A 20-month-old girl with a complex chromosomal disorder had first presentation of febrile status epilepticus and was admitted to the hospital. Two days after her initial seizure, she died suddenly and unexpectedly during a video EEG monitoring study. An advanced analysis of the physiologic changes in the hours and minutes leading up to death was undertaken. The electrocardiography over the last 19 minutes of life was reviewed, and the R-R intervals were manually measured. Heart rate variability was assessed through calculation of the SD of the R-R intervals and the root mean square of successive differences over successive 100 beat periods. Instantaneous heart rate, SD of the R-R intervals, the root mean square of successive differences, and oxygen saturation were plotted against time over the last 19 minutes of life. Diffuse cerebral suppression on EEG was observed 10 minutes before death, followed minutes later by severe bradycardia and increased heart rate variability. Although the child did not meet criteria for a diagnosis of epilepsy, the sequence of physiologic changes leading up to death suggests a pathophysiology similar to sudden unexplained death in epilepsy. A comparable pattern of diffuse cerebral suppression preceding parasympathetic overactivity has been suggested in some rare cases of adults who have experienced sudden unexplained death in epilepsy during video EEG monitoring.
- ECG —
- HRV —
- heart rate variability
- RMSSD —
- root mean square of successive differences
- SDNN —
- SD of the R-R interval
- SUDC —
- sudden unexpected death in childhood
- SUDEP —
- sudden unexpected death in epilepsy
- VEM —
- video EEG monitoring
Febrile status epilepticus is defined as a febrile seizure or series of febrile seizures without intervening recovery lasting ≥30 minutes.1 Although febrile status is a medical emergency, the mortality risk seems low. The FEBSTAT (Consequences of Prolonged Febrile Seizures in Childhood) study longitudinally followed up 119 children with febrile status epilepticus and reported 2 deaths, neither of which was related to prolonged seizure.2
Death after a seizure is an important issue, as individuals with epilepsy have an increased risk of sudden death. This phenomenon (ie, sudden unexpected death in epilepsy [SUDEP]) is a rare but tragic occurrence, with devastating effects on the individual’s family and friends. SUDEP is less commonly reported in children than in adults,3 potentially because many adult risk factors are less relevant for young children (eg, alcohol use, living alone, poor adherence to medication).
The current article presents the case of a young girl who died suddenly during video EEG monitoring (VEM), days after her first presentation with febrile status epilepticus. We conducted an advanced analysis of the physiologic changes occurring in the hours and minutes before death.
A 20-month-old girl with a complicated medical history underwent prolonged VEM during an admission after first presentation of convulsive febrile status epilepticus.
The child’s medical history was significant for a chromosomal disorder involving distal monosomy 9p, proximal monosomy 15q and 16p13.11 duplication (arr(hg19) 9p24.3p23(204 090–11 371 865)×1, 15q11.1q14(20 686 203–39 443 222)×1, 16p13.11(14 910 213–16 194 575)×3) and associated dysmorphic features, bilateral cleft palate, short webbed neck, congenital heart disease (mild valvular pulmonary stenosis and trivial patent ductus arteriosus), and severe bilateral bronchomalacia. She had a difficult medical course, experiencing respiratory insufficiency requiring ventilator support, chronic hypoglycemia, and frequent infections, all of which had led to numerous prolonged hospitalizations. Neurologic abnormalities included hypotonia from birth and global developmental delay. An MRI of the brain as a neonate showed a simplified gyral pattern, hypo-opercularization of the temporal lobes, and hypoplastic olfactory bulbs and tracts. In discussion with the child’s family and palliative care, a decision was made not to intubate or perform chest compressions in the event of acute deterioration.
While in respite care, at 20 months of age, the child had her first seizure (convulsive febrile status epilepticus lasting 45 minutes) and was admitted to the hospital for treatment of vomiting and diarrhea associated with a rotavirus infection. A 50-minute seizure occurred the next day, associated with hypoglycemia (serum glucose level, 1.8 mmol/L), and levetiracetam treatment was started along with 5% dextrose in water infusion. On day 3, VEM was initiated to assess for subclinical seizures. During the recording, the girl was administered 2 L of 100% oxygen by nasal prongs and the dextrose infusion; serum electrolyte levels were normal that morning, and glucose levels were normal throughout the day. At 6.5 hours into the recording, Code Blue was called for cessation of breathing. Cardiopulmonary resuscitation was briefly initiated but then stopped per the family’s wishes; the girl subsequently died.
During the VEM, electrocardiography (ECG) (lead I) and oxygen saturation (finger sensor) were monitored. The EEG recording was reviewed and the ECG tracing examined beat-by-beat, with each R-R interval manually measured and recorded over the 19 minutes leading up to terminal asystole. Instantaneous heart rate (1/R-R interval) was calculated, as well as 2 heart rate variability measures (SD of the R-R interval [SDNN] and the root mean square of successive differences [RMSSD]). Increases in these heart rate variability (HRV) measures were considered to represent a shift toward parasympathetic over sympathetic autonomic activity.4 Changes over time of SDNN and RMSSD were estimated by calculating values for the 100 R-R intervals surrounding each beat (50 intervals before and 50 intervals after each beat). These values were plotted against time, along with oxygen saturation values at 30-second intervals.
We did not measure PR, QRS, or QT intervals because such measurements differ when calculated from standard 12-lead ECGs compared with single-lead ECG recordings.5 We attempted to assess the respiratory rate because premorbid breathing abnormalities have been described in previous SUDEP studies6,7; however, the child was covered with a blanket for much of the recording, barring reliable measurement.
For the majority of the VEM, the EEG showed diffuse, high-amplitude delta activity, with relative attenuation over the left hemisphere, consistent with a diffuse encephalopathy. Frequent multifocal spikes were seen from the left frontal, right posterior temporal, and left anterior temporal regions, suggesting a predisposition for focal seizures from these regions. Occasional periods of diffuse attenuation lasting 2 to 5 seconds were observed (Fig 1A), although the recording did not resemble burst suppression or hypsarrhythmia, patterns seen in some developmental epileptic encephalopathies.
No clinical or electrographic seizures were observed during 6.5 hours of VEM before death. Ten minutes before terminal asystole (T-10 minutes), there was an abrupt onset of diffuse attenuation, which did not significantly improve (Fig 1B).
Single-lead ECG showed sinus rhythm for the first 6 hours of recording, including up to the time of diffuse EEG suppression at T-10 minutes (Fig 2A).
At T-8 minutes, profound sinus bradycardia was seen with ventricular escape beats (Fig 2B). At T-7 minutes, ongoing ventricular escape rhythm was established (Fig 2C). Sinus bradycardia was again observed by T-4 minutes (Fig 2D). The terminal rhythm was sinus bradycardia progressing to asystole.
Heart rate was steady at 120 to 135 beats/min until T-8 minutes, after which the heart rate decreased and plateaued at ∼70 beats/min (the ventricular escape rate) (Fig 3). SDNN and RMSSD both remained low until T-8.5 minutes. At that point, SDNN began to rise, followed less than 1 minute later by RMSSD. Both measures peaked at approximately T-6.5 minutes, returned almost to baseline by T-4 minutes, and then spiked dramatically in the last 3 minutes before death. The increase in HRV correlates with ventricular escape rhythm with atrial capture beats causing significant R-R interval variability. The dramatic variability spike during the terminal event correlates with profound sinus bradycardia with prolonged pauses, with ventricular escape beats occurring at increasing R-R intervals before asystole.
Peripheral oxygen saturation did not drop significantly until T-5 minutes (Fig 3). This measure may have been influenced by the patient’s oxygen therapy.
The present study describes the sequence of events leading up to sudden unexpected death in childhood (SUDC), days after initial presentation with febrile status epilepticus. The girl did not meet the criteria for a diagnosis of epilepsy because her 2 seizures were provoked,8,9 although the EEG suggested a high risk of further seizures. Physiologic monitoring data suggest the cause of death was rooted in the central nervous system, following a pattern reminiscent of recently published adult SUDEP cases.7,10
The first observed physiologic change was diffuse cerebral suppression on the EEG, a feature universally reported in adult SUDEP during VEM. This widespread cortical suppression usually occurs immediately after convulsive status epilepticus; however, a recent series demonstrated that the phenomenon may occur without preceding seizure.6,7 In such cases, a central nervous system precipitating event is suspected, although the underlying mechanism remains unclear. In retrospect, the main warning sign for our patient was brief periods of diffuse cerebral suppression leading up to the final, severe attenuation, suggesting cerebral dysfunction was the initial trigger for the cascade of events leading to death. Oxygen saturation and heart rate were stable throughout the recording, effectively ruling out global hypoperfusion or hypoxia as inciting factors.
The diffuse cerebral suppression was followed minutes later by increased parasympathetic activity. This change, manifested by severe sinus bradycardia, increased HRV and emergence of ventricular escape rhythm, briefly abated, and then increased again in the last minutes of life. Increased parasympathetic activity before death was shown convincingly by Jeppesen et al,10 who performed HRV analysis in an adult patient with SUDEP during VEM. The phenomenon is also supported by the recent series of Lhatoo et al7 that documented progressively worsening bradycardia after onset of cerebral suppression. These findings suggest that some cases of SUDEP, as well as our case, may occur due to pathologically increased parasympathetic activity leading to severe bradycardia and reduced cardiac contractility,11 followed by respiratory arrest.
Risk of sudden death in children with seizures may be increased if brief periods of generalized EEG suppression are observed. This scenario is perhaps not surprising given that postictal cerebral suppression has been suggested as a marker of SUDEP risk.12 However, our findings emphasize that the diffuse suppression need not follow a seizure to be clinically significant.
Autonomic dysfunction may predispose to sudden death, and altered HRV has been proposed as an independent risk factor for SUDEP.10,13–15 Our patient had relatively low baseline SDNN and RMSSD for the first 6 hours of the prolonged recording, with HRV only spiking in the minutes before death. This scenario is considerably different from the adult case reported by Jeppesen et al,10 in which HRV progressively increased over many hours leading up to death. Age-dependent patterns of autonomic dysfunction before sudden death may not be surprising given that HRV decreases with age.16 Although this report is the first to describe SUDEP-like death after febrile seizure, the phenomenon may be underrecognized given that a disproportionate number of children who experience SUDC (31.7%) have a history of febrile seizures.17 Interestingly, certain genes have been implicated in both epilepsy and Brugada syndrome, a cardiac conduction abnormality that can cause sudden death, in some cases provoked by fever.18
We have presented a case of SUDC during VEM and undertaken an advanced analysis of the physiologic changes leading up to death. Adult SUDEP frequently involves diffuse cerebral suppression on EEG, followed by increased parasympathetic activity, patterns which were also observed in our patient. Although almost certainly a rare occurrence, this case shows that children with febrile seizures can die suddenly and unexpectedly in a manner reminiscent of adult SUDEP. The case may be representative of the physiologic changes preceding death in pediatric SUDEP; however, this theory cannot be confirmed without identification and investigation of definite SUDEP cases involving children during VEM.
- Accepted February 23, 2017.
- Address correspondence to Kenneth A. Myers, MD, PhD, Epilepsy Research Centre, Melbourne Brain Centre, Austin Health, 245 Burgundy St, Heidelberg, VIC 3084, Australia. E-mail:
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.
FUNDING: No external funding.
POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose.
- Katona PG,
- Jih F
- Silvani A,
- Calandra-Buonaura G,
- Dampney RA,
- Cortelli P
- Sivakumar SS,
- Namath AG,
- Tuxhorn IE,
- Lewis SJ,
- Galán RF
- Umetani K,
- Singer DH,
- McCraty R,
- Atkinson M
- Copyright © 2017 by the American Academy of Pediatrics