Clinical manifestations

Presentation and course

Depressive disorders in patients with epilepsy may be identical to primary depressive disorders. Nonetheless, in approximately 30% to 50% of patients with epilepsy, depressive episodes can present with atypical clinical manifestations and fail to meet the diagnostic criteria included in the Diagnostic and Statistical Manual of Mental Disorders-IV (DSM-IV). They consist of symptoms of depression intermixed with brief euphoric moods, irritability, anxiety, paranoid feelings, and somatoform symptoms (anergia, atypical pain, and insomnia). They tend to follow a chronic course with recurrent symptom-free periods.

More often than not, depressive and anxiety disorders occur together. For example, in a study of 188 consecutive patients with epilepsy, 31 met DSM-IV criteria of a current major depressive episode. Twenty-one of these patients had a mixed major depressive episode and an anxiety disorder (51). Recognition of such comorbidity is of the essence as failure to include it in the treatment plan can result in recurrence of the depressive disorder. In addition, comorbid depressive and anxiety disorders in patients with and without epilepsy are associated with a worse course of the depressive disorder and a worse quality of life compared to that of patients with only a depressive disorder.

Often, depressive disorders can be associated with personality disorders. In such cases, the response to treatment is poor and the overall prognosis is worse. Of note, personality disorders are rarely recognized in the evaluation of patients with epilepsy, and their prevalence is yet to be investigated in population-based studies. Most of the data available are derived from case series from tertiary centers (59).

Clinical manifestations particular to epilepsy. The occurrence of psychiatric symptoms in patients with epilepsy can be temporally related to that of seizures, either preceding it (preictal), following it (postictal), as an expression of the seizure (ictal), and independent of the of the seizure (interictal). In general, interictal depressive disorders are the ones more frequently identified whereas peri-ictal expressions often go unrecognized.

Preictal symptoms of depression. Preictal symptoms of depression typically present as a dysphoric mood in which the symptoms may extend for hours or even 1 to 3 days prior to the onset of a seizure. For example, Blanchet and Frommer assessed mood changes over a period of 56 days in 27 patients with epilepsy who rated their mood on a daily basis (09). Mood ratings pointed to a dysphoric state 3 days prior to a seizure in 22 patients (81%). This change in mood was greatest during the 24 hours preceding the seizure. Patients or parents of children with epilepsy often report that dysphoric symptoms completely resolve the day after the ictus.

Postictal symptoms of depression. Postictal symptoms of depression have been recognized for decades but have been investigated in only a few studies. The largest study that investigated the prevalence of postictal psychiatric symptoms in a systematic manner included 100 consecutive patients with treatment-resistant focal epilepsy (55). Postictal symptoms of depression were identified in 43 of 100 consecutive patients with refractory partial seizure disorders. These symptoms occurred after more than 50% of seizures during the previous 3 months and their duration ranged from 0.5 to 108 hours, with a median duration of 24 hours. These symptoms could occur within 72 hours of the last seizure, and typically a symptom-free period of 12 to 24 hours between the seizure and the onset of the psychiatric symptoms was recorded. Furthermore, postictal suicidal ideation was identified in 13% of patients, and the median duration of these symptoms was also 24 hours. Other studies have suggested that postictal symptoms of depression can outlast the ictus for up to 2 weeks and, at times, have led patients to suicide (Anatassopoulos and Kokkini 1969; 35).

Of note, patients with epilepsy may experience interictal depressive episodes with postictal exacerbation in severity of these symptoms (55). Furthermore, in patients with interictal depressive episodes in remission, postictal symptoms of depression can occur despite the presence of adequate doses of antidepressant medication (Kanner unpublished data). This observation suggests a different pathogenic mechanism operant in postictal and interictal depressive symptoms.

Ictal symptoms of depression. Ictal symptoms of depression are the clinical expression of a simple partial seizure in which the depressive symptoms are the sole (or predominant) semiology. Ictal symptoms of depression ranked second after symptoms of anxiety or fear as the most common type of ictal affect in one study (85). This presentation occurred in 21% of 100 patients with epilepsy who reported auras consisting of psychiatric symptoms (Weil 1955; 19). Yet, the actual prevalence of ictal symptoms of depression is yet to be established in larger studies. The most frequent symptoms include feelings of anhedonia, guilt, and suicidal ideation. Such mood changes are typically brief, stereotypical, occur out of context, and are associated with other ictal phenomena. More typically, however, ictal symptoms of depression are followed by an alteration of consciousness as the ictus evolves from a simple to a complex partial seizure.

Depressive episodes as an iatrogenic effect.

Caused by antiepileptic drugs. All antiepileptic drugs can potentially cause psychiatric symptoms, particularly when used at high doses (66). However, antiepileptic drugs with GABAergic properties, primarily phenobarbital, primidone, the benzodiazepines, tiagabine, and vigabatrin are more likely to cause depression (27; 12; Smith et al 1987; 04; 79). Other antiepileptic drugs that have been associated with the development of symptoms of depression include felbamate, topiramate, levetiracetam, zonisamide, and, among the third-generation antiepileptic drugs, perampanel (67; 54; 70). The effects of depression using lacosamide, a third-generation antiepileptic drug, has been evaluated in several studies. Moseley and colleagues examined the NDDI-E and GAD-7 scores prior and following lacosamide treatment and showed no significant effects in the GAD-7 scores and significant improvements in the NDDI-E scores, suggesting that lacosamide is safe to use in patients with mood disorders (68).

Psychiatric adverse events are more likely to be identified in patients susceptible to develop psychiatric disorders, such as those with a past psychiatric history or with a family psychiatric history (70). Accordingly, obtaining personal lifetime and family psychiatric histories is of the essence in identifying patients who may be at risk of developing iatrogenic depressive episodes.

The addition of antiepileptic drugs with mood stabilizing properties, such as carbamazepine, valproic acid, and lamotrigine can very rarely cause depressive episodes. On the other hand, these antiepileptic drugs are associated with the occurrence of depression on their discontinuation in patients with a prior history of depressive disorders that had been kept in remission by these antiepileptic drugs (58).

In 2008, the Food and Drug Administration instructed all manufacturers of antiepileptic drugs to add a warning to their labeling indicating that use of these drugs increases risk for suicidal thoughts and behavior (05). The announcement was based on an FDA meta-analysis of 199 clinical multicenter-randomized, double-blind placebo controlled and parallel trials of 11 antiepileptic drugs. The FDA concluded that patients randomized to antiepileptic drugs had twice the risk for suicidal ideation and behavior than those randomized to placebo (0.43% vs. 0.24%). The FDA data were carefully reviewed by a task force of the American Epilepsy Society, which identified several methodological flaws including: (1) grouping all antiepileptic drugs (which have different mechanisms of action) into one class, and (2) basing the recommendation on spontaneously reported adverse events rather than on systematically acquired data (40). Furthermore, attempts to replicate the data from the FDA meta-analysis in 5 large studies yielded contradictory findings. Thus, certain but not all antiepileptic drugs can cause psychiatric adverse events that can lead to suicidal ideation and behavior, but this is likely to occur in patients with a predisposition for psychiatric illness. Accordingly, before starting an antiepileptic drug like topiramate, zonisamide, perampanel, vigabatrin, barbiturates, benzodiazepines, or levetiracetam, it is essential to inquire about a prior personal or family psychiatric history, both of which are clear risk factors for the development of depressive disorders.

Depression following epilepsy surgery. There have been an increasing number of reports of depressive disorders following an anterotemporal lobectomy during the first 3 to 6 months after surgical resection (50). Furthermore, it is not unusual to see “mood lability” within the initial 6 weeks after surgery. Often these depressive disorders subside by 6 to 12 months after surgery. The severity of postsurgical depressive disorders varies from mild to very severe, including suicidal attempts. In most instances, these depressive disorders respond readily to pharmacologic treatment with antidepressant drugs. Patients with a prior history of depression are at greater risk. A German study found that patients with personality disorders are at higher risk of suffering from postoperative psychiatric complications, including psychotic episodes, as compared with patients with other preoperative psychiatric conditions (such as depression) or with patients with no preoperative psychiatric diagnosis whatsoever (59). Although some studies have not found a relationship between depression with the postsurgical control of seizures, others have (Pintor et al 2007; 50). All patients undergoing epilepsy surgery, therefore, should be advised of this potential complication prior to surgery.

Prognosis and complications

Impact on mortality risk. A population-based study from Sweden found that mortality is 11-fold higher in patients with epilepsy than controls (24). Close to 16% of all deaths were attributed to external causes, which included death by suicide and accidents (in a motor vehicle, drug poisoning, falls, drowning, assault, other). In fact, patients with epilepsy had a greater than 3-fold higher risk of premature death from these causes than the control groups. Furthermore, among patients with epilepsy who died from external causes, comorbid psychiatric disorders were identified in 75.2%, the majority of which included depression and substance misuse. It is worth noting that the lifetime prevalence of psychiatric disorders in patients with epilepsy was 40.7%, compared to 10.3% in controls. In a separate population-based study conducted in Denmark, the presence of mood disorders increased the risk of suicide by 32-fold (14). In a review of 21 studies, suicide accounted for 12% of all deaths in patients with epilepsy (44).

Impact on the management of the seizure disorder.

Worse tolerance of antiepileptic drugs. Two studies published in 2011 showed that patients with epilepsy with major depressive episodes endorsed more severe adverse events of antiepileptic drugs (48). The study by Kanner and colleagues demonstrated that the worse tolerance was not limited to major depressive episode but was of comparable magnitude in patients with less severe forms of depression, such as sub-syndromic depressive episodes. Tolerance was worse among patients with epilepsy who experienced comorbid major depressive episodes and anxiety disorders (48).

Worse response to the pharmacologic and surgical treatment of the seizure disorder. Three studies have suggested a worse response of the seizure disorder to pharmacotherapy with antiepileptic drugs. Likewise, a lifetime history of psychiatric disorders and, in particular, depression has been associated with an increased likelihood of persistence of auras or even focal seizures with loss of consciousness following an anterotemporal lobectomy. For example, in a study of 100 consecutive patients, only 12% of patients who reached complete seizure-freedom after surgery (mean follow-up period 8.3±3.3 years) had a lifetime history of depression, whereas such history was identified in 67% of patients who had auras but no complex partial or secondarily generalized tonic-clonic seizures and in 79% of patients with persistent disabling seizures (53). These data were confirmed in 2 other studies that included patients with mesial temporal sclerosis: in one study of 280 patients, those with a preoperative psychiatric diagnosis (38% of the entire cohort) were significantly less likely to remain seizure-free (OR 0.53, 95% CI 0.28-0.98, p = 0.04) (15), whereas in the second study of 115 patients, a pre-surgical history of major depressive disorders was a risk factor for persistent postsurgical seizures (OR=5.23; p=0.003) (20). Other studies have questioned the findings of these studies (01); hence, there is still equipoise with respect to this question. In a study from a large German epilepsy center, persistence of seizures following an anterotemporal lobectomy was identified in patients with epilepsy and a history of depression associated with a personality disorder (60).

Impact on the quality of life. Several studies of patients with treatment-resistant epilepsy demonstrated that depressive disorders are the most powerful predictor of poor quality of life, even after controlling for seizure frequency, severity, and other psychosocial variables (73; 32; 63; 17; 51). In one of these studies, depression was significantly associated with poor quality-of-life independently of seizure type, but seizure-freedom for the previous 3 months improved the quality-of-life ratings (17). Of note, comorbid depressive and anxiety disorders have a worse impact on the quality of life of patients with epilepsy than the depressive or anxiety disorders occurring alone and particularly when a major depressive episode is comorbid with more than one anxiety disorder (51).

Impact on health care costs. Patients with untreated depression were found to use significantly more health resources of all types, independent of seizure type or duration (18). Mild-to-moderate depression was associated with a 2-fold increase in medical visits compared with nondepressed controls whereas severe depression was associated with a 4-fold increase. The presence and severity of depression was a predictor of lower disability scores, irrespective of the duration of the seizure disorder. Also, depressive disorders have been found to worsen the tolerance of antiepileptic drugs.

Clinical vignette

A 30-year-old female with history of focal epilepsy presented with new symptoms of depression. The patient started having focal seizures with loss of consciousness at the age of 20 years. Previous workup revealed a focal epilepsy of left temporal lobe origin of undetermined cause. She was seizure-free on lamotrigine monotherapy for the past 10 years. She was diagnosed with tuberculosis 2 months prior to presentation, requiring treatment with isoniazid, rifampin, pyrazinamide, and ethambutol. Her past medical history was significant for a previous episode of depression at the age of 18 years. Her family history was significant for depressive disorders and anxiety disorders in her mother and sister.

She had been seizure-free for 2 years, but 1 month prior to presentation she had 4 focal seizures with loss of consciousness, all within a 48-hour period for which she was taken to the nearest emergency room where levetiracetam was added to her antiepileptic regimen. In the 2 weeks prior to presentation, she started having symptoms of decreased energy, anhedonia, constant crying, irritability, and overall sadness. She also reported feeling life is not worth living anymore although denied an active suicidal ideation.

This case illustrates iatrogenic depressive episodes caused by 2 possible effects: (1) the introduction of levetiracetam in a patient who is known to be at risk of iatrogenic effects because of a past history of depressive disorder and positive family history of mood and anxiety disorder and (2) the pharmacokinetic interaction between lamotrigine and rifampin, which resulted in increased clearance of lamotrigine and, in-turn, may have unmasked a depressive disorder that was kept in remission by the antidepressant and mood-stabilizing properties of this antiepileptic drug.

The patient had a routine EEG that showed slowing and interictal epileptiform activity over the left temporal region but was basically was unchanged when compared to previous studies. A lamotrigine level of 4.5 mcg/mL was lower than a level (8 mcg/ml) obtained one year earlier. The dose was increased to achieve previous serum concentration, levels were monitored while she was on rifampin therapy, and levetiracetam was discontinued. Her depression as well as seizures remitted.

Biological basis

Etiology and pathogenesis

The causes of depressive disorders in patients with epilepsy are multiple and more often than not are the result of the interaction of several variables that include: (1) a recurrence of a mood disorder that may have preceded the onset of epilepsy; (2) a genetic predisposition for mood disorders; (3) changed in neurotransmitters, hormonal and inflammatory abnormalities associated with the seizure disorder, which are similar to those of primary mood disorders; (4) iatrogenic factors; and (5) adjustment to the obstacles associated with a life with seizures (eg, loss of independence, stigma, need to change work, etc.).

In the past, depressive disorders were considered a “complication” of epilepsy. Yet, data from population-based studies have revealed a more complex relationship--a bidirectional relation between the two disorders, whereby not only are people with epilepsy at greater risk of developing depressive disorders, but people with primary depressive disorders are at greater risk of developing epilepsy. For example, in one study, Hesdorffer and colleagues used the United Kingdom’s General Practice Research Database to compare the incidence ratio of primary psychiatric disorders (eg, depression, anxiety, psychosis) during the 3 years that preceded and followed the onset of the seizure disorder between 3,773 patients with epilepsy and 14,025 controls 10 to 60 years of age (37). The incidence rate ratios of depression were 2- to 3-fold higher during the 3 years preceding and the 3 years following the diagnosis of epilepsy in the subjects who went on to develop the seizure disorder than in controls; for suicidality, the incidence rate ratio during the 3 years was 4-fold higher before and 1 year after the diagnosis of epilepsy. In a previous population-based study conducted in Sweden, investigators compared the risk of developing unprovoked seizures or epilepsy among patients who had been hospitalized for a psychiatric disorder (n = 1,885) with that of a group of controls, matched for gender and year of diagnosis selected randomly from the register of the Stockholm County population (03). The age-adjusted odds ratio for development of unprovoked seizures was 2.5 for people with major depressive disorder, 2.7 for patients with bipolar illness, 2.7 for anxiety disorder, and 2.6 for suicide attempt. Three other population-based studies published between 1990 and 2006 demonstrated that patients with a depressive disorder had a 3- to 7-fold higher risk of developing epilepsy (30; 38; 39).

The burden of comorbidity in people with epilepsy is high. Several diseases, including depression, anxiety, dementia, migraine, heart disease, peptic ulcers, and arthritis are up to 8 times more common in people with epilepsy than in the general population. Several mechanisms explain how epilepsy and comorbidities are associated, including shared risk factors and bidirectional relations (56).

The use of antidepressant drugs has been traditionally associated with an increased risk of developing seizures and, in fact, seizures are listed as a potential adverse event of all antidepressant drugs. Thus, is the bidirectional relation reported in the studies listed above the expression of an iatrogenic effect of antidepressant drugs used to treat the mood disorder? The available evidence does not appear to support this supposition. Indeed, one study conducted in patients with primary major depressive and obsessive-compulsive disorders compared the incidence of seizures in the course of Phase II and III multicenter-randomized placebo-controlled trials between those randomized to placebo or an antidepressant (02). Data for this study encompassed 75,873 patients enrolled in trials submitted to the Food and Drug Administration for regulatory purposes between 1985 and 2004. The trials included antidepressants of the tricyclic, selective serotonin-reuptake inhibitor (SSRI), and serotonin-norepinephrine reuptake inhibitor (SNRI) families as well as bupropion. The investigators found that subjects randomized to antidepressants were 52% less likely (69% less likely when excluding bupropion immediate release formulation) to develop seizures compared to those randomized to placebo. In addition, patients randomized to placebo were 19 times more likely to experience a seizure compared to the expected incidence in the general population. A higher incidence of seizures was found in patients randomized to clomipramine and bupropion (in its immediate-release formulation) than to placebo. Furthermore, antidepressant drugs of the SSRI and tricyclic families have been found to display antiepileptic effects in several animal models of epilepsy (47) whereas reduction in seizure frequency has been suggested in 4 open trials with SSRIs in patients with treatment-resistant epilepsy (25; 26; 81; 78) and in one double-blind placebo controlled trial with imipramine (Frost et al 1978).

In addition, a lifetime history of depression has been associated with a worse response to pharmacotherapy with antiepileptic drugs and epilepsy surgery. For example, in one study of 780 consecutive patients with new -onset epilepsy followed for a median 79-month period, those with a psychiatric history, in particular depressive disorders, preceding the onset of epilepsy were 2-fold more likely to develop treatment-resistant epilepsy than those without (41). In a separate study of 138 consecutive patients with new onset epilepsy, those who endorsed symptoms of depression and anxiety before the start of antiepileptic therapy were significantly less likely to be seizure-free after 12 months of therapy (Petrovski et al 2010). In a more recent population-based study, patients with a history of depression before the onset of epilepsy and who required pharmacologic treatment and psychotherapy were less likely to be seizure-free (46). By the same token, several studies have now reported that patients with intractable temporal lobe epilepsy and a lifetime psychiatric history were less likely to achieve complete seizure freedom following an anterotemporal lobectomy (53; 15).

Clearly, the data presented above suggest not only that mood disorders may be a risk factor for epilepsy, but they may also increase the likelihood of a treatment-resistant seizure disorder. The existence of common neurobiologic pathogenic mechanisms operant in depressive disorders and epilepsy may explain this phenomenon. These mechanisms include: (1) disturbances of neurotransmitters in the central nervous system, such as serotonin, norepinephrine, dopamine, glutamate and gamma-amino-butyric acid (GABA), (2) endocrine disturbances, such as hyperactive hypothalamic-pituitary-adrenal axis, resulting in high serum concentrations of cortisol, and (3) inflammatory mechanisms (48).

For example, although disturbances of secretion or receptor sensitivity of serotonin have been identified as pivotal pathogenic mechanisms of completed suicide, major depression, and anxiety disorders, decreased secretion has been identified in several animal models of epilepsy, including the genetically epilepsy-prone rat (43) and the pilocarpine status epilepticus model in the Wistar rat (65) as well as in animal models of epilepsy tested in cats, rabbits, and rhesus monkeys (43).

Glutamate is the “excitatory” neurotransmitter with a key pivotal role in the epileptogenic process. High glutamate concentrations in plasma and CSF have been identified in humans with mood disorders, whereas high concentrations in frontal and occipital cortex have been reported in studies with magnetic resonance spectroscopy (80). In addition, there are data suggesting that NMDA antagonists may exert an antidepressant effect in animal models of depression as well as in humans with treatment-resistant major depressive disorders (86).

In the case of GABA, an inhibitory neurotransmitter, which also plays a pivotal role in the epileptogenic process, low CSF, plasma, and intracortical concentrations have been found in humans with mood disorders (75).

A hyperactive hypothalamic pituitary adrenal axis leading to high secretion of cortisol was identified as one of the neurobiologic markers of major depression almost 40 years ago (16). It is typically tested with the dexamethasone suppression test, which demonstrates a failure to suppress the secretion of cortisol. Likewise, several studies have shown that pretreatment with corticosterone accelerates the kindling process in rats (62).

With respect to inflammatory disturbances, proinflammatory cytokines, in particular interleukin-1beta (IL-1beta), have been found to play a pathogenic role in patients with mood disorders. In animal models of epilepsy and humans with temporal lobe epilepsy, IL-1beta, has been shown to have pro-convulsant properties, which are blocked by its naturally occurring antagonist (IL-1RA). The pro-convulsant mechanism has been associated to a reduction in glutamate uptake by glial cells or an enhanced release of glutamate from these cells (Vezani et al 1999; Vezani et al 2000; Vezani et al 2008).

Epidemiology

Depressive disorders are the most frequent psychiatric comorbidity in patients with epilepsy, with lifetime prevalence rates harboring around 30% to 35% (82). In cross-sectional studies, patients with persistent seizures have been found to have higher prevalence rates (21% to 33%) than seizure-free patients (4 to 6%) (21; 42; 71). Furthermore, a population-based survey of 185,000 households, which screened for lifetime symptoms of depression, yielded higher rates in patients with epilepsy than those with diabetes or asthma and healthy subjects (22); indeed, at least 1 symptom of depression was reported by 32% of patients with epilepsy, 13% of patients with diabetes, 16% with asthma, and 8.6% of healthy respondents. Similar prevalence rates of depression have been identified in children and adolescents with epilepsy (13).

Prevention

Despite their frequent occurrence, depressive disorders remain under-recognized and under-treated. For example, among 97 patients with epilepsy with depressive episodes severe enough to warrant the use of psychotropic drugs, more than 1 year lapsed before 63% of patients with spontaneous and 54% with iatrogenic episodes were referred for treatment (54). Other investigators identified a current untreated major depressive disorder in 43% of patients with epilepsy, but 38% of patients with lifetime histories of major depressive disorder had never been referred for treatment (84). This problem is not restricted to adults; for example, Ettinger and colleagues found depressive symptomatology in 26% of 44 children with epilepsy who had never been previously evaluated or treated (23). There needs to be a constant effort to increase awareness about the recognition and treatment of these psychiatric comorbidities in patients with epilepsy.

Although rapid identification and treatment of primary depressive disorders is of the essence, anticipation of iatrogenic depressive disorders is also important in patients with epilepsy. Indeed, as stated above, antiepileptic drugs with negative psychotropic properties (eg, barbiturates, topiramate, benzodiazepines, zonisamide, vigabatrin, levetiracetam) can result in depressive symptoms, including major depressive episodes in patients with epilepsy at risk of developing psychiatric disorders (eg, patients with a prior psychiatric history or family history). By the same token, discontinuation of antiepileptic drugs with mood-stabilizing properties in patients with epilepsy with a prior history of depressive disorders may unmask such disorders. Furthermore, the addition of antiepileptic drugs with enzyme-inducing properties may lower the serum concentrations of concomitant antidepressant drugs and limit their efficacy, causing a recurrence of psychiatric symptoms. Finally, clinicians must be on the lookout for postsurgical depressive disorders occurring in 20% to 30% of patients with epilepsy during the first 3 to 6 months following an anterotemporal lobectomy. This risk is increased in patients with a family or personal prior history of depressive disorders. Accordingly, presurgical psychiatric evaluations are of the essence to identify those patients with an increased risk.

Differential diagnosis

When considering a diagnosis of depressive disorder, it is of the essence to establish the nature of the psychiatric symptomatology as the treatment differs significantly. Clinicians must consider whether the symptoms at hand are the expression of: (1) an interictal depressive episode, (2) a peri-ictal depressive episode, (3) an iatrogenic episode, or (4) an organic process mimicking a mood disorder. The first three, were discussed extensively in the previous sections.

Among the organic processes that can be also associated with epilepsy and can present as mood disorders, clinicians must consider the possibilities of an endocrinopathy affecting the thyroid, pituitary, and suprarenal glands as well as neurologic disorders secondary to brain tumors involving primarily frontal and temporal lobe structures and degenerative disorders of the brain, including dementias.

Diagnostic workup

Depressive disorders are not difficult to diagnose; furthermore, self-report screening instruments have become available to identify symptoms of depression and major depressive episode--its most severe clinical expression--in patients with epilepsy. These instruments can be given to patients to fill-out in the clinic while waiting to be seen by their doctor, and they are probably the best solution to facilitate their identification in patients with epilepsy in a busy outpatient clinic. These instruments include the Beck Depression Inventory (BDI-II), the Center of Epidemiologic Studies-Depression (CES-D), which have been validated for patients with epilepsy (45), and the Neurological Disorders Depression Inventory in Epilepsy, which is a 6-item instrument that was validated to screen for major depressive episode in patients with epilepsy specifically (33). The latter has the advantage of having been constructed specifically to minimize confounding symptoms that plague other instruments, such as somatic symptoms that may also be the expression of adverse events of antiepileptic drugs or of cognitive problems associated with epilepsy. Patients can fill-out the instrument in less than 3 minutes. A score of higher than 15 is suggestive of a major depressive episode. These instruments are not diagnostic by themselves, and a subsequent, more detailed evaluation is necessary. Once the diagnosis of a mood disorder has been established by psychiatric evaluation, the self-rating screening instruments can be given at every visit to document symptom remission.

As already discussed, comorbid anxiety disorders or symptoms of anxiety are a common occurrence, and their identification and effective treatment is of the essence. The 7-item Patient’s Health Questionnaire Generalized Anxiety Disorder-7 (GAD-7) is an ideal self-rating instrument to screen for generalized anxiety disorder (61). It takes 2 to 3 minutes for patients to complete; a score of higher than 10 is suggestive of a generalized anxiety disorder.

Management

To date, antidepressants of the selective-serotonin-reuptake inhibitor (SSRI) and serotonin-norepinephrine reuptake inhibitors (SNRI) type have become the first-line of treatment of primary depressive disorders, replacing the first generation of tricyclic antidepressants because of their lower toxicity. Unfortunately, to date there are no published double-blind placebo controlled trials that have tested the efficacy of these two families of antidepressants in patients with epilepsy. In their absence, two consensus statements have been published outlining pharmacological and psychotherapeutic approaches for patients with epilepsy with affective disorders using data from the management of primary mood disorders and from open trials with SSRIs in patients with epilepsy (06; 57). Both documents recommend the use of these antidepressant drugs in patients with epilepsy and emphasize the safety of these drugs when used at therapeutic doses. Four antidepressant drugs should be avoided in patients with epilepsy, however: chlorimipramine, bupropion, amoxapine, and maprotiline; these drugs have been found to be associated with increased risk of seizures.

Before starting pharmacotherapy, it is important to determine if the depressive episodes resulted from an iatrogenic process, which may result from:

	• The introduction of an antiepileptic drug with negative psychotropic properties (eg, barbiturates, topiramate, benzodiazepines, zonisamide, vigabatrin, levetiracetam), particularly in patients with epilepsy at risk of developing psychiatric disorders (eg, patients with a prior psychiatric history or family history).
	• The discontinuation of an antiepileptic drug with mood-stabilizing properties (eg, carbamazepine, oxcarbazepine, valproic acid, lamotrigine) in patients with a personal (or family) history of mood disorder that was in remission or masked by one of these antiepileptic drugs.
	• The addition of an antiepileptic drug with enzyme-inducing properties (carbamazepine, phenytoin, barbiturates) in patients with epilepsy taking an SSRI, which had resulted in remission of the depressive episode. In these patients, these antiepileptic drugs will increase the clearance of the antidepressant, lower its serum concentration, and limit its efficacy.

Another cautionary note is in order: before starting an antidepressant drug, clinicians must rule out a history of a manic or hypomanic episode that may be suggestive of a bipolar disorder as these drugs can potentially trigger a manic or hypomanic episode in the short term, but they may worsen the course of the bipolar disorder in the long-term, particularly in the case of rapid cycling bipolar disease. In such cases, an antiepileptic drug with mood stabilizing and antidepressant properties such as lamotrigine must be considered. Carbamazepine and valproate may be added in case of persistent symptoms. Lithium should be considered if these antiepileptic drugs cannot yield a euthymic state. If an antidepressant is required, it should not be started in the absence of a mood-stabilizing drug.

Do antidepressant drugs worsen seizures? For a long time, clinicians have been reluctant to prescribe antidepressant drugs to patients with epilepsy, because of a concern that this type of drugs lowers the seizure threshold. Yet, this appears to be a misunderstanding that needs to be clarified (49). First, reports of seizures in patients taking antidepressant drugs have included patients taking overdoses. Second, based on the evidence suggesting a bidirectional relationship between depression and epilepsy that suggests a higher incidence of seizures in patients with depression, it is important to reconsider the question of whether antidepressants increase the risk of seizures or whether we are identifying the seizure occurrence associated with such increased risk. The study by Alper and colleagues cited above helped to clarify this point as it showed that compared to the expected incidence of seizures of the general population, all depressed patients had a higher incidence, but patients randomized to placebo had a significantly higher incidence of seizures than those treated with the actual antidepressant drug (02).

A review of the literature has found an increased incidence of seizures with the use of 4 specific antidepressant drugs; these include: maprotiline, amoxapine, clomipramine, and bupropion (Swinkels et al 1995; 66). As stated above, SSRI and SNRIs are expected to be safe in patients with epilepsy when used at therapeutic doses (02). In one study, sertraline was found to definitely worsen seizures in only 1 in 100 patients with refractory epilepsy (54) whereas several open trials have suggested a potential antiepileptic effect of fluoxetine and citalopram (25; 26; 81). In a published study by Ribot and colleagues, treatment of depressive or anxiety disorders with an SSRI or SNRI was associated with a reduction in seizure frequency by almost 30% among patients with more than 1 seizure per month (78). This effect was not related to the psychiatric effect of these drugs.

Blumer has also reported that using tricyclic antidepressants alone and in combination with SSRIs in patients with epilepsy did not lead to any seizure exacerbation (11). Monoamine oxidase-A-inhibitors (MAO-I) are not known to cause seizures in patients without epilepsy. Patients treated with tricyclic antidepressants should be started at low doses with small increments until the desired clinical response is reached to minimize the risk of adverse events. Furthermore, the use of this class of antidepressant drugs requires a baseline EKG as well as one when reaching the target dose, as changes have been reported. The latest review of the use of SSRIs in patients with epilepsy concludes that antidepressant drugs of the SSRI and SNRI families are safe for these patients and should be used at therapeutic doses for the management of depressive and anxiety disorders (49).

Pharmacokinetic interaction between antidepressant drugs and antiepileptic drugs. Most antidepressant drugs are metabolized in the liver, and their metabolism is accelerated in the presence of antiepileptic drugs with enzyme-inducing properties. This pharmacokinetic effect is not observed with the new antiepileptic drugs gabapentin, lamotrigine, tiagabine, levetiracetam, zonisamide, pregabalin, vigabatrin, and perampanel.

Conversely, some of the SSRIs are inhibitors of one or more isoenzymes of the cytochrome P450 (CYP 450) system. These antidepressants include fluoxetine, paroxetine, fluvoxamine, and to a lesser degree, sertraline (72; 31; 34). Escitalopram and citalopram, on the other hand, do not have pharmacokinetic interactions with antiepileptic drugs (66). Sertraline has been shown rarely to increase phenytoin levels, and this is thought to be associated with displacement by tight protein binding, or by inhibition of the CYP 450 system (36).

Choice of antidepressant. The criteria used to choose an antidepressant must be based on the following:

	• Type of depressive episode: retarded versus agitated. In the former, SNRIs may be more effective, given their noradrenergic effect. In the latter, SSRIs should be considered first.
	• Presence of comorbid anxiety disorders and their type. In such cases, the clinician must identify the SSRI or SSRI that can cover the symptoms of both conditions. Failure to treat comorbid anxiety disorders will result in increased risk of recurrence of the depressive disorder or failure to reach symptom remission.
	• Adverse event profile of the antidepressant drug, in particular those drugs with a propensity to cause weight gain.

Dosing of the antidepressant must be the same as in the treatment of primary mood disorders. The aim is to achieve complete symptom remission as persistence of symptoms is associated with an increased risk of recurrence of a major depressive episode. Furthermore, clinicians must adjust the dose in the presence of enzyme-inducing antiepileptic drugs by increasing it by 30%.

Other types of psychiatric treatments. Lithium was the first “mood stabilizing drug” used for the treatment of patients with primary bipolar disorder. Its use has been fraught with several problems, including changes in EEG recordings and pro-convulsant effects at therapeutic serum concentrations in patients without epilepsy (07). Lithium’s neurotoxicity and related increase in seizure risk increases with the concurrent use of neuroleptic drugs, in the presence of EEG abnormalities, and with a history of CNS disorder.

Electroconvulsive therapy is not contraindicated in depressed patients with epilepsy (Sackeim et al 1983; 77; 28). It is a well-tolerated treatment and is worth considering in patients with epilepsy with very severe depression that fails to respond to antidepressant drugs. Furthermore, there is no evidence that electroconvulsive therapy increases the risk of epilepsy (08), and in fact, it has been used in the treatment of refractory status epilepticus.

Vagal nerve stimulation is used as adjunctive treatment for epilepsy and has also shown longer-term mood effects (64). The impact of other neurostimulation interventions, such as deep brain stimulation or transcranial magnetic stimulation on mood in patients with epilepsy, has not been studied in controlled trials.

In addition to pharmacological intervention, the value of psychotherapy for the treatment of depression in patients with epilepsy should not be overlooked. Surveys reveal that fear of the next seizure is rated as the greatest concern in patients with epilepsy (29). Counseling and psychotherapy, in particular cognitive behavior therapy, can be useful in helping the patient deal with the stressors and limitations of living with epilepsy (76).

More recently, the randomized, controlled crossover study UPLIFT demonstrated the efficiency of a mindfulness-based cognitive therapy intervention for preventing major depressive disorder episodes in patients with epilepsy (83).

Special considerations

Pregnancy

Patients with epilepsy may have a de-novo or recurrence of previous depressive disorders during the postpartum period.

Anesthesia

Anesthetics do not affect seizures or depressive disorders in patients with epilepsy. It should be emphasized that the medications should be taken as prescribed prior to, during, and after the surgical procedures and the use of anesthetics.