Premature mortality in epilepsy: is it preventable?

Keywords::

• drowning
• epidemiology
• epilepsy
• mortality
• SUDEP

People with epilepsy are at higher risk of dying at a younger age than would be expected when compared with the general population, and this has been consistently seen in many studies [1–5]. The main measure of premature mortality is the standardized mortality ratio (SMR), which is the ratio of the observed number of deaths in the study population to the expected number of deaths in the study population if it had experienced the same age- and sex-specific mortality rates of the population from which it was drawn.

If a cohort has a SMR of 1 it suggests that mortality is similar to that of the general population, whilst a SMR of 2 would suggest that people in the cohort are twice as likely to die in any time period as their peers in the general population. Comparing SMRs in longitudinal prospective incident cohort studies gives an insight into how this risk changes over the course of the condition [6]. Proportionate mortality ratios (PMRs), although not a direct mortality measure, are also helpful as they indicate the proportion of deaths in a cohort due to a particular cause.

In a meta-analysis of mortality studies in people with epilepsy, most of the variation in reported SMRs was dependent on the source population, with institutional studies having SMRs between 1.9 and 5.1, while population-based studies having SMRs between 1.3 and 3.1 [7].

Deaths in individuals with epilepsy can be stratified by cause: deaths as a direct result of epilepsy (status epilepticus and accidents including drowning and sudden unexpected death in epilepsy [SUDEP]); deaths due to the underlying etiology of the epilepsy (brain tumors and cerebrovascular disease); and deaths from unrelated causes. The risk of premature death appears to be significantly higher in the first few years after onset [2,8–11], and decreases thereafter. Most of the initial risk (i.e., in the early years) appears to be related to the underlying etiology [2,8]. A prospective study assessed the 1-year mortality in people with a first afebrile epileptic seizure (either provoked or unprovoked) [10]. The overall SMR was 9.3 (95% CI: 7.9–10.9), with significantly increased mortality for individuals with remote symptomatic epilepsy (SMR: 6.5), provoked seizures (SMR: 10.1) and seizures due to a progressive neurological condition (SMR: 19.8). In contrast, mortality was not significantly elevated in those with cryptogenic seizures (SMR: 1.6; 95% CI: 0.4–4.1). The cause of death was attributable to the underlying cause in 64% of individuals, to unrelated conditions in 20%, to seizure-related causes in 6% and due to unknown causes or suicide in the remainder.

In a retrospective institution-based study, it was estimated that approximately a third of deaths that occur after the initial few years are epilepsy-related [1]. A meta-analysis of epilepsy cohort studies found an overall SMR of 18.7 (95% CI: 15.0–23.1) for drowning, which was even more marked in those with epilepsy and learning disability (SMR: 25.7) and people in institutional care (SMR: 96.9) [12]. This estimate of an 18-fold increased risk of drowning in people with epilepsy may be an under- or over-estimate depending on the setting. In a recent study in rural China of approximately 3500 people with newly treated epilepsy, mortality over a 4-year period was increased almost fivefold compared with that of the general population (SMR: 4.9; 95% CI: 4.0–6.1) [13]. The most common cause of death was accidental death (PMR: 59%) including drowning (PMR: 45.1%), probable SUDEP (14.7%), status epilepticus (6.9%) and neoplasms (6.9%). The risk was particularly high in young people, similar to the findings from a previous study from China where the SMRs were highest in those aged 15–29 years (SMR: 23.3–40.2) [14]. Overall, the risk of drowning was 82-times higher in the cohort than in the general population. Female gender and higher doses of phenobarbital were associated with a lower risk of premature death [13].

Status epilepticus is an important epilepsy-related cause of mortality with a PMR probably comparable to that of accidental death and drowning in some studies [4,15]. The primary determinant of outcome is the underlying etiology [16], with the majority of people with epilepsy who die as a result of status epilepticus having severe refractory epilepsy related to the underlying cause. There is some suggestion that the overall mortality from status epilepticus may be decreasing [17], possibly as a result of more aggressive earlier treatment [18], but further prospective studies are needed to verify this.

Sudden unexpected death in epilepsy is defined as the sudden, unexpected, witnessed or unwitnessed, nontraumatic and nondrowning death of a person with epilepsy with or without evidence of a seizure, excluding documented status epilepticus, and in whom post-mortem examination does not reveal a structural or toxicological cause for death [19]. SUDEP has been estimated to be responsible for up to 17% of deaths in epilepsy [20], although this percentage may be higher in children with remote symptomatic epilepsy [15]. The risk of SUDEP appears to correlate with the severity of the epilepsy, with the highest incidence rates recorded in those with epilepsy and learning disability, individuals attending tertiary epilepsy clinics, those with a long duration of epilepsy and particularly in individuals in epilepsy surgery programs [21]. In contrast, the incidence of SUDEP in community-based studies appears to be low, with only one report of suspected SUDEP in the first 10 years of a large prospective community-based cohort study [22]. The relative risk of SUDEP increases in proportion to the annual seizure frequency and the number of concomitant antiepileptic drugs (AEDs), although this may simply reflect the underlying severity of epilepsy [23].

There is evidence that SUDEP is mediated by a recent epileptic event. In a post-mortem study of 18 cases of SUDEP (and 22 controls), a marker of acute neuronal injury, heat shock protein (HSP-70), was significantly elevated in the hippocampus in cases of SUDEP compared with epilepsy and cardiac controls [24].

In a cohort of 245 children with epilepsy prospectively followed for a median of 40 years, 24% (60) died during follow-up; 55% (33) of deaths were related to epilepsy. In total, 30% of deaths were attributed to SUDEP, 10% to drowning and 7% to status epilepticus. In total, 85% (51) of deaths occurred in the 107 people not in 5-year terminal remission (15.9 deaths per 1000 person-years) compared with only 6.7% (4) in the 103 people in 5-year terminal remission not taking AEDs (1.5 per 1000 persons-years; p < 0.001). Individuals with epilepsy due to a remote symptomatic etiology had a higher rate of deaths (11.1 deaths per 1000 person-years) compared with those with idiopathic or cryptogenic epilepsy (3.2 per 1000 person-years; p < 0.001). Absence of terminal remission, remote symptomatic etiology and a prior history of status epilepticus were predictive of premature death on univariate analysis but only absence of terminal remission was predictive of death (and SUDEP) on multivariate analysis [15].

Standardized mortality ratios have been shown to decrease consistently after the initial years in population-based incident cohort studies, often to a rate comparable with that of general population. However, there is a suggestion that the mortality rate increases again in some studies after 10 years [4,5,11] or 20 years [2], although not all studies report this [3]. Most of this late increase seems to be confined to people with symptomatic epilepsy, although interestingly, the SMR for people with idiopathic/cryptogenic epilepsy increased significantly in a large prospective cohort study in the last 10 years of follow-up [25]. An increased rate of mortality in individuals with idiopathic/cryptogenic epilepsy was seen in a long-term study [2] but not in others [3,5]. In another study, the overall SMR after 25 years for people with epilepsy was 2.2 (95% CI: 2.0–2.5) with few epilepsy-related deaths [25]. The main causes of death in this study were pneumonia, cerebrovascular disease and neoplasms in the first 15 years, while in the last 5 years of follow-up, ischemic heart disease and pneumonia predominated. The overall SMR remained persistently elevated despite the fact that over 80% of people in the cohort were in terminal remission [26].

Whether or not the risk of premature mortality is preventable ultimately depends on what is driving this higher mortality rate. If, as has been suggested by some (but not all [11]) of the childhood epilepsy studies [15], ongoing seizures (and therefore epilepsy-related causes) are the primary cause of death, then rendering people with severe epilepsy seizure-free either by surgery [27,28] or by medication in people with remote symptomatic epilepsy [29] will reduce the mortality. Indeed, it has been suggested in one study that individuals rendered seizure-free do not have an increased mortality rate [30]. One recent study suggested that people taking higher doses of an AEDs appear to have a decreased risk of premature mortality compared with people who were on lower doses [13]. Although this needs appropriate testing, it may provide some circumstantial evidence that effective treatment reduces the premature mortality risk in epilepsy.

Education of individuals with epilepsy regarding the risk of drowning, the importance of AED compliance (AED noncompliance may be associated with increased mortality [31]) and a discussion on the possibility of SUDEP [32] may further modify the risk. If, on the other hand, the long-term mortality is primarily due to non-epilepsy-related causes, as suggested by some studies [20,25], then modifying the long-term mortality risk may not be possible until a better understanding of the pathophysiology of mortality in epilepsy is obtained. Highlighting to healthcare professionals that people with a history of epilepsy are at higher risk of dying from unrelated causes such as pneumonia may help modify this risk.

Acknowledgement

The authors are grateful to Dr GS Bell for her helpful comments and review of the manuscript.

Financial & competing interests disclosure

Josemir W Sander has received research grants, honoraria and travel support from various pharmaceutical companies that are involved in the manufacturing of antiepileptic drugs including Union Chimique Belge (UCB), Eisai, Janssen-Cilag and GlaxoSmithKline. Aidan Neligan has received travel grants and honoraria from Eisai and Janssen-Cilag. This work was undertaken at UCLH/UCL, which received a proportion of funding from the Department of Health’s NIHR Biomedical Research Centres funding scheme. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.