An official website of the United States government
NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.
Noebels JL, Avoli M, Rogawski MA, et al., editors. Jasper's Basic Mechanisms of the Epilepsies. 5th edition. New York: Oxford University Press; 2024. doi: 10.1093/med/9780197549469.003.0038
Currently, there are no established biomarkers which predict the risk of epilepsy following an epileptogenic brain insult, or which predict outcomes of pharmacological or surgical interventions following epilepsy diagnosis. Epilepsy appears to share bidirectional mechanistic relationships with its associated behavioral comorbidities such as neuropsychiatric and cognitive impairments, with several pieces of evidence suggesting that behavioral abnormalities exist before the initial diagnosis of epilepsy. Here, we discuss the preclinical and clinical evidence of these relationships, and the potential for behavioral and cognitive disorders to be harnessed as predictive biomarkers for risk of epilepsy or treatment outcomes.
Identifying biomarkers for epileptogenesis and epilepsy severity is a major goal in epilepsy research, and it is an unmet clinical need recognized by the International League Against Epilepsy and the National Institutes of Health (Galanopoulou et al., 2013; Simonato et al., 2014). Such biomarkers will enable the identification of individuals who could benefit from the preventive therapeutic strategies after an epileptogenic brain insult such as a traumatic brain injury (TBI). Similarly, a biomarker for predicting treatment outcomes at the initiation of therapy or surgical resection of epileptogenic tissue in treatment-resistant patients would be highly beneficial. Despite the several candidates being identified, including the molecular, electroencephalographic, and imaging biomarkers, their clinical use has not been achieved, with potential explanations for this discussed in Simonato et al. (2021). These include, but are not limited to, the fragmented and poorly coordinated efforts and study designs with inappropriate statistical approaches, issues with intermodel reproducibility, and difficulties with validation in patients (Simonato et al., 2021).
Epilepsy is commonly associated with comorbidities such as cognitive and psychosocial dysfunction that impart a major burden on the quality of a patient’s life (Galanopoulou et al., 2013; Pohlmann-Eden et al., 2015; Semple et al., 2019). There is strong evidence indicating a mechanistic bidirectional relationship between epilepsy and cognitive and neuropsychiatric comorbidities (Catena-Dell’Osso et al., 2013; Hermann et al., 2017), and there is an increasing view in the field that these comorbidities may represent an integral part of the epileptic condition (Brooks-Kayal et al., 2013). The mechanisms may include common neurobiological changes in anatomical structures such as the hippocampus and alterations in the expression of neurotransmitter receptors that are relevant to both the epilepsy and the cognitive and neuropsychiatric comorbidities (Mazarati and Sankar, 2016). However, there is evidence, as described later in this chapter, that the neurobehavioral and cognitive abnormalities may already be existent before the diagnosis of epilepsy. Similarly, the neurobehavioral abnormalities may be the psychological consequence of having epilepsy and/or the neurobiological effects of the antiseizure medications (ASMs) (El Sabaa et al., 2020; Loring et al., 2007). This may make the identification of etiological mechanisms highly challenging. Nevertheless, the presence of neurobehavioral and cognitive abnormalities before the epilepsy diagnosis has implications for understanding the mechanisms and initiation of early treatments for these disorders, but it could also be utilized to predict epilepsy development in at-risk patients who have experienced a potential epileptogenic brain insult. In this chapter, we discuss the common neurobiological mechanisms that may be related to the bidirectional relationship between the epilepsy and its associated neurocognitive and neuropsychiatric comorbidities. We also discuss evidence for existence of such comorbidities before the diagnosis of epilepsy and at the time of initiation of epilepsy therapy. We explore whether neurocognitive and neuropsychiatric measures could be a predictive biomarker for epileptogenesis or therapeutic response to the ASMs, or to the resective brain surgery to treat patients with drug-resistant focal epilepsy.
A comorbidity occurs during the course of an index disease, such as epilepsy. Comorbidities are generally defined in broad terms, including clinical diseases and syndromes and signs or symptoms that are distant from the index disease (Feinstein, 1970; Keezer et al., 2016). The most common notion is that epilepsy is at least partly the cause of behavioral comorbidities and if the epilepsy is resolved, the behavioral comorbidities will do likewise (Helmstaedter and Witt, 2017). However, it is important to note that a comorbidity, by definition, neither excludes nor implies a causal relationship with the index disease (i.e., epilepsy) (Helmstaedter and Witt, 2017; Valderas et al., 2009). In this section, we will review the different lines of research that challenge this epilepsy-centric notion, highlighting the evidence that shows that in some patients the neurobehavioral comorbidities precede the development of epilepsy and may even determine the outcome of epilepsy.
Several large population-based studies have reported different medical conditions that are up to eight times more prevalent in patients with epilepsy compared to the general population (Gaitatzis et al., 2012; LaFrance et al., 2008; Téllez-Zenteno et al., 2005; Tellez-Zenteno et al., 2007). However, recent evidence has also shown that patients with psychiatric disorders are at higher risk of developing epilepsy (Kanner et al., 2014). The most common comorbidities seen in patients with epilepsy are cognitive deficits (70%–80% of patients) and mood disorders (up to 60% of patients).
The bidirectional relationship between epilepsy and mood disorders was highlighted in population-based studies showing that the incidence rate ratios and prevalence ratios for depression are significantly elevated both before and after an epilepsy diagnosis (Josephson et al., 2017). This evidence raised the prospect that depression may be established before the onset of epilepsy, which could even be viewed as a biomarker for developing epilepsy (Hesdorffer et al., 2006, 2007). A study led by Forsgren and Nystrom found that depression preceded the diagnosis of epilepsy and was seven times more common than among age- and sex-matched controls. Interestingly, when the analysis was restricted to those individuals with focal-onset epilepsy syndromes, depression was 17 times more common than matched controls (Forsgren and Nyström, 1990). An epidemiological investigation performed in Minnesota described that the diagnosis of depression occurred 3.7 times more frequently among adults with epilepsy than among controls. It is worth pointing out that an episode of major depression tends to occur closer to the time of the first seizure in comparison to controls (Hesdorffer et al., 2000). In an epidemiological study performed in the Icelandic population, major depression and attempted suicide independently increased the risk for unprovoked seizure even after correcting for sex, age, alcohol intake, or the number of major depression symptoms (Hesdorffer et al., 2006). Another study using the Health Improvement Network database of the United Kingdom found that the severity of depression appears to augment the risk of developing epilepsy and the odds of worse seizure outcomes (Josephson et al., 2017).
In regard to other mood disorders, a UK population-based study revealed that a history of depressive and anxiety disorders, suicidality, and psychosis were identified 2- to 4-fold more frequently during the 3 years before the onset and diagnosis of epilepsy in patients compared to that of controls (Hesdorffer et al., 2012). The findings were further supported by a Swedish epidemiological study, in which investigators evaluated the risk of developing unprovoked seizures and epilepsy among patients who had been hospitalized for a psychiatric disorder (Adelöw et al., 2012). Furthermore, the age-adjusted odds ratio for developing unprovoked seizures was significantly higher for all psychiatric disorders, including major depression, suicide attempt, bipolar disorder, and anxiety disorders. Interestingly, the odds ratio to develop genetic generalized epilepsy among depressed patients was higher than for focal epilepsies.
Different lines of research have shown the close and bidirectional relations among epilepsy, stress, and anxiety disorders (Becker et al., 2015; Hingray et al., 2019; Zijlmans et al., 2017). Anxiety disorders have been described to play an important role in the vulnerability to develop epilepsy (Zijlmans et al., 2017), as the diagnosis of the first seizure has been associated with a stressful life event (Kanner, 2011; Zijlmans et al., 2017). Evidence has shown that individual factors (such as early life stress or previous diagnosis of anxiety) and environmental factors (such as a stressful event) contribute to developing epilepsy (Becker et al., 2015; Kumar et al., 2011). Similarly, anxiety may trigger a seizure, while conversely, seizures can trigger anxiety symptoms (Hingray et al., 2019). Anxiety disorders are mainly seen in temporal lobe epilepsy (TLE) and are associated with abnormal emotional processing with a bias toward negatively valence emotion (Lanteaume et al., 2012). In regards to the genetic epilepsies, studies have shown that juvenile myoclonic epilepsy has a strong association with anxiety disorders (Filho et al., 2008).
The prevalence of psychoses in people with epilepsy has been estimated to be 2%–8% (Toone et al., 1982). Psychosis has been described in ictal, interictal, and postictal periods, as well as postepilepsy surgery and ASM treatment (Krishnamoorthy et al., 2002; Toone et al., 1982). Patients with psychotic episodes treated with antipsychotic drugs, clozapine, olanzapine, or quetiapine, have also a higher incidence of seizures in comparison to placebo-treated patients (Alper et al., 2007). Furthermore, patients with epilepsy are at greater risk of developing a psychotic disorder, and conversely patients with a primary psychotic disorder are also at greater risk of developing epilepsy (Kanner and Rivas-Grajales, 2016). A retrospective study found that psychoses are more commonly seen in people suffering mesial-TLE (Filho et al., 2008), and it is often associated with more severe structural pathology than in TLE patients without psychosis (Gutierrez-Galve et al., 2012; Tebartz Van Elst et al., 2002).
Likewise, a Taiwan-based population study described a bidirectional relationship between schizophrenia and epilepsy. The researchers found that the incidence of epilepsy was higher in the schizophrenia cohort in contrast to controls; and that the incidence of schizophrenia was higher in the epilepsy cohort than controls. Moreover, the effect of schizophrenia to develop epilepsy was more significant in women, but the association between epilepsy and incidence of schizophrenia was more pronounced in men (Chang et al., 2011). Previous diagnosis of attention-deficit/hyperactivity disorder (ADHD) was 2.5-fold more common among children with newly diagnosed seizures than among control subjects. However, this association was restricted in ADHD predominantly inattentive type and was not affected by the type of seizures, etiology, sex, or seizure burden at diagnosis (Hesdorffer et al., 2004). In another population-based case-control study, children with migraine with aura had a 4-fold increased risk of developing epilepsy, but not those who developed migraine without aura (Ludvigsson et al., 2006).
Cognitive deficits are determined by multiple factors, including the presence of structural brain lesions, epilepsy burden, treatment, and individual memory reserve capacities (Helmstaedter, 1999). Prior to the epilepsy onset, congenital, developmental, or acquired brain lesion and behavioral and psychiatric disorders can have a negative effect on cognition (Witt and Helmstaedter, 2015). Accordingly, it is expected that cognitive problems may already be present before epilepsy onset (Taylor et al., 2010; Witt and Helmstaedter, 2012). However, depending on the precipitating factor that causes the cognitive dysfunction, these deficits can be transient (Helmstaedter and Witt, 2017). Studies have demonstrated that cognitive impairments are present in newly diagnosed, untreated new-onset epilepsies (Taylor et al., 2010; Witt and Helmstaedter, 2015). Furthermore, performance on psychometric tests on attention, executive functions, and memory to assess the different dimensions of cognition have shown impairment in recently diagnosed adults with untreated epilepsy (Pulliainen et al., 2000; Witt and Helmstaedter, 2015). Similarly, in children with new-onset genetic generalized epilepsy, evidence has shown that academic and behavioral problems may precede the first recognized seizure (Austin et al., 2001; Jones et al., 2007), and often the cognitive impairments are described since the beginning of the disease (Hermann et al., 2012). On the other hand, as assessed by IQ tests, sound cognitive abilities have been widely demonstrated as an indicator of treatment success and a greater capacity to cope with any transient or more permanent disease (Chelune et al., 1998; Lieb et al., 1982). Furthermore, cognition can be negatively affected due to the presence of epileptic activity, both convulsive and nonconvulsive (Aldenkamp and Arends, 2004), and interictal events such as epileptiform discharges (Binnie et al., 1987).
The epilepsy-centric view as the cause of behavioral comorbidities is no longer valid and may actually prevent the search for underlying etiological factors and the proper identification and treatment of behavioral comorbidities. Comprehensive preclinical investigations are warranted to answer neurobiological questions that can dissect the bidirectionality of epilepsy and its associated comorbidities. Therefore, a practical clinical approach favors screening, evaluation, and treatment of behavioral comorbidities at the onset of the disease.
In conclusion, the evidence suggests that the bidirectionality between comorbidities and epilepsy does exist since behavioral comorbidities can affect the development and prognosis of epilepsy, and epilepsy can modify behavioral outcomes. Current evidence raises the possibility that epilepsy and the associated comorbidities may mark different disease stages of a common underlying condition (Helmstaedter et al., 2014). Overlapping neuroanatomical regions, such as the temporal, orbitofrontal, and inferior prefrontal areas, and altered expression and function of specific neurotransmitters have been identified as potential substrates for depression and epilepsy, which will be reviewed in the next section.
Given the highly prevalent neurocognitive and neuropsychiatric comorbidities associated with epilepsy, as described in the previous sections, it is perhaps likely that biological perturbations in the brain caused by the epileptic condition contribute to the onset and severity of these comorbidities. However, given the evidence of bidirectionality, such that psychiatric disorders can predate the onset of epilepsy (Hesdorffer et al., 2012) and that neurocognitive functioning at epilepsy onset can predict antiseizure treatment response (Petrovski et al., 2010), it is also possible that common pathobiological mechanisms exist which explain the co-occurrence of these diseases. The purpose of this section is to highlight these potential mechanisms which may be shared between seizures/epilepsy and behavioral comorbidities, including cognitive and psychiatric disorders. As the remit of our chapter is to discuss the use of behavioral biomarkers in epilepsy, the description of common causal mechanisms will be relatively brief.
When considering neurocognitive comorbidities, structural or functional disruptions to neural circuitry and connectivity may represent a causal mechanism (Lenck-Santini and Scott, 2015). Our primary understanding of the cause of all epilepsy syndromes involves some form of alteration to neuronal circuit or network connectivity, either structurally, functionally, or both, rendering the brain or brain regions in a state which is vulnerable to experience bouts of excessively active or synchronous activity—seizures. Clearly, structural changes to epileptogenic regions, such as the temporal lobe in case of hippocampal sclerosis, may be expected to influence cognitive processes requiring these regions. Also, nonpathological changes in network activity may also lead to neurocognitive disturbances relevant in epilepsy. For example, neural synchrony, measured using electrophysiology, appears to be aberrant in many instances of brain disorders characterized by cognitive symptoms (Uhlhaas and Singer, 2006), including Alzheimer disease (AD), autism, and schizophrenia, as well as epilepsy. With regard to epilepsy, apart from the obvious evidence of aberrant neural synchrony being associated with seizures, interictal activity can also exhibit abnormal synchrony, such as in the case of interictal spikes and pathological high-frequency oscillations often present in the seizure-onset zone (Zijlmans et al., 2012). The use of functional magnetic resonance imaging (fMRI) can also be utilized to gain measures of network connectivity, both at rest and during seizure (Centeno and Carmichael, 2014), and using this technology has also identified abnormalities in network engagement in epilepsy patients when conducting a word generation paradigm (Vlooswijk et al., 2011). Such examples provide clear evidence of network abnormalities associated with cognitive disturbances in epilepsy patients.
Another potential shared mechanism linking cognitive disturbances in epilepsy may revolve around assessment of molecular pathologies, such as those associated with AD, including tau and beta-amyloid, a disease characterized by cognitive dysfunction. There is increased prevalence of seizures in AD compared with age-matched controls, and this evidence, coupled with an abundance of literature regarding animal modeling (Chan et al., 2015; Minkeviciene et al., 2009), suggests that tau and beta-amyloid species can lead to excessive excitotoxicity and seizures (Palop and Mucke, 2010). Notably, although usually linked to dementia, hyperphosphorylated tau has been identified in surgically resected tissue from patients with temporal lobe, which correlates to cognitive decline (Gourmaud et al., 2020). Indeed, in a modified kainic acid-induced SE (KASE) in which animals developed chronic drug-resistant TLE, animals that received an anti-tau and protein phosphatase 2A enhancer showed improved cognitive performance and decreased neuromotor deficits. Moreover, the reduction of behavioural deficits correlated to a reduction in spontaneous recurrent seizures (Casillas-Espinosa et al., 2023). Furthermore, the improvement of behavioral comorbidities and seizures correlated with a reduction of phosphorylated tau, oxidative stress and modification of other relevant molecular pathways (Casillas-Espinosa et al., 2023). Interestingly, an intervention study with a T-type calcium channel modulator given immediately after terminating the kainic-acid-induced SE (KASE) showed that the drug-treated animals did not develop depressive-like behaviour or cognitive deficits and that such animals also did not develop TLE (Casillas-Espinosa et al., 2019). The findings from these intervention pre-clinical studies highlight the bidirectional relationship between the development of behavioural comorbidities and epilepsy, early during the epileptogenesis process but also chronically in drug-resistant epilepsy.
Regarding affective comorbidities, such as depression and anxiety, several good reasons exist why causative biological mechanisms may overlap with that of epilepsy, particularly TLE. First, these disorders are strongly linked to stress, either in early life (Heim and Nemeroff, 2001) or adulthood (McEwen, 2004), and the temporal lobe regions are key areas in the brain that mediate the response to stress, and glucocorticoid stress receptors are highly expressed in these regions. Also, stress hormones are recognized to influence neuronal excitability, and many patients report that stress is a primary trigger of their seizures (McKee and Privitera, 2017). Finally, as described above, bidirectional relationships exist between these comorbidities. With regards to biological mechanisms, in addition to anatomical linkage, abnormalities to HPA axis function, monoamine systems including serotonin and noradrenaline, as well as disturbances to the balance of GABA and glutamate neurotransmission, may also be very relevant as potential mechanisms explaining the co-occurrence of psychiatric conditions and epilepsy. These topics and the mechanistic implications are covered comprehensively in Chapter 61, this volume.
Given the common prevalence of neurobehavioral comorbidities of epilepsy, and that such comorbidities may share similar mechanisms and common anatomical substrates, it is not surprising that such comorbidities are present even before the diagnosis of epilepsy, as will be discussed in this section. Such evidence may provide possibilities of using these as a biomarker for epileptogenesis or enrich the other biomarker panels to improve into a possible combinatory biomarker. This is indeed something relatively easier to establish in animal models of epilepsy due to practical ease of such measurements. However, relatively few studies have investigated this question (Table 38–1). One pioneering study in this regard used a modification of the lithium-pilocarpine model of TLE where status epilepticus (SE) was induced in rats and then terminated after either 90 minutes or 60 minutes. The authors knew from their earlier report that the animals with the 60 minutes of SE do not develop epileptic seizures. Seizure threshold was tested by timed infusion of pentylenetetrazole, and behavioral assessments were performed prospectively during the first few weeks after SE induction and the epilepsy outcomes were confirmed several months later following the vEEG monitoring (Bröer and Löscher, 2015). This study showed that the best intergroup discrimination was provided by combining all measured parameters (receiver operating characteristic [ROC] area under curve [AUC] of 0.96; p = 0.01), which accurately predicted the development of epilepsy in animals. The behavioral test evaluated in this study was a pick-up test measuring the hyperexcitability and aggressive behavior in rats that independently had a predictive value at week 1 after SE induction (ROC AUC of 0.87; p = 0.02) but was not effective in the next 2 weeks of measurements (Bröer and Löscher, 2015).
Summarizing the Literature Relating to the Potential of Using Behavioural Biomarkers for Epileptogenesis and Outcomes of Pharmacological and Surgical Interventions.
A follow-up study (2016) used a similar approach as that of Broer and Loscher (2015), where they limited the duration of self-sustained SE induced by electrical stimulation of amygdala to 2.5 hours instead of 4, leading to development of epilepsy in 50% of the rats. This study showed that behavioral hyperexcitability could predict the development of epilepsy, whereas contrary to the findings by Broer and Loscher, seizure threshold was ineffective as a biomarker for epilepsy development. The finding of no predictive value of seizure threshold suggests that the biomarker evaluations may be model-specific and that multiple/combinatorial biomarkers must be pursued for effective translation to the clinic.
Pascente et al. (2016) used a different strategy of epilepsy modeling to allow only a proportion of animals to become epileptic with the aim to evaluate potential biomarkers for epileptogenesis. Pascente et al. (2016) induced pilocarpine-induced SE in 3-week-old rats where 60%–70% animals develop epilepsy. They evaluated the learning behavior on the Morris water-maze test and the neuroinflammation measured by magnetic resonance spectroscopy for their potential to discriminate epileptic animals (Pascente et al., 2016). Although all SE rats displayed learning deficits, only the rats that became epileptic displayed accelerated forgetfulness (d-15, ROC AUC = 0.79, p = 0.01 and d-65, ROC AUC = 0.84, p = 0.009). In addition, myo-inositol, a marker of astrocyte activation, increased persistently in animals that developed epilepsy when compared to the animals that did not. Accordingly, a combination of myo-inositol levels at day 72 and behavioral outcomes increased the biomarker sensitivity and specificity (ROC AUC to 0.9 with p = 0.002) (Pascente et al., 2016).
Given these intriguing results in the animal models of TLE, it is surprising that no further studies have been reported or conducted in the preclinical setting for further validation and translation to the clinical settings. A recent study by Nizinska et al. (2021) evaluated whether behavioral abnormalities would be able to stratify the animals with short versus long latency to the first seizures or the animals with high or low seizure numbers in the self-sustained electrical status model of epilepsy. The authors show that a higher motor activity or different swimming strategies (low proportion of scanning versus incursion) at week 26 post SE was associated with high number of seizures in the animals (Nizinska et al., 2021). However, the experimental design in this study is not appropriate to evaluate the potential of behavioral measurements as biomarkers for epileptogenesis, since the behavioral tasks were completed at the late chronic stage of epilepsy where rats already displayed epileptic seizures. The behavioral outcomes in this case may have been influenced by the prior presence of epilepsy. Indeed, such behavioral abnormalities are also commonly observed in patients with posttraumatic epilepsy who display personality disorders such as mood swings, irritability, disinhibited behavior, socialization problems, and depressed mood when compared to the TBI patients without posttraumatic epilepsy (Bushnik et al., 2012; Haltiner et al., 1996; Mazzini et al., 2003; Raymont et al., 2010). However, most of these clinical studies with posttraumatic epilepsy (PTE) patients lacked any behavioral assessment before epilepsy diagnosis with the exception of the study from Asikainen et al. (1999). Although the neurological deficits were not significantly different between the patients with PTE and without PTE, the authors proposed that such defecits still constituted a risk factor for posttraumatic epilepsy (odds ratio of 1.74) (Asikainen et al., 1999). Another preclinical study utilized a clinically translatable model of posttraumatic epilepsy induced by fluid percussion injury that imparts a proportion of animals to become epileptic, allowing for evaluation of predictive biomarkers for posttraumatic epileptogenesis (Shultz et al., 2013). In this study, behavioral evaluations for anxiety-like behavior, depression-like behavior, and learning and memory impairments were conducted 1, 3, and 6 months post injury along with the measurement of volumetric changes and metabolic functions measured by magnetic resonance imaging (MRI) and 18F-FDG positron emission tomography (PET), respectively. Although, some PET and MRI imaging-based abnormalities were related to the animal becoming epileptic, none of the behavioral outcomes were useful in this context, either alone or in combination with the other evaluated parameters (Shultz et al., 2013).
Behavioral comorbidities have also been related to the development of genetic epilepsies. Genetic absence epilepsy rats from Strasburg (GAERS) display neurobehavioral comorbidities such as depression-like behavior evidenced by lower preference for sucrose in the sucrose preference test and lower percentage of entries in the open arm of the Elevated Plus maze and inner area of the Open Field test (Jones et al., 2008). These behavioral abnormalities were present at the 7-week time point, which is before the emergence of spike-wave discharges that typically are observed in these rats at around an age of 10–12 weeks (Jones et al., 2008).
Apart from adult-onset epilepsy, there are several reports of behavioral abnormalities in children at the time or even prior to the diagnosis of epilepsy (Hermann et al., 2012). Austin et al. (2001) reported that behavioral problems such as attention deficits and thought and somatic complaints were observed in children 6 months before the first recognized seizure, with the rate being the highest in children who had previous events that were perceived to be seizures. Similarly, a study by Jones et al. (2007) showed that children recently diagnosed with epilepsy displayed an elevated rate of lifetime-to-date Diagnostic and Statistical Manual of Mental Disorders, fourth edition (DSM-IV) Axis I disorder, including depressive disorders, anxiety disorders, and attention-deficit/hyperactivity disorder. The authors report that around 45% of the children exhibited these disorders before the first recognized seizures (Jones et al., 2007). Moreover, a history of major depression is reported to be 1.7-fold and a history of suicide attempt being 5.1-fold more common in patients displaying unprovoked seizures than the controls in an Icelandic population with children and adults (Hesdorffer et al., 2006). Finally, there are several reports of the need of special academic and teaching services in up to 25% of the school going children before the first diagnosis of epilepsy (Berg et al., 2011, 2005). These findings suggest the prevalence of psychiatric comorbidities antedating epilepsy onset and may represent underlying neurobiological influences independent of seizures, epilepsy syndrome, and medication treatment. Despite such findings, a detailed clinical investigation for a potential epileptogenesis biomarker has not yet been attempted. Performing such an extensive prospective clinical study would be challenging, but it is highly warranted, considering the preclinical and clinical literature relating to the presence of behavior impairments before the epilepsy onset.
In addition to the potential of being a predictive biomarker for the epileptogenesis, presence of behavioral comorbidities at the time of diagnosis could also predict the severity of epilepsy or the responsiveness to the antiepileptic treatment (Table 38–1). In this context, a previous study had demonstrated that prior or current psychiatric comorbidity such as depression was associated with the refractory epilepsy (odds ratio 2.35, 95% confidence interval 1.48–3.73); the other risk factors reported in this study were high number of seizures, family history of epilepsy, febrile seizures, or TBI as a cause of epilepsy (Hitiris et al., 2007). In a prospective study specifically designed to determine the role of neuropsychiatric comorbidities in epilepsy pharmacoresistance, the A-B Neuropsychological Assessment Scale (ABNAS) questionnaire was completed by the patients newly commencing the AED treatments (Petrovski et al., 2010). The ABNAS scale has been shown to correlate with the scale of memory, anxiety, and depression (Brooks et al., 2001). Petrovski et al. (2010) showed that treatment-nonresponsive patients had higher pretreatment ABNAS scores than patients whose seizures were controlled by the AEDs (n = 93) (p = 0.007). The neuropsychiatric comorbidities along with the MRI lesion displayed a high predictive value for the AED nonresponders with a ROC AUC of 0.75 (p = 0.03) (Petrovski et al., 2010). Moreover, the presence of neuropsychiatric comorbidities is also associated with more severe perceived AED-related adverse events (Gómez-Arias et al., 2012; Kanner et al., 2012) which would affect the adherence and compliance to the prescribed AEDs and complicate the evaluation of treatment responsiveness.
Presence of neuropsychiatric and cognitive abnormalities may also have the potential to predict outcomes following the surgical resection of epileptogenic brain tissue (Table 38–1). In a study that followed-up children 2 years after the temporal lobe resections, it was observed that the children with previous cognitive impairments were more likely not to be seizure-free compared to the ones that had no cognitive impairments (Danielsson et al., 2002). Similar findings have also been reported in the adult epilepsy patients following temporal lobectomy. A retrospective study of 100 epilepsy patients who underwent anterior temporal lobectomy demonstrated that a lifetime history of mood and psychiatric disorders was predictive of worst postsurgical seizure outcome at 2 years follow-up (Kanner et al., 2009). This study showed that patients without any history of neuropsychiatric disorders at time of surgery had a 16.6-fold chance (CI: 1.4–688) of having complete seizure freedom (Kanner et al., 2009). Another study enrolling 280 patients with TLE surgery reported an odds ratio of 0.53 (95% CI: 0.28–0.98, p = 0.04) for the patients to remain seizure-free in the follow-up period if they had a history of neuropsychiatric disorders before the epilepsy surgery (Cleary et al., 2012). Similarly, another study has reported an odds ratio of 5.23 (p = 0.003) for nonfavorable seizure outcomes after a cortico-amygdalohippocampectomy if the patients had a history of major depressive disorder at the time of surgery (de Araújo Filho et al., 2012). Despite a number of clinical reports, predictive value of behavioral impairments on pharmacoresistance has not been tested in animal models with appropriate study design using recommended statistical tools such as ROC curve to determine the sensitivity and specificity of such predictions. There is evidence that rats with resistance to the phenobarbital displayed severe behavioral and cognitive changes compared to the rats that responded well to the AED treatment (Gastens et al., 2008). However, whether the abnormal behavior could predict treatment responsiveness is not determined yet in the preclinical settings.
Evidence indicates that neurobehavioral comorbidities can exist before the onset of seizures in many epilepsy cases. This evidence is derived from both clinical studies, as well as in purpose-designed, statistically powered preclinical studies and raises the potential of behavior comorbidities being a predictive biomarker for epilepsy. This also clearly points to the shared etiological mechanisms between epilepsy and the neurobehavioral comorbidities that not only lead to a bidirectional relationship between the two but also suggest that therapeutic strategies may also be interrelated. The literature discussed above clearly suggests that neuropsychiatric and cognitive dysfunction at the time of AED initiation or before the time of epilepsy surgery as a major risk factor for the worst outcomes with regard to seizure control with medical and surgical treatment. Such findings provide a strong rationale for conducting neuropsychiatric assessments before the treatment initiation and epilepsy surgery. This would enhance recognition of psychiatric comorbidities in epilepsy, provide detailed insights of the diversity of phenotypes in a shared anatomical and functional network, provide the chance for an early intervention, and in the context of this chapter, provide an early predictor for unfavorable treatment or surgical outcome.
This is an open access publication, available online and distributed under the terms of a Creative Commons Attribution-Non Commercial-No Derivatives 4.0 International licence (CC BY-NC-ND 4.0), a copy of which is available at https://creativecommons.org/licenses/by-nc-nd/4.0/. Subject to this license, all rights are reserved.
Your browsing activity is empty.
Activity recording is turned off.
See more...
