The Elephant in the Room: Suicide in Patients With Epilepsy

Commentary

Identifying Clinical Correlates for Suicide Among Epilepsy Patients in South Korea: A Case–Control Study.

Park S-J, Lee HB, Ahn MH, Park S, Choi EJ, Lee H-J, Ryu HU, Kang J-K, Hong JP. Epilepsia 2015;56:1966–1972. 10.1111/epi.1322626530473.

OBJECTIVE: Suicide is a major cause of premature mortality in patients with epilepsy. We aimed to identify the clinical correlates of suicide in these patients. METHODS: We conducted a matched, case–control study based on a clinical case registry of epilepsy patients (n = 35,638) treated between January 1994 and December 2011 at an academic tertiary medical center in Seoul, Korea. Each epilepsy patient in the suicide group (n = 74) was matched with three epilepsy patients in the nonsuicide group (n = 222) by age, gender, and approximate time at first treatment. The clinical characteristics of the patients in both groups were then compared. RESULTS: In a univariate analysis, seizure frequency during the year before suicide, use of antiepileptic drug polytherapy, lack of aura before seizure, diagnosis of temporal lobe epilepsy, use of levetiracetam, psychiatric comorbidity, and use of antidepressants were all significantly higher in the suicide group than in the nonsuicide group. Multivariate analysis revealed that a high seizure frequency (odds ratio [OR] 3.3, 95% confidence interval [CI] 1.04–10.2), a lack of aura before seizure (OR 4.0, 95% CI 1.7–9.3), temporal lobe epilepsy (OR 3.7, 95% CI 1.6–8.6), and use of levetiracetam (OR 7.6, 95% CI 1.1–53.7) and antidepressants (OR 7.2, 95% CI 1.5–34.1) were all associated with a higher probability of suicide. SIGNIFICANCE: Patients with temporal lobe epilepsy who experience seizures weekly or more frequently, experience a lack of aura, use levetiracetam, or take antidepressants are all at a higher risk of suicide and should be monitored closely.

This large Korean study of patients treated at a tertiary epilepsy center adds to the international studies on suicide in patients with epilepsy. It confirms that completed suicide is more likely in temporal lobe epilepsy (TLE) patients who have increased seizure frequency and are treated with antidepressants and multiple antiepileptic drugs (AEDs). Park et al. also found that treatment with levetiracetam was associated with an increased likelihood of suicide. Although this drug can trigger depression in epilepsy patients (1), the few patients treated with levetiracetam in this study underscore the need to replicate this finding.

The results of this study clearly remind clinicians of the importance of screening TLE patients for psychiatric comorbidity, particularly depression, before selecting AED regimens. This message is particularly relevant for AEDs that can trigger depression, such as vigabatrin, topiramate, and phenobarbital (see review in Mula and Sander [2]). Use of the reliable, valid, free, easily administered and scored Neurological Disorder Depression Inventory for Epilepsy (NDDIE) (3) should help clinicians achieve this goal, and prevent this highly morbid outcome in some TLE patients.

But the risk for suicide is not only associated with depression. Anxiety disorders, depression and anxiety disorder, bipolar disorder, and psychosis—psychiatric disorders prevalent in epilepsy patients (4)—are associated with suicidal behavior (thoughts, plans, attempts, and successful acts) (5). The availability of screening instruments for anxiety disorders (the 7-item Generalized Anxiety Disorder [GAD-7] self-report scale [6]) and for depression and anxiety disorders (the Hospital Anxiety and Depression Scale [HADS] [7]) can help identify patients with these comorbidities who then need to be assessed for suicidality. Furthermore, the increased risk for AED adverse effects in epilepsy patients with clinical and subclinical depression and anxiety (8, 9) can contribute to a downward emotional spiral with worsening of their psychiatric symptoms and quality of life. This, in turn, can exacerbate a sense of hopelessness and suicidal behavior. The role of substance misuse to self-treat depression, anxiety, and bipolar disorder in patients both with (10) and without epilepsy (11) in this cascade of events should not be underestimated and needs to be evaluated in youth and adults with epilepsy.

Should clinicians screen for suicidal behavior in all epilepsy patients before prescribing AEDs? Clearly, this should be part of the assessment of patients who screen positively for psychiatric comorbidity. In the absence of prospective randomized controlled AED drug trials, with baseline assessment of suicidality and psychiatric comorbidities, this question remains to be answered. Given the FDA 2008 warning, however, such screening should be part of clinical practice (12). Suicidal ideation is an important predictor of subsequent acts (13) and is an indication for psychiatric referral. The NDDIE and nonepilepsy-specific depression instruments such as the Beck Depression Inventory and the Patient Health Questionnaire (PHQ-9) include items on suicidal thoughts. Alternatively, clinicians can simply ask patients if they have had suicidal thoughts recently or in the past.

What other take-home messages do we learn from the study by Park et al.? Knowledge of risk factors would help clinicians identify those patients at highest risk for suicide. However, the study's wide age range, from 18 to 75 years, and lack of information on the age distribution of the patients precluded conclusions about age-related risk factors. This is particularly important because adolescents and young adults have different risk factors for suicide than geriatric patients. In adolescence and young adulthood the risk factors include past suicide attempt, suicidal ideation, recent romance breakup, family history of suicide and of depression, learning problems, and substance abuse (14, 15). For geriatric patients, medical comorbidities, psychiatric disorders (mainly depression), functional impairment, and stressful life events (see Conwell and Thompson [16]) increase vulnerability for suicide. Age-related factors among the risk factors for suicide need to be examined in large representative community samples of young and old epilepsy patients for them to have generalizable clinical implications.

In their cohort-controlled population study of suicide in individuals with epilepsy and in the general population, Kwon et al. (17) found no increased prevalence of suicide in epilepsy patients when controlling for psychiatric and medical comorbidities. The inclusion of community samples of epilepsy patients is essential to determine if, in fact, the risk for suicide is specifically increased in epilepsy above and beyond the effects of having these comorbidities. As mentioned above, medical comorbidities are particularly relevant in the suicide risk of geriatric patients. In fact, the window for suicide in the 3 years before and 1 year after diagnosis (whether of epilepsy or of a psychiatric disorder) might simply reflect the lack of treatment for the underlying psychiatric conditions given the two-way relationship between epilepsy and psychiatric disorders (18). More specifically, patients with psychiatric disorders are at higher risk for epilepsy and for suicide. Similarly, patients with epilepsy are at increased risk for psychiatric disorders and suicide. The lack of treatment or inadequate treatment for psychiatric disorders might be the common underlying variable in this finding.

Yet, despite accumulating evidence for suicide and the psychiatric comorbidity in patients with epilepsy and the availability of relatively easy to administer screening instruments for psychiatric comorbidity, the mental health needs of these patients remain unmet. In terms of suicide, the elephant in the room might be that clinicians are reticent to evaluate suicidal behavior. There may be several reasons for their reluctance: a lack of training in how to evaluate suicidal behavior, a concern about potential liability incurred if one identifies suicidality in a patient, the shortage of mental health professionals willing to treat epilepsy patients, and the limited psychiatric insurance coverage for the psychiatric comorbidity. How to remove the elephant from the room in order to prevent this significant epilepsy morbidity should be an important quality-of-care goal of the epilepsy community. An integrated research, clinical, and policy approach will be needed to achieve this goal.