitis, or other acquired insults (Teasellet al., 2007b). Common sequelae of ABIs include well-known cognitive, emotional, and behavioral changes, though there may also be a variety of less visible consequences, including epileptogenesis. Epileptogenesis refers to the process whereby a nonepileptic brain transitions to a brain with an enduring predisposition to generate unprovoked seizures (i.e., an epileptic brain) (Jensen, 2009). Seizures are an abnormal, excessive, or synchronous neuronal activity in the brain resulting in signs or symptoms (Fisher et al., 2005). As defined by the International League Against Epilepsy (ILAE), epilepsy is defined by any of the following conditions: (a) at least two unprovoked (or reflex) seizures occurring greater than 24 hours apart; (b) one unprovoked (or reflex) seizure and a probability of further seizures similar to the general recurrence risk after two unprovoked seizures, occurring over 10 years; and (c) a diagnosis of an epilepsy syndrome (Fisher et al., 2014). Figure 11.1 demonstrates the progression of epileptogenic acquired brain injury and therapeutic opportunities.
Epilepsy is estimated to affect approximately 70 million individuals worldwide, with a lifetime incidence of up to 5 per 1,000 in developed countries and two- to threefold greater in developing countries (Ngugi, Bottomley, Kleinschmidt, Sander, & Newton, 2010). The impact of unpredictable seizures in epilepsy has a significant effect on the quality of life, which can result in serious injuries, limitations on activities including driving restrictions, and stigma, which can further limit employment and social interactions (Jacoby & Austin, 2007). Furthermore, cognitive deficits and mental health disorders are frequently seen as comorbidities with epilepsy (LaFrance, Kanner, & Hermann, 2008), even outside the setting of ABI. Current pharmacological therapies fail to control seizures in 25% to 30% of individuals (Kwan & Brodie, 2000; Mattson, Cramer, & Collins, 1996), and despite the introduction of several new antiseizure medications, including those with novel mechanisms of action, the proportion of medically intractable cases remains stable (Boon, Chauvel, Pohlmann-Eden, Otoul, & Wroe, 2002; Cross & Riney, 2009; Leppik, 2002; Matsuo & Riaz, 2009). Additionally, antiseizure medications are known to have a variety of significant adverse effects (Swann, 2001), including impairing cognitive functioning (Park & Kwon, 2008), which may be particularly problematic in the setting of ABI.
ETIOLOGY AND EPIDEMIOLOGY
PTE is a common as well as particularly debilitating long-term consequence of TBI that accounts for nearly 20% of all acquired epilepsies in the general population (Agrawal, Timothy, Pandit, & Manju, 2006; Teasell, Bayona, Lippert, Villamere, & Hellings, 2007a). ABI and TBI are recently gaining recognition as major health concerns (Annegers, Grabow, Kurland, & Laws, 1980). For example, in military populations, TBI accounts for a third of all combat-related injuries, and TBI is seen in 60% of veterans with a blast-related injury (Okie, 2005). In all, 500,000 individuals suffer from PTE in the United States and the European Union (Pitkänen, Bolkvadze, & Immonen, 2011). Epilepsy following TBI is particularly challenging as other sequelae of TBI are exacerbated by unpredictable seizures and the potential adverse effects of antiseizure medications. PTE, with acute and chronic changes resulting from TBI and epileptogenesis, and with focal and global structural pathology, is a prototype for epileptogenesis in other forms of ABI-related epilepsy in which the more extensive nature of the disease is often more difficult to appreciate.
Over the past several decades, clinical research into the various aspects of acquired epilepsy have included diagnostic methodology, risk factors, recurrence, treatment, and experimental models. A TBI is defined by the Department of Defense and the Department of Veterans Affairs (Traumatic Brain Injury Task Force, 2007) as any traumatically induced structural injury and/or physiological disruption of brain function as a result of an external force that is indicated by new onset or worsening of at least one of the following clinical signs, immediately following the event: (a) any period of loss of or a decreased level of consciousness; (b) any loss of memory for events immediately before or after the injury; (c) any alteration in mental state at the time of injury; (d) neurological deficits that may or may not be transient; (e) intracranial lesion. Recent data from nonmilitary populations (Faul, Xu, Wald, & Coronado, 2010) estimate that 1.7 million people sustain a TBI each year in the United States. Of those individuals, 52,000 die as a result of their injury, which contributes to nearly one-third of all injury-related deaths (Institute of Medicine Committee on Gulf War and Health, 2008). Another 275,000 are hospitalized, and the remaining 1.365 million are treated and released from the emergency department. Furthermore, TBI-related emergency department visits have increased 14.4% and hospitalizations have increased 19.5% from 2002 to 2006. The number of TBIs that do not receive medical care is also not known, but is estimated to be significant for mild injuries. The costs of TBI in the United States are estimated to be $60 billion annually, which includes the significant costs due to TBI-related disability. As the vast majority of those suffering from TBI survive (greater than 95%), these individuals are at risk for long-term consequences, such as PTE. And as the incidence of TBI-related emergency department visits and hospitalizations has been increasing, the mortality from TBI has been declining, likely due to preventative measures such as helmets and seat belts as well as improved treatment (Coronado et al., 2011). Therefore, more individuals survive TBI and potentially suffer long-term consequences, such as PTE.
Following a TBI, seizures can occur immediately (<24 hours), early (1–7 days), or late (>7 days). While immediate and early seizures are thought to relate to the acute severity of the TBI, late seizures suggest that epileptogenesis secondary to TBI has occurred and produced a permanent propensity for seizures. The length of time between injury and the emergence of clinical seizures is variable and may be prolonged, with 12.6% of veterans presenting with a first seizure more than 14 years after the TBI (Raymont et al., 2010). As the definition of epilepsy is a brain with an enduring predisposition to generate seizures and at least one unprovoked seizure (Fisher et al., 2014), those with at least one unprovoked, late seizure following TBI could qualify for a diagnosis of PTE. Furthermore, in those with TBI who suffer a single late seizure, 86% will have a second seizure within 2 years (Haltiner, Temkin, & Dikmen, 1997). Data from the Vietnam conflict demonstrate that over 50% of individuals with penetrating head trauma develop PTE (Raymont et al., 2010; Salazar et al., 1985). Other studies (Annegers, Hauser, Coan, & Rocca, 1998) have demonstrated the 30-year cumulative incidence of epilepsy following TBI is 2.1% for mild injuries, 4.2% for moderate injuries, and 16.7% for severe injuries. Therefore, even the least affected group of TBI demonstrates a two- or threefold increase in the risk for epilepsy. The risk of developing new onset epilepsy following a mild TBI is increased even 10 years following the injury (Christensen et al., 2009).
With these continuing risks and the debilitating features of PTE, and likely increase in prevalence in the future due to a variety of factors, there is major unmet clinical need for improved understanding about the vulnerabilities to late effects of TBI and identification of therapeutic targets to prevent PTE. Currently there are no good predictors of PTE, other than risk based on the severity of the initial injury, and no treatments exist to modify the development of PTE despite significant efforts (Beghi, 2003; Temkin, 2009; Temkin et al., 1990; Temkin et al., 1999). Furthermore, PTE is often pharmaco-resistant to available antiseizure drugs (Herman, 2002), which magnifies the impact of PTE once seizures begin.
As previously stated, epileptogenesis is the process of structural and functional changes, including the collective molecular-, cellular-, and systems-level mechanisms, that transform the normal brain to one that generates spontaneous recurrent seizures occurring following a brain insult, such as TBI, stroke, infection, and status epilepticus (Dudek & Sutula, 2007; Pitkänen & Lukasiuk, 2009; Vezzani et al., 2016). The “latentcy period” is a concept suggesting that after ABI occurs, there is a variable critical period of time from or during the time of the brain insult before the occurrence of the first spontaneous seizure (Pitkänen & Lukasiuk, 2009; Vezzani et al., 2016), which suggests a progressive series of cellular changes may be involved as well as different pathophysiological processes (Agrawal et al., 2006; Hunt, Boychuk, & Smith, 2013). The latent period may last months or years, and during this time the structure and physiology of the brain experiences epileptogenic changes that result in the development of epilepsy (Vezzani et al., 2016). Furthermore, many molecular and cellular studies have been oriented to the analysis of changes that may occur during the latent period (Dudek & Staley, 2011).
The presence of spontaneous burst discharges that develop after a mechanical brain insult seems to be a necessary minimal criterion for the human “PTE” model (Hunt et al., 2013). It is likely that other models of epilepsy and epileptogenesis, in which a slowly evolving process leads to circuit plasticity with permanent structural and functional changes and eventually to spontaneous seizures such as the kindling model (McNamara, 1986; Morimoto, Fahnestock, & Racine, 2004), share many mechanisms with PTE. Additionally, the molecular mechanisms activated in response to ABI overlap with mechanisms identified as important for epileptogenesis. For example, the neurotrophin brain-derived neurotrophic factor (BDNF) and its receptor TrkB have been demonstrated to play an important role in epileptogenesis in the kindling model (Binder, Routbort, Ryan, Yancopoulos, & McNamara, 1999; He et al., 2004; Kotloski & McNamara, 2010) and following TBI (Hu et al., 2004; Rostami et al., 2014). BDNF has been demonstrated to activate the JAK-STAT pathway, which regulates transcription of GABA-A receptor subunits (Lund et al., 2008), and the JAK-STAT pathway is also activated following TBI (Raible, Frey, & Brooks-Kayal, 2014; Raible, Frey, Cruz Del Angel, Russek, & Brooks-Kayal, 2012). Tau, an axonal protein with a microtubule-binding domain in its phosphorylated forms, has been demonstrated to aggregate into intracellular tangles following brain trauma (Corsellis, Bruton, & Freeman-Browne, 1973; McKee et al., 2009), as well as in other chronic neurologic diseases including Alzheimer’s disease (Grundke-Iqbal et al., 1986). Tau has also been linked to hyperexcitability (Holth et al., 2013), and genetic removal of tau decreases seizures in a mouse model of Alzheimer’s disease (Roberson et al., 2011), perhaps due to loss of GABAergic interneurons (Andrews-Zwilling et al., 2010; Li et al., 2009). Plaques of Aβ protein have also been shown in numerous studies of TBI (Ikonomovic et al., 2004; Roberts, Allsop, & Bruton, 1990; Roberts et al., 1994). The burden of Aβ in the brain has also been shown to increase seizure susceptibility in mouse models of Alzheimer’s disease, and the apoE4 mice develop spontaneous tonic-clonic seizures (Hunter et al., 2012). Microglial activation, as measured by glial fibrillary acidic protein (GFAP) immunoreactivity, increases following TBI (Carbonell & Grady, 1999; Regner et al., 2001; Shitaka et al., 2011), and GFAP in serum has been proposed as a biomarker of TBI in humans (Schiff, Hadker, Weiser, & Rausch, 2012). Microglial activation is also seen following seizures (Shapiro, Wang, & Ribak, 2008), and microglial activation is hypothesized to participate in cerebral inflammation that leads to hyperexcitability and seizures (Vezzani, Friedman, & Dingledine, 2013). Animal models used to study the epileptogenesis mechanisms due to infection can be complicated because most of the infectious agents that cause encephalitis in rodents are associated with high mortality (Vezzani et al., 2016). Mice infected with the Theiler’s murine encephalomyelitis virus (TMEV) respond differently. For example, SJL/J mice exhibiting mononuclear cell infiltration into the CNS with demyelination in response to the infection are used as a model of multiple sclerosis, whereas C57BL/6J mice infected with TMEV develop acute and late seizures as well as hippocampal damage (Vezzani et al., 2016). The mechanisms of early seizures during the acute symptomatic period after an infection is often multifactorial and varies depending on the type of infection (Vezzani et al., 2016). A common underlying factor in most CNS infections (i.e., meningitis and encephalitis) is the triggering of the inflammatory cascade with release of inflammatory cytokines (Vezzani et al., 2016). Furthermore, the mechanisms of epilepsy following CNS infections are not well established, but data from advances in experimental models may enhance our knowledge of the mechanisms (Vezzani et al., 2016). Individuals with brain infections may demonstrate structural damage, such as infarction, cortical necrosis, and gliosis, which may represent epileptogenic foci (Vezzani et al., 2016).
Stroke is a common ABI that may result in epileptogenesis and epilepsy. In the elderly, stroke is the most commonly identified etiology for new onset seizures (Annegers, Hauser, Lee, & Rocca, 1995; Loiseau et al., 1990; Stephen & Brodie, 2000). Even in apparently idiopathic cases, cerebrovascular disease is likely a contributor, as these individuals have elevated risk factors for stroke including hypertension, hyperlipidemia, coronary artery disease, and peripheral vascular disease (Li et al., 1997). Factors that increase the likelihood of poststroke epilepsy include stroke severity, large infarct size, and hemorrhagic transformation (Keller, Hobohm, Zeynalova, Classen, & Baum, 2015) as well as cortical involvement (Conrad et al., 2013). Retrospectively, 30% to 40% of elderly with seizures have had a stroke (W. Allen Hauser, Hesdorffer, & Epilepsy Foundation of America., 1990). Prospectively, 8.6% of those with ischemic stroke suffered a seizure within 9 months after the stroke, while 10.6% of those with a hemorrhagic stroke had a seizure within the same period (Bladin et al., 2000). In those with a poststroke seizure, the risk of a second seizure, and thereby a suggestion of an enduring predisposition to epilepsy, may be as high as 20% (Silverman, Restrepo, & Mathews, 2002). Likely secondary to the strong link between stroke and epilepsy, prescriptions for statins are associated with a decreased risk of new-onset epilepsy (Pugh et al., 2009).
Brain tumors are another common cause of new-onset epilepsy. Estimates suggest that tumors represent the etiology for 10% to 12% of epilepsy in the elderly, the population most likely to suffer from cancer (Mousali et al., 2009; Roberts, Godfrey, & Caird, 1982). Seizures can often be the presenting symptom in patients with brain tumors, whether primary or metastatic and whether intra-axial or extra-axial (Hamasaki et al., 2013). The probability that seizures will be associated with a CNS tumor depends on the tumor type, grade, and its location within the brain or, if extra-axial, its location within the cranial vault (van Breemen et al., 2007). While all types of primary and secondary tumors may present with seizures (Lee et al., 2013; Lu-Emerson & Eichler, 2012; Rajneesh & Binder, 2009), tumors of glial origin are very likely to be associated with seizures, with astrocytomas having seizure rates in the 50% to 80% range, and oligodendroglioma with rates of 46% to 78% (Chang et al., 2008; Hamasaki et al., 2013; Kahlenberg et al., 2012). These gliomas grow slowly, invading the surrounding tissue and causing gliosis and chronic inflammatory response. With their growth rate, high-grade tumors such as glioblastoma multiforme (GBM) likely effect epileptogenic changes in the peritumoral region due to mass effect and as a result of local neuronal network disruption. An incidence rate of 22% to 62% (Chaichana et al., 2011; Hamasaki et al., 2013; Kerkhof et al., 2013) is seen in individuals with GBM. Meningiomas, although common (Ostrom et al., 2013), are among the least epileptogenic intracranial tumors with a reported seizure rate between 13% and 26% (Chaichana et al., 2013; Lieu & Howng, 2000). Nearly half of those with TAE from metastatic disease did not have a diagnosis of cancer previously (Lynam et al., 2007). Overall, lower-grade tumors are more likely to present with seizures than higher-grade or metastatic tumors (Johnston & Smith, 2010), and when the inherent epileptogenicity of the tumors is combined with prevalence, gliomas, meningiomas, and metastatic tumors are the most common causes of TAE (Roberts et al., 1982; Sundaram, 1989). The mechanisms behind TAE are likely multifactorial, involving metabolic and pH changes, alterations in levels of neurotransmitters and their receptors, and disruption of localized neural networks in the region of brain tissue surrounding the tumor (Cowie & Cunningham, 2014).
In patients following central nervous system (CNS) infection, seizures are characterized into early seizures (1–2 weeks following injury), which are believed to represent acute symptomatic seizures; and late seizures (months or years following injury), which occur spontaneously and represent the clinical onset of acquired epilepsy (Vezzani et al., 2016). The mechanisms of early and late seizures are thought to be different. As previously stated, posttraumatic seizures are characterized into immediate or impact-associated (<24 hours after injury), early (1–7 days after injury), or late (> 7 days after injury) (Agrawal et al., 2006; Annegers et al., 1998; Frey, 2003; Pitkänen et al., 2014), and this classification scheme is also believed to represent different underlying pathophysiological processes (Agrawal et al., 2006). Furthermore, when studying patients after brain injury from the acute phase to 1 year, Angeleri et al. (1999) found that in all patients, early seizures, which in their protocol referred to the 4 weeks’ interval following injury, occurred within the first 10 days with a subsequent free period of >1 month. This suggested that the interval period was due to two types of seizures having different epileptogenic processes (Angeleri et al., 1999; Phillips, 1954).