Sudden Unexpected Death in Epilepsy (SUDEP): A Review of Risk Factors and Possible Interventions in Children

Sudden unexpected death in epilepsy (SUDEP) is a rare, but serious outcome in patients with epilepsy. The unexpected nature of SUDEP, especially in children and adolescents, has significant impacts on friends, family, and caregivers. This review will summarize the current literature and provide practical recommendations for health care providers as they care for pediatric patients with epilepsy.

Definition. SUDEP is defined as a “sudden, unexpected, witnessed or unwitnessed, nontraumatic and nondrowning death in patients with epilepsy, with or without evidence for a seizure and excluding documented status epilepticus, in which postmortem examination does not reveal a toxicologic or anatomic cause of death.”¹  SUDEP can be further classified on the basis of whether the patient was successfully resuscitated after the event, whether an autopsy confirmed diagnosis, and whether comorbid conditions contributed to death (Table 1).¹  However, these classifications can be difficult to apply to individual cases, leading to challenges when estimating incidence of SUDEP.² 

Incidence. Owing to the rarity of SUDEP cases, varying methodologic approaches, and heterogeneous patient populations, the reported incidence rate of SUDEP in children varies from 0.22 to 1.11 per 1000 person-years.^3–9  SUDEP occurs less commonly in children than adults, with the incidence in adults ranging from 0.35 to 6 per 1000 person-years.³ Table 2 summarizes studies reporting incidence rates of SUDEP in various pediatric cohorts.

Etiology. The etiology of SUDEP is poorly understood but may involve dysregulation of the cardiac, respiratory, and autonomic nervous systems (Figure). Structural and functional changes within the central nervous system may be involved, including brain volume loss in the brainstem and temporal lobe.^10,11  Studies from epilepsy monitoring units (EMUs) are particularly helpful in elucidating possible etiologies for SUDEP because of the difficulty of capturing these events. The MORTEMUS study by Ryvlin et al¹²  compiled retrospective data from 147 EMUs resulting in 16 SUDEP cases and 9 near-SUDEP cases in patients 10 to 62 years of age. Through video electroencephalogram (VEEG) monitoring, all of the SUDEP cases experienced a generalized tonic-clonic seizure (GTCS) prior to cardiorespiratory arrest. In 10 of these cases, patients developed rapid breathing in the immediate postictal phase followed by generalized VEEG suppression and the development of bradycardia, apnea, and asystole.¹²  This study suggests that many SUDEP cases begin with a seizure event followed by cardiorespiratory compromise. Pursuing this hypothesis, Singh et al¹³  used VEEG, pulse-oximetry, electrocardiography, and respiratory inductance plethysmography data from an EMU to study risk factors leading to cardiopulmonary abnormalities following seizures in 26 patients aged 2 to 20 years. This study identified that ictal cardiopulmonary abnormalities, including apnea and bradycardia, are common in children with epilepsy. Ictal apnea and bradycardia were associated with male sex, localization-related epilepsy, increased seizure duration, and increased number of antiseizure medications (ASMs).¹³ 

Several case reports have analyzed occurrences of SUDEP and near-SUDEP not associated with epileptic events, both in a child and adults.^14,15  Given the heterogeneous nature of the data and the complex physiologic interplay of the autonomic, cardiac, and respiratory systems, it is likely that SUDEP cases may arise from a variety of etiologic phenomena.

Risk Factors. A number of risk factors have been identified for SUDEP. We will discuss these risk factors in the following categories: epilepsy type and severity, pharmacologic treatment, and genetics.

The predominant risk factor for SUDEP is a history of GTCS. Patients experiencing 3 or more GTCSs per year are at higher risk.^3,16  Of those for whom seizure type was reported, 53 of 60 (88%) patients experienced GTCs with an average of 16.3 GTCSs per month.³  A separate Swedish case-control study, which included 225 cases of definite or probable SUDEP in patients 4 to 92 years of age, found that patients with a history of GTCs had a 10-fold increased risk of SUDEP.¹⁷  Patients who experienced 1 to 3 GTCSs in the previous year had a 22-fold increase in SUDEP risk, while patients with 4 to 10 GTCSs in the previous year had a 32-fold increase in SUDEP risk.¹⁷ 

Other disease-related risk factors for SUDEP include onset of epilepsy in childhood or adolescence, symptomatic epilepsy, nocturnal seizures, and intellectual impairment or developmental delay.^3,7,17–21  Patients with onset of epilepsy at 0 to 15 years of age have a 7.7 times higher risk of SUDEP compared with those with onset of epilepsy at 45 years of age or later.³  In another study, patients with an onset of epilepsy before 16 years of age had a 1.72-fold increased risk for SUDEP compared with those with an onset of epilepsy at 16 to 60 years of age.¹⁸  In a cohort of 154 patients with SUDEP compared with living age-matched controls, nocturnal seizures were associated with a 2.6-fold increased risk for SUDEP after controlling for other risk factors.¹⁹  Any history of nocturnal GTCS was associated with a 9-fold increased risk of SUDEP, while nocturnal GTCSs in the past year were associated with a 15-fold increased risk of SUDEP.¹⁷ 

While developmental delay and intellectual impairment have been proposed as risk factors, few studies have evaluated whether intellectual disability is an independent risk factor for SUDEP. Abdel-Mannan et al³  noted that 33 (45%) of the evaluated SUDEP cases had developmental delay or intellectual impairment.³  However, 14 children in this study were diagnosed with epileptic encephalopathies known to be associated with intellectual impairment, and etiology was not known for 27 children.³  Because the incidence of epilepsy is known to increase in children with more profound intellectual impairment, it is difficult to determine if this is associated with an increased risk of SUDEP.²² 

Systematic reviews of case-control studies have found that patients receiving more than 1 ASM or experiencing changes in their ASM regimen 3 to 5 times per year were more likely to die of SUDEP.^7,18,21  This relationship could be confounded because patients with uncontrolled or treatment-resistant epilepsy are more likely to have more complicated medication therapy. A case-control study also suggests an association between subtherapeutic drug concentrations and SUDEP because 78 of the 91 patients who died from SUDEP had subtherapeutic drug concentrations, while only 8 of 91 control patients had subtherapeutic drug concentrations (p < 0.0001).²³  Patients not actively treated with ASMs may also be at higher risk.¹⁶ 

Multiple studies have identified genetic risk factors associated with an increased risk for SUDEP. Several genes have been associated with both an increased risk for specific seizure diagnoses and SUDEP, which are summarized in Table 3.^24–26  No specific genes have been associated with a decreased risk of SUDEP. However, patients with idiopathic epilepsy, which is classically understood to be genetic in origin, were disproportionately underrepresented in one study, suggesting that SUDEP may occur less frequently in these patients.³  Genetic associations not directly correlated with epilepsy may also play a role. Genes associated with cardiac disorders, including cardiomyopathies and long QT syndrome, have been identified as possible risk factors for SUDEP.^24–26  Variants of these genes could impact sodium and potassium ion channels located in both the nervous system and heart, possibly leading to a correlation with seizures and cardiac arrest.²⁴  Lastly, variants in genes that regulate opioid signaling pathways, glutamate, and gamma aminobutyric acid in neurotransmission may also be associated with SUDEP.²⁷ 

Medication Adherence. Because the risk of SUDEP increases with the frequency of GTCS, achieving effective seizure control could reduce a patient's risk of SUDEP. Along with rational treatment selection and optimization of therapy as quickly as the patient will tolerate, improved medication adherence may reduce the risk of SUDEP by increasing seizure control.^28–30 

Patterns of medication adherence have been defined in children diagnosed with epilepsy. The mean adherence rate within the first month of treatment following epilepsy diagnosis has been reported as 79.4%.³¹  In another study of pediatric adherence for the first 6 months of treatment following epilepsy diagnosis, near-perfect adherence was achieved by only 42% of children.³²  Thirteen percent of patients were non-adherent.³²  At 1 year 26.6% of patients were not taking their prescribed ASM.³³  However, these studies all excluded patients with significant developmental disorders or other co-morbidities requiring daily medication use, which limits generalizability to children at a higher risk of SUDEP.^31–33 

Single-parent homes,³¹  lower socioeconomic status,^31–34  poorer family communication and problem-solving skills,³⁴  and greater parent fears and stressors³⁴  have been correlated with poorer adherence in children with epilepsy. White coat adherence, the phenomenon of increased adherence in the days prior to clinic visits, has been observed in pediatric patients with epilepsy.³⁵  Through the Pediatric Epilepsy Medication Self-Management Questionnaire parents of children ages 2 to 12 years with new-onset epilepsy identified forgetting to give the medication and their child's disliking the taste as the most common barriers to medication adherence.³⁶  Similarly, higher socioeconomic status, increased incidence of side effects, fewer adherence barriers, greater caregiver responsibility, and lower family conflict are associated with improved medication adherence in adolescents.^37,38 

To date, no studies have directly correlated improved medication adherence with lower rates of SUDEP. However, this may be due to methodologic constraints of such research. Regardless, medication adherence should be promoted in all children diagnosed with epilepsy. Important steps in managing non-adherence include screening for risk factors for non-adherence, identification of reasons for non-adherence, and use of strategies to combat those factors.³⁹  Potential strategies include simplifying drug regimens, epilepsy and medication education, addressing adverse drug reactions, and assisting with memory aids.³⁹  Patients who achieve seizure freedom with their medication may become less strict about their medication adherence; therefore, frequent discussions about medication adherence with patients are important.⁴⁰ 

One proposed and studied method to improve medication adherence in pediatric epilepsy is the Supporting Treatment Adherence Regimen (STAR) intervention.^41,42  The STAR intervention consists of 4 sessions focused on providing information about epilepsy and problem-solving for identified barriers including behavioral contracts and frequent follow-up.^41,42  The authors noted that children were included in the problem-solving process in age-appropriate ways including brainstorming ideas for improving adherence and choosing rewards for good adherence.⁴¹  Though this pilot study was limited by a small sample size, families in the treatment group demonstrated an average improved adherence of 31.5%, while families in the control group had an average improved adherence of 9.3%.⁴¹  The follow-up study found that the STAR method significantly improved patient adherence when compared with the control group following the third and fourth sessions.⁴²  However, at 3-month follow-up there was not a statistically significant difference in the adherence rates between groups.⁴²  These findings highlight the need for continual follow-up and discussion of adherence with patients and their families. Further research into the STAR method is forthcoming.⁴³ 

Another proposed method for improving medication adherence includes the use of text messaging or mobile application–based reminders. A pilot study of the use of technologic reminders for teenagers demonstrated that parents and teenagers found these interventions beneficial although significant changes in adherence were not seen.⁴⁴  While improving medication adherence could decrease seizure frequency and ultimately the risk of SUDEP in children, more research is needed to evaluate the impact of interventions to increase medication adherence and on SUDEP risk.

Non-pharmacologic Approaches. Surgical Intervention. For patients with treatment-refractory epilepsy, surgical intervention is often helpful in improving seizure control. Improved seizure control following epilepsy surgery is correlated with lower mortality overall, and potentially a decreased risk of SUDEP.^45–47  The degree to which surgical interventions impact mortality in epilepsy may be dependent on epilepsy type and specific surgical intervention. In a recent analysis of primarily adult patients with treatment-resistant focal epilepsy, the standardized mortality ratio for SUDEP was 1.9 per 1000 person-years in the surgery group compared with 4.6 per 1000 person-years in the comparator group.⁴⁵  All SUDEP cases in both groups occurred in those with seizures arising from one or both temporal lobes, potentially highlighting the importance of temporal lobe epilepsy and SUDEP risk.⁴⁵  In a separate study of patients who had undergone temporal lobe epilepsy surgery, 6 cases of SUDEP were reported in 299 patients over the study period, translating to an incidence of 1 in 455 person-years.⁴⁸  The small sample sizes limit interpretation of these studies, and more data are needed to fully elucidate the impact of various epilepsy surgery approaches on SUDEP risk.

Vagus nerve stimulation (VNS) therapy has also been reported to decrease the risk for SUDEP. In an analysis of 40,443 patients receiving VNS therapy, age-adjusted SUDEP rates significantly decreased to 2.47 per 1000 person-years in the first 2 years following implantation.⁴⁹  The reported SUDEP was 1.68 per 1000 person-years during years 3 to 10 following VNS.⁴⁹  Of note, the reported rates of SUDEP in this surgical cohort are higher owing to the increased incidence of treatment-resistant epilepsy. Data on the rates of SUDEP following other epilepsy surgical procedures including corpus callosotomy and deep brain stimulation are lacking.

Sleep Safety and Nighttime Supervision. Sleeping in the prone position has been associated with SUDEP, as one cohort study identified that 85 of 97 (87.6%) patients who were asleep when SUDEP occurred were in the prone position.⁵⁰  Because most SUDEP incidents happen during sleep, night-time supervision and the use of lattice pillows have been suggested as measures to prevent SUDEP.²⁸  Lattice pillows are constructed with mesh designed to increase air flow when someone is sleeping in the prone position. One study of patients in an EMU showed that prompt response may improve outcomes when 7 patients were resuscitated from near-SUDEP because Cardiopulmonary Resuscitation was initiated within 3 minutes of the event.¹²  This also highlights the importance of providing parents and caregivers with information about SUDEP and opportunities to learn basic life support.

Diet. Another intervention for pediatric patients with treatment-resistant epilepsy is the ketogenic diet. Research on the ketogenic diet has shown seizure reduction in pediatric patients, but there are many adverse effects that can make the ketogenic diet intolerable.⁵¹  A systematic review and meta-analysis completed by Rezaei et al⁵²  concluded that a modified Atkins diet had similar efficacy as the ketogenic diet with reduced adverse effects. For patients with treatment-resistant epilepsy, the ketogenic diet or modified Atkins diet may be considered to decrease seizure burden and associated complications or concerns including possibly the risk of SUDEP.

The Institute of Medicine, in its publication Epilepsy Across the Spectrum, mentions SUDEP as a basic educational need of all individuals with epilepsy.⁵³ Table 4 displays the recommendations from the American Academy of Neurology and American Epilepsy Society SUDEP Guidelines concerning SUDEP communication.⁵⁴  Most parents express a desire to discuss SUDEP in person with their neurologist upon diagnosis and receive additional resources to help them process this information.^55,56  Patient resources are available through organizations such as the Epilepsy Foundation and Danny Did Foundation. Discussions should also include the known modifiable risk factors of SUDEP, namely optimization of drug therapy to reduce seizure occurrence.⁵⁷  There is debate about when to tell children and their caregivers about SUDEP. Owing to various levels of maturity, development, and independence there is not a defined age when children should be told.⁵⁶  We suggest that education should be provided to parents or caregivers soon after diagnosis, and that timing of communication with the child be part of a shared decision-making partnership with the caregivers.

While most patients desire to know about SUDEP, only 8.7% to 29% of health care professionals in any setting report discussions of SUDEP with all or most of their patients.⁵⁸  Among pediatricians, reasons reported for not discussing SUDEP were a lack of provider knowledge and assumption that specialists such as neurologists would discuss SUDEP with their patients.⁵⁹  One survey of pediatric primary care providers demonstrated that 85.6% of providers were not familiar with SUDEP.⁵⁹  Another reason that providers may not discuss SUDEP with their patients is the belief that their patients are not at risk.

A SUDEP and seizure safety checklist has been developed to aid providers in their discussions with patients, but thus far it has been used only in adult patients.⁶⁰  The SUDEP-7 Inventory was developed to help identify patients at highest risk for SUDEP, but a case-control study of patients ages 10 to 73 years found no statistically significant difference in scores between the cases and controls (p = 0.39).⁶¹  Others have suggested using technology within electronic medical records to identify patients at risk for SUDEP in order to prompt providers to have SUDEP discussions with their patients.⁶²  A clinical trial of a computerized clinical support program to remind providers to discuss SUDEP with pediatric patients and their caregivers has been conducted and results are pending.⁶³  Regardless, there is a discrepancy between information that patients desire to receive about SUDEP and what they are presented by their health care providers.⁶⁴  Providers should be prepared to discuss SUDEP with their patients in an accurate and sensitive manner.⁶⁵ 

Etiology and Risk Factors. Research continues to identify the etiology and risk factors associated with SUDEP. A North American SUDEP registry has been established to include data from medical records and family interviews of patients who die of SUDEP to better characterize the incidence and risk factors of SUDEP.⁶⁶  Another case-control study aims to further understand the risk factors associated with SUDEP, including whether ASMs and nighttime supervision modify this risk.⁶⁷  Another clinical trial is testing the hypothesis of a relationship between SUDEP-7 scores and neural-circulatory control to help identify patients at risk of SUDEP.⁶⁸  Two registered clinical trials are reviewing carbon dioxide and SpO₂ levels and their association with SUDEP.^69,70  Additionally, the Institute of Medicine recommends prevention studies in high-risk individuals to further understand risk factors for SUDEP.⁵³ 

Medication Therapy. There are several novel pharmacotherapeutic agents that may decrease the risk of SUDEP. Selective serotonin receptor inhibitors have been proposed as a possible therapy to prevent SUDEP by improving seizure control and reducing postictal central apnea.⁷¹  This approach has been hypothesized owing to observed similarities between SUDEP and sudden infant death syndrome, which is associated with changes in the serotonergic system.⁷²  A phase 2 clinical trial is testing the impact of fluoxetine on minute ventilation during hypercapnic ventilatory response.⁷³  The use of naloxone immediately following a GTCS in an effort to increase postictal oxygen saturation is also being studied.⁷⁴  Any impact of these agents will require further analysis in the pediatric population, because these studies do not include pediatric patients.

Wearables and Other Technologies. Many watches and bed alarms are being tested and developed.^75–78  These devices could aid in the understanding of SUDEP and help to identify biomarkers while also measuring the frequency of seizures and impact of therapy.⁴⁹  One challenge is creating a technology with both the specificity and sensitivity to accurately detect potential SUDEP events based on physiologic parameters.⁷⁸  The 2019 Prevent21 Summit, a summit held in the United Kingdom focused on SUDEP prevention, recommended that all technologic alarms are only one part of minimizing the risk of SUDEP, because while monitoring and recording seizures is valuable, in order to help prevent SUDEP someone must be available to intervene when a seizure occurs.⁷⁹ 

While many unknowns about SUDEP remain, health care professionals can provide evidence-based information about SUDEP to children and adolescents with epilepsy and their caregivers (Table 5). Health care professionals should work with families to reduce SUDEP risk factors by supporting medication adherence and discussing other treatment options including surgery and dietary interventions. As we await future research further defining the etiology, risk factors, and prevention measures for SUDEP, health care professionals should ensure that families are informed about the risks of SUDEP and steps they can take to help reduce their risk of SUDEP.

ASM

antiseizure medication

EMU

epilepsy monitoring unit

GTCS

generalized tonic-clonic seizure

STAR

Supporting Treatment Adherence Regimen

SUDEP

sudden unexpected death in epilepsy

VEEG

video electroencephalogram

VNS

vagus nerve stimulation