Association of Peri-ictal Brainstem Posturing With Seizure Severity and Breathing Compromise in Patients With Generalized Convulsive Seizures

Abstract

Objective

To analyze the association between peri-ictal brainstem posturing semiologies with postictal generalized electroencephalographic suppression (PGES) and breathing dysfunction in generalized convulsive seizures (GCS).

Methods

In this prospective, multicenter analysis of GCS, ictal brainstem semiology was classified as (1) decerebration (bilateral symmetric tonic arm extension), (2) decortication (bilateral symmetric tonic arm flexion only), (3) hemi-decerebration (unilateral tonic arm extension with contralateral flexion) and (4) absence of ictal tonic phase. Postictal posturing was also assessed. Respiration was monitored with thoracoabdominal belts, video, and pulse oximetry.

Results

Two hundred ninety-five seizures (180 patients) were analyzed. Ictal decerebration was observed in 122 of 295 (41.4%), decortication in 47 of 295 (15.9%), and hemi-decerebration in 28 of 295 (9.5%) seizures. Tonic phase was absent in 98 of 295 (33.2%) seizures. Postictal posturing occurred in 18 of 295 (6.1%) seizures. PGES risk increased with ictal decerebration (odds ratio [OR] 14.79, 95% confidence interval [CI] 6.18–35.39, p < 0.001), decortication (OR 11.26, 95% CI 2.96–42.93, p < 0.001), or hemi-decerebration (OR 48.56, 95% CI 6.07–388.78, p < 0.001). Ictal decerebration was associated with longer PGES (p = 0.011). Postictal posturing was associated with postconvulsive central apnea (PCCA) (p = 0.004), longer hypoxemia (p < 0.001), and Spo₂ recovery (p = 0.035).

Conclusions

Ictal brainstem semiology is associated with increased PGES risk. Ictal decerebration is associated with longer PGES. Postictal posturing is associated with a 6-fold increased risk of PCCA, longer hypoxemia, and Spo₂ recovery. Peri-ictal brainstem posturing may be a surrogate biomarker for GCS severity identifiable without in-hospital monitoring.

Classification of Evidence

This study provides Class III evidence that peri-ictal brainstem posturing is associated with the GCS with more prolonged PGES and more severe breathing dysfunction.

Sudden unexpected death in epilepsy (SUDEP) is the leading category of death in patients with refractory epilepsy, with an incidence of 6.3 to 9.3 per 1,000 person-years in this population.^1,2 Frequent generalized convulsive seizures (GCS) in patients with long-standing, early-onset epilepsy have the most significant risk.³ Recent studies have focused on determining pathophysiology and electroclinical risk factors for SUDEP as well as markers of GCS severity. These factors include prolonged ictal central apnea (ICA), postconvulsive central apnea (PCCA), hypoxemia severity, postictal blood catecholamine rise, and prolonged (>50 seconds) postictal generalized electroencephalographic suppression (PGES).^4–9 PGES was observed in all monitored SUDEP cases in the Mortality in Epilepsy Monitoring Unit Study (MORTEMUS) along with cardiorespiratory instability.¹⁰ Although its role as a risk marker of SUDEP has not been prospectively confirmed, prolonged PGES is seen with severe GCS, cardiorespiratory compromise, and delayed arousal.^6,10–12 GCS tonic phase semiology and duration are strongly linked to PGES incidence, particularly when characterized by bilateral symmetric tonic arm extension (decerebrate) posturing.^13–15 Tonic or dystonic posturing can also be postictal, although its symptomatogenic brain areas and its relationship to postictal cardiorespiratory compromise are unknown.¹⁶ Brainstem seizure spread may potentially explain both.^14,17 Semiologic clinical features such as posturing can be recognized without the need for multimodal monitoring and thus may have value in seizure severity assessment. We sought to precisely study GCS features, including tonic phase semiology and postictal posturing and their association with potential SUDEP biomarkers such as PGES and peri-ictal breathing dysfunction.

Methods

The primary research question is to determine the association between peri-ictal brainstem posturing and presence of PGES and its duration as well as breathing compromise.

Standard Protocol Approvals, Registrations, and Patient Consents

Written informed consent was obtained prospectively from all the participants in the National Institute for Neurological Disorders and Stroke Center for SUDEP Research's Autonomic and Imaging Biomarkers of SUDEP multicenter project (U01-NS090407) and its preliminary phase, the Prevention and Risk Identification of SUDEP Mortality (PRISM) project (P20NS076965). These studies were approved by the Institutional Review boards of the participating centers.

Patient Selection

Patients with intractable epilepsy (failure of adequate trials of ≥2 antiepileptic medications)¹⁸ who were ≥18 years of age and were undergoing video-EEG (VEEG) evaluation in the adult epilepsy monitoring units of participating centers from February 2011 until April 2018 were selected. We included patients with recorded GCS that were successfully analyzed until April 2018, including generalized tonic-clonic seizures, focal to bilateral tonic-clonic seizures, and focal-onset motor bilateral clonic seizures.¹⁹ Exclusion criteria were status epilepticus and obscured or unavailable video. Demographic and clinical data were collected, including epilepsy duration, seizure type and frequency, semiologic seizure features, awake or asleep states at seizure onset, and presence of major cardiac (cardiac ischemic disease, known arrhythmia, valvulopathy) or respiratory (obstructive sleep apnea, asthma, chronic obstructive pulmonary disease, bronchiectasis, cystic fibrosis) disease. We considered the use of serotonin or serotonin-noradrenaline reuptake inhibitors. We assessed the impact of antiepileptic drug regimen during admission on tonic phase semiology. Epilepsy type was classified as generalized (genetic generalized epilepsy in all cases), focal, both, or unknown.²⁰ GCS duration was defined as time from onset of bilateral motor signs of tonicity or clonicity to clinical seizure end, and GCS phases were classified as tonic, jittery, and clonic.

Data Collection

Semiology Classification

Tonic phase semiology was classified into 4 categories based on a modified classification proposed by previous authors¹³: (1) ictal decerebration (bilateral symmetric tonic arm extension), (2) ictal decortication (bilateral symmetric tonic arm flexion without progression to decerebration), (3) ictal hemi-decerebration (tonic extension of 1 arm with flexion of contralateral arm without progression to decortication or decerebration), and (4) absence of ictal tonic phase. Examples of brainstem posturing are provided in videos 1 through 3.

Video 1

Decerebration during tonic phase (audio and images have been edited to protect patients' identity).Download Supplementary Video 1 ^{(10.8MB, mov)} via http://dx.doi.org/10.1212/011274_Video_1

Video 2

Decortication during tonic phase (audio and images have been edited to protect patients' identity).Download Supplementary Video 2 ^{(13.4MB, mov)} via http://dx.doi.org/10.1212/011274_Video_2

Video 3

Asymmetry (hemi-decerebration) during tonic phase (audio and images have been edited to protect patients' identity).Download Supplementary Video 3 ^{(12.5MB, mov)} via http://dx.doi.org/10.1212/011274_Video_3

Postictal posturing referred to patients adopting decerebration or decortication after the last clonic jerk of the GCS. An example is provided in video 4.

Video 4

Post-ictal posturing (audio and images have been edited to protect patients’ identity).Download Supplementary Video 4 ^{(11.2MB, mov)} via http://dx.doi.org/10.1212/011274_Video_4

Cardiorespiratory Monitoring and VEEG Monitoring

All patients underwent prolonged surface VEEG monitoring with the 10-20 International Electrode System. EEG and ECG were acquired with Nihon Kohden (Tokyo, Japan), Micromed (Modigliani Veneto, Italy), and Xtlek and Nicolet (Natus Medical Inc, Pleasanton, CA) acquisition platforms. Peripheral capillary oxygen saturation (Spo₂) was monitored with pulse oximetry (Nellcor OxiMax N-600x [Covidien, Dublin, Ireland], Masimo Radical-7 [Irvine, CA], and SenTec Digital Monitoring System [Therwil BL, Switzerland]), and chest wall and abdominal excursions were recorded with inductance plethysmography (Ambu [Ballerup, Denmark] Sleepmate and Perfect Fit 2 [Dymedix, Shoreview, MN]).

Breathing analysis for apnea used a composite analysis of inductance plethysmography, EEG breathing artifact, and visually inspected thoracoabdominal excursions 2 minutes before seizure onset (clinical or electrographic, whichever occurred first) and up to 3 minutes after clinical seizure end. Central apnea (cessation of thoracoabdominal breathing movements) was defined as >1 missed breath without other explanation (i.e., speech or intervention), with a minimum duration of 5 seconds. ICA referred to apnea occurring in the preconvulsive phase of GCS. PCCA referred to apnea after GCS; we preferred this term to post-ICA because apnea could occur after convulsions but with ongoing EEG seizure discharges. Incidences and durations of ICA and PCCA were determined. Apnea was not assessed during the GCS phase because of invariable artifact in breathing channels.

Baseline Spo₂ was determined as the mean value in a 15-second page at 2 minutes before EEG onset or clinical onset, whichever occurred first. We defined change in Spo₂ as the difference between baseline and the lowest Spo₂ value (nadir Spo₂) recorded during or up to 3 minutes after clinical seizure end. Hypoxemia was defined as Spo₂ <95%. When baseline Spo₂ was already <95%, a >1% drop was considered significant. If transient loss of Spo₂ signal occurred during monitoring but hypoxemia persisted after signal recovery, hypoxemia duration was determined but not Spo₂ nadir (and thus change in Spo₂). If Spo₂ signal did not return or hypoxemia had resolved, we made no comment on change in Spo₂ or hypoxemia duration. Finally, to avoid the effect of seizure duration and following previous studies, we determined time to recovery to mild hypoxemia (Spo₂ 90%) after clinical seizure end, which we called Spo₂ recovery.⁷ We considered early oxygen administration when it was applied during the seizure or within 5 seconds of seizure termination.¹³

Presence and duration of PGES⁶ were determined by a validated automated EEG suppression detection tool²¹ and supplemented by visual analysis by the same 2 epilepsy neurophysiologists in all cases when the tool gave no solution. The visual inspection was masked to VEEG results for 1 neurophysiologist but not for the other one.

Statistical Analysis

Descriptive statistics (mean, SD, frequency, percentage, etc) were provided for demographic and clinical variables based on patients and seizures (table 1). Descriptive statistics for continuous outcomes (PGES duration, change in Spo₂, hypoxemia duration, and Spo₂ recovery) are provided in tables 2 and 3. Mean and SD of the continuous outcomes across seizures were provided for categorical demographic and clinical variables. Considering that the outcomes are repeated measures, p values were obtained from the generalized estimating equation (GEE) method to account for within-participant correlation. For continuous demographic and clinical variables, covariate coefficient estimates, standard error, and corresponding p values from the GEE method were provided. Descriptive statistics for dichotomous outcomes (PGES, PCCA) are provided in table 4. Frequency and percentage and mean and SD were provided for categorical and continuous variables, respectively. The p values were obtained from GEE as well, with the binomial distribution and logit link. Based on the univariate analysis shown in tables 2 through 4, the multivariable analysis is presented in tables 5 and 6. Variables from the univariant analysis with p < 0.1 were included in the final models, and age at study and sex were treated as force-in variables. A value of p < 0.05 in the final models was considered significant. All analyses were performed in SAS 9.4 (SAS Institute Inc, Cary, NC).

Table 1.

Demographic and Phenotypic Variables

Table 2.

Univariate Analysis for Continuous Variables With Categorical Independent Variables

Table 3.

Univariate Analysis for Continuous Variables With Independent Continuous Variables

Table 4.

Univariate Analysis for Categorical Variables

Table 5.

Multivariate Analysis for PGES Incidence and Duration

Table 6.

Multivariate Analysis for Respiratory Outcomes

Data Availability

The datasets used and analyzed during the current study are available from the corresponding author on request.

Results

Demographics and Clinical Phenotype

We identified 307 GCS in 187 patients. VEEG recordings meeting study criteria were available in 295 seizures in 180 patients (90 female). Two hundred thirty-seven seizures were included in 2 previous publications on peri-ictal breathing dysfunction.^5,22 Mean age at monitoring was 36.7 ± 13.3 years (median 34; range 18–77 years). Mean age at epilepsy onset was 19.6 ± 15.5 years (16; 1–68 years), and mean epilepsy duration was 16.8 ± 12.1 years (15; 1 month–45 years). Epilepsy type was generalized in 29 patients (16.1%), focal in 145 (80.6%), and unknown in 5 (2.8%). One patient had both focal and generalized epilepsy. Details regarding demographic and phenotypic characteristics are summarized in table 1.

Seizure Characteristics

One hundred forty-eight seizures occurred during wakefulness; 144 occurred during sleep; and 3 occurred during postictal stupor in a seizure cluster in 1 patient.

Total GCS duration was 52.3 ± 17.9 seconds (51; 5–154 seconds). Tonic phase was present in 197 of 295 (67%) seizures (mean duration 7.9 ± 4 seconds [7; 1–22 seconds]), and jittery phase was present in 238 of 295 (80.7%) seizures (mean duration 9.5 ± 7.3 seconds [7; 1–55 seconds]). All seizures had a clonic phase, with a duration of 39.3 ± 17.7 seconds (36; 5–123 seconds). Ictal decerebration was observed in 122 of 295 (41.4%) seizures; ictal decortication was seen in 47 of 295 (15.9%) seizures; and ictal hemi-decerebration was observed in 28 of 295 (9.5%) seizures. We found no association between antiepileptic drug regimen or medication reduction/cessation and tonic phase semiology (p > 0.05).

Postictal posturing occurred in 18 of 295 (6.1%) seizures in 12 patients (6.6%). In 16 of 18 (88.8%) seizures in 10 of 12 patients (83.3%), tonic flexion of the upper extremities, identical to ictal decortication, was observed. In the remaining 2 of 18 seizures (2 of 10 patients), tonic extension of the upper extremities was noted, similar to ictal decerebration. Electrographic burst discharge was simultaneous with decortication in 2 seizures (2 patients) followed by PGES. In the remainder, this occurred concurrently with PGES. Posturing occurred 7 ± 7.9 seconds (4; 1–30 seconds) after the last clonic jerk.

PGES was present in 197 of 293 (67%) GCS in 132 patients, with a mean duration of 36.5 ± 21.4 seconds (35; 1–169 seconds); it could not be assessed in 2 seizures due to electrode artifact.

ICA was observed in 83 of 205 (40.4%) seizures in 48 patients (mean duration 14.7 ± 8.6 seconds [12; 5–39 seconds]), and PCCA was seen in 45 of 285 (15.8%) seizures in 34 patients (mean duration 11.2 ± 12 seconds [8; 5–85 seconds]). No comment could be made on the incidence of ICA in 90 seizures and the incidence of PCCA in 10 seizures due to movement artifact or loss of polygraphic data.

Hypoxemia duration, available in 127 seizures, was 142.6 ± 65.5 seconds (124; 25–314 seconds). In analysis of Spo₂ recovery from clinical seizure end, available in 120 seizures, it was 43.2 ± 34.2 seconds (35.5; −27 to 179 seconds). Finally, Spo₂ change (baseline to nadir), available in 119 seizures, was 34.4 ± 14.5% (33%; 2–77%).

Association of Peri-Ictal Semiology With PGES and Breathing Dysfunction

In univariate analyses, tonic phase semiology was related to PGES presence (p = 0.000), PGES duration (p = 0.034), and change in Spo₂ (p = 0.024). Tonic phase semiology was not related to the presence of ICA (p = 0.906) or PCCA (p = 0.546). In the univariate analysis, for the subset of patients with tonic phase, its duration was associated with total hypoxemia duration (p = 0.027) and Spo₂ recovery (p = 0.049). However, there was no significant association of tonic phase duration with PGES presence (p = 0.376) or duration (p = 0.791) or with change in Spo₂ (p = 0.822). There was also no association of tonic phase duration with ICA (p = 0.965) or PCCA (p = 0.712). Postictal posturing was associated with PGES (p = 0.001, tables 2–4).

In multivariate analysis, the presence of ictal decerebration (odds ratio [OR] 14.79, 95% confidence interval [CI] 6.18–35.39, p < 0.001), ictal decortication (OR 11.26, 95% CI 2.96, 42.93, p < 0.001), or ictal hemi-decerebration (OR 48.56, 95% CI 6.07–388.78, p < 0.001) was associated with increased risk for PGES compared to the absence of any tonic phase. PGES duration was significantly longer in those seizures with ictal decerebration (β estimate [Est] 20.45 seconds, 95% CI 4.74–36.15, p = 0.011) compared to seizures without tonic phase. No differences were noted in PGES duration between seizures with ictal decortication (Est 11.09, 95% CI −4.41 to 26.59, p = 0.161) or hemi-decerebration (Est 5.22, 95% CI −10.16 to 20.61, p = 0.506) and those seizures without tonic phase. PGES duration was also longer with increasing age at the time of study (Est 0.51 second, 95% CI 0.13–0.89, p = 0.008, table 5).

Ictal decerebration (Est 9.57%, 95% CI 3.83–15.32, p = 0.001), ictal decortication (Est 11.37%, 95% CI 4.32–18.42, p = 0.002), and ictal hemi-decerebration (Est 12.52%, 95% CI 4.19–20.84, p = 0.003) were also related to larger drops in Spo₂ compared to patients without tonic phase. Changes in Spo₂ were smaller in patients with respiratory comorbid conditions (Est −9.38%, 95% CI −15.26 to −3.50, p = 0.002, table 6).

Postictal posturing was associated with increased risk of PCCA (OR 6.06, 95% CI 1.76–20.89, p = 0.004). Other variables associated with PCCA were sex (male, relative risk 0.26, 95% CI 0.09–0.73, p = 0.010), epilepsy type (focal, relative risk 0.29, 95% CI 0.11–0.80, p = 0.017), and shorter duration of GCS (OR 0.95, 95% CI 0.91–0.99, p = 0.017). Postictal posturing was associated with prolonged hypoxemia duration (Est 47.87 seconds, 95% CI 24.47–71.27, p < 0.001). Hypoxemia duration was also longer in men (Est 40.14 seconds, 95% CI 16.61–63.67, p < 0.001), increased with GCS duration (Est 0.87 second, 95% CI 0.03–1.70, p = 0.041), and decreased with increasing age at study (Est −1.61 seconds, 95% CI −2.44 to −0.78, p < 0.001). Postictal posturing was also associated with longer Spo₂ recovery (Est 27.84 seconds, 95% CI 1.98–53.69, p = 0.035). Conversely, Spo₂ recovery was shorter with early administration of oxygen (Est −17.69 seconds, 95% CI −29.56 to −5.83, p = 0.003) and with increased duration of the GCS (Est −0.53 second, 95% CI −0.92 to −0.14, p = 0.009, table 6). Given the apparent paradoxical results regarding GCS duration and its association with PCCA and Spo₂ recovery, we sought to determine the ratio of tonic phase duration to clonic phase duration and its association with overall GCS duration. An increase in tonic/clonic duration ratio was associated with a decrease in total GCS duration (Est −19.29 seconds, 95% CI −29.52 to −9.07, p < 0.001).

Discussion

Our findings suggest that peri-ictal semiology is related to markers of GCS seizure severity such as PGES and peri-ictal breathing dysfunction in the form of PCCA and oxygen desaturation. We found a clear gradation of semiologic severity such that the presence of ictal decerebration, decortication, and hemi-decerebration was associated with the most striking signs of compromise (presence of PGES and larger drops in Spo₂), with ictal decerebration being associated with prolonged PGES. Absence of GCS tonic phase was associated with less profound changes. We also made the novel observation that postictal brainstem-type posturing is related to a 6-fold increased risk for PCCA and to longer hypoxemia duration and Spo₂ recovery periods after GCS seizures. Because PCCA has been observed in SUDEP and near-SUDEP, postictal brainstem posturing may suggest a semiologic marker of seizure severity and reflect a brainstem mechanism for SUDEP and near-SUDEP phenomena.

Decerebration and decortication are release phenomena in animal brainstem transection and stimulation studies^23–25 and are used to grade severity of encephalopathy in the Glasgow Coma Scale.²⁶ Brainstem transection between the red nucleus and vestibular nuclei produces decerebration, resulting from loss of inhibitory cerebral and cerebellar input on tonic vestibular responses, and disruption of rubrospinal function, resulting in opisthotonic posturing.²⁷ Brainstem transection above the red nucleus effectively removes most cortical influences, leaving unrestrained intact cerebellar afferents to vestibular nuclei.^27,28 Human studies provide less precise anatomic correlates, although flexor (decorticate) responses likely reflect more rostral and less severe supratentorial involvement than extensor (decerebrate) responses.^26,29 Functional, reversible decerebrate and decorticate responses similar to those found in GCS occur in human hepatic and other nonstructural causes of coma.^30,31 Similar posturing can occur in the postictal state. Immediate postictal tonic contractions were described by Gastaut and Broughton³² at times as being “as intense as that of the tonic phase of the tonic-clonic attack,” with trismus, and limb and back extension, indicating what they described as a “functional decerebrate state” in the absence of scalp EEG discharges. However, these have not hitherto been associated with seizure severity or SUDEP risk.

Tonic posturing during and after GCS may indicate dysfunction in cortical and diencephalic influences on descending pathways exerted through brainstem and cerebellar nuclei, likely through disinhibitory processes. The various patterns of observed posturing may reflect extent of seizure spread, with most caudal bilateral spread causing the most severe tonic semiologies. Sensitive respiratory structures amenable to descending seizure influences include the periaqueductal gray (PAG) and parabrachial pons, putative pre-Botzinger area, raphe nuclei, solitary tract nucleus, and nucleus ambiguus.^33–36 The PAG integrates multiple cortical and subcortical afferent signals and influences several respiratory-regulatory nuclei such as the pre-Botzinger complex. Ventrolateral caudal PAG activation in the cat decreases spontaneous activity and responsiveness to surrounding stimuli and elicits irregular breathing, hypotension, and bradycardia.³⁵ The ventrolateral medulla shows serotoninergic neuronal loss in patients with SUDEP; seizure spread to such brainstem levels, as evidenced by characteristic posturing, may produce postictal respiratory compromise in high-risk patients.³⁷ At the same time, disruption of ascending pathways, which impinge on cortical, basal ganglia, and other rostral motor control structures, may prolong the comatose postictal state and impair the protective effect of arousal.³⁸ PGES may reflect both cortical descending dysfunction and disruption of ascending inputs.²²

Another potential explanation for posturing during and after GCS is brainstem depolarization.³⁹ Brainstem seizures have not been elicited in humans but have been triggered in animals after stimulation of the mesencephalic reticular formation, pons, and medulla.^24,40 PAG hyperactivation occurs in audiogenic seizures.⁴¹ In a rodent model of 4-aminopyridine–induced hippocampal seizures, only those rats receiving high doses of 4-aminopyridine with tonic-clonic seizures and longer hippocampal discharges exhibited brainstem discharges. Longer brainstem discharges (>30 seconds) were associated with a respiratory arrest and accompanying cortical and hippocampal EEG flattening. In this study, spreading depression in the brainstem was not noted before respiratory dysfunction.¹⁷ Similarly, in Kv1.1 knockout and Scn1a mice, an animal model of SUDEP, postictal spreading depolarization in the dorsal medulla after seizures produced cardiorespiratory arrest, preceded by EEG suppression and apnea.⁴² Spreading depolarization has also been reproduced recently in a homozygous Cacna1a mouse model, in this case coincidently with apnea.⁴³ Specific subcortical structures such as superior olivary complex, PAG, pontine and midbrain reticular formation, substantia nigra pars reticularis, and amygdala, as well as Kolliker-Fuse, facial nucleus, and rostroventrolateral medullar, were significantly activated in an MRI study of DBA/1 mice with audiogenic seizures and seizure-induced respiratory arrest.⁴⁴ These findings suggest widespread but unsuccessful activation of compensatory mechanisms needed to overcome respiratory arrest. PAG stimulation in DBA/1 mice and C57BL/6 mice (nonepileptic mice) produced significant intensity-related decreases in interbreathing interval in both strains.⁴⁵ However, the effects were significantly reduced in DBA/1 compared to C57BL/6 mice, suggesting that PAG-deficient responses would confer susceptibility to seizure-induced cardiorespiratory failure.⁴⁵ Lastly, in the same animal model of SUDEP, neural activity in PAG was enhanced when a selective serotonin reuptake inhibitor was administered, preventing seizure-induced SUDEP.⁴⁶ These results are broadly in line with human neuroimaging and neuropathologic studies that show damage in brainstem structures responsible for breathing modulation.^37,47

Lastly, hypoxemia has been reported to cause reversible decerebration and decortication in humans. This suggests that hypoxemia during GCS could functionally transect the cerebrum from caudal structures, which would be reflected as postictal posturing. Ictal decerebration is associated with PGES,^13,48 although none of the previous studies observed decortication, which occurred in 16% of seizures in our study. We found any ictal brainstem posturing to be associated with PGES compared to seizures without tonic phase. However, when decerebration occurred, PGES duration was significantly longer; this lengthening did not occur with other semiologies. We postulate that ictal decerebration may be a clinical manifestation of caudal brainstem seizure spread, which in turn causes more severe cortical deafferentation, reflected by longer PGES duration. Thus, ictal decerebration may be a potential clinical biomarker of SUDEP.

Our finding of a relationship between postictal brainstem posturing and PCCA is intriguing. The former is a known phenomenon³² further described in 31 GCS in 16 patients, in whom 48% of seizures had postictal clinical motor manifestations, including focal dystonic posturing.¹⁶ Although precise descriptions of such posturing were not provided by the authors, the very specific brainstem-type posturing described in our study was found in only a minority of our study seizures (6%). Such postictal phenomena may represent seizure discharges in unrecorded brain regions such as the brainstem.¹⁶ Direct human recordings of brainstem-propagated seizures are lacking, although there is some animal evidence to this effect.¹⁷ There appears to be no direct causal relationship between ictal and postictal brainstem posturing, although it is clear that ictal brainstem posturing is associated with larger changes in Spo₂ and that decerebration is particularly related to prolonged PGES. Thus, there is a setting for severe breathing compromise in patients with ictal decerebration, and the subsequent occurrence of postictal brainstem posturing and PCCA in such patients may prove fatal. The 6-fold elevation in PCCA risk with postictal brainstem posturing and the prolongation in Spo₂ recovery and hypoxemia duration are striking findings and encourage scrutiny of the postictal VEEG recording in patients with high-risk SUDEP phenotypes.

Our study is a multicenter, prospectively designed study with a large sample size and detailed cardiorespiratory polygraphy compared to previous studies.^13,48 However, several limitations should be considered. Regarding consideration of false positives, the results for our main findings remain significant (p < 0.01 or 0.001), even after adjustment for multiple testing on 6 primary outcomes; thus, our conclusions remain. Our definition of apnea differs from previous extended definitions (10-second duration) based on sleep studies. Our definition is pragmatic, reflecting stimulation studies for symptomatogenic zones underpinning ICA, which has a consistent minimum duration of 5 seconds, even with brief 2-second stimulation bursts.⁴⁹ Thus, our definition is more sensitive to transient disturbances of breathing but may overdetect apnea.^5,22 Information regarding hypoxemia was available in <43% of seizures due to absence of Spo₂ sensors or loss of signal during monitoring from tonic-clonic movements, which is a difficulty consistently reported in prior literature.⁷ However, we confirmed earlier observations regarding the effect of oxygen administration on Spo₂ recovery after GCS, which validates the reliability of the results.^7,50 Paradoxically, in our study, we found that PCCA was associated with shorter duration of GCS and similarly that Spo₂ recovery decreased with longer GCS duration. However, shorter duration of the GCS was associated with a more prolonged tonic phase compared to clonic phase. Our hypothesis is that not the tonic phase duration itself but its duration in comparison to the clonic phase duration may explain the seemingly paradoxical results. There were only 45 seizures in 34 patients with PCCA and 16 seizures in 12 patients with postictal brainstem posturing, and validation is required in a larger dataset, which we hope to achieve at the conclusion of this multicenter study. Our analysis did not include SUDEP outcomes in our patients, and thus, extrapolation of our findings to the SUDEP and near-SUDEP settings is speculative. Nonetheless, we believe that ictal and postictal brainstem posturing is associated with biomarkers of GCS severity, determined by PGES presence and duration and breathing compromise in the form of oxygen desaturation and PCCA. Further prospective follow-up is required to validate this hypothesis and to elucidate the role of peri-ictal semiology and SUDEP risk.

Glossary

CI

confidence interval

EST

β estimate

GCS

generalized convulsive seizures

GEE

generalized estimating equation

ICA

ictal central apnea

MORTEMUS

Mortality in Epilepsy Monitoring Unit Study

OR

odds ratio

PAG

periaqueductal gray

PCCA

postconvulsive central apnea

PGES

postictal generalized electroencephalographic suppression

PRISM

Prevention and Risk Identification of SUDEP Mortality

SUDEP

sudden unexpected death in epilepsy

VEEG

video-EEG

Appendix. Authors

Study Funding

Samden Lhatoo is funded by the Center for SUDEP Research: NIH/National Institute for Neurological Disorders and Stroke (NINDS) U01-NS090405 and NIH/NINDS U01-NS090407. Orrin Devinsky is funded by the Center for SUDEP Research: NIH/NINDS U01-NS090407 and NS090415. He has equity interest in Empatica, Tilray, Receptor Life Sciences, Egg Rock, Rettco, Qstate Biosciences, Tevard, and Engage. George Richerson is funded by the Center for SUDEP Research: NIH/NINDS U01-NS090414.

Disclosure

Laura Vilella, Nuria Lacuey, Johnson P. Hampson, Liang Zhu, M.R. Sandhya Rani, Shirin Omidi, Manuela Ochoa-Urrea, Shiqiang Tao, and Rup K. Sanju report no disclosures. Daniel Friedman receives salary support for consulting and clinical trial related activities performed on behalf of The Epilepsy Study Consortium, a nonprofit organization. Dr. Friedman receives no personal income for these activities. NYU receives a fixed amount from the Epilepsy Study Consortium toward Dr. Friedman’s salary. Within the past year, The Epilepsy Study Consortium received payments for research services performed by Dr. Friedman from Adamas, Axcella, Biogen, Crossject, CuroNZ, Engage Pharmaceuticals, Eisai, GW Pharmaceuticals, Pfizer, SK Life Science, Takeda, Xenon, and Zynerba. He has also served as a paid consultant for Eisai and Penumbra. He has received honorarium from Neuropace, Inc. He has received travel support from Medtronics and the Epilepsy Foundation. He receives research support from the Centers for Disease Control and Prevention, NINDS, Epilepsy Foundation, Empatica, Epitel, UCB, Inc, and Neuropace. He serves on the scientific advisory board for Receptor Life Sciences. He holds equity interests in Neuroview Technology and Receptor Life Sciences. Maromi Nei, Kingman Strohl, Catherine Scott, Brian K Gehlbach, Norma J. Hupp, Jaison Hampson, Nassim Shafiabadi, Xiuhe Zhao, Victoria Reick-Mitrisin, Stephan Schuele, Jennifer Ogren, Ronald M. Harper, Beate Diehl, Lisa M. Bateman reports no disclosures. Philippe Ryvlin reports no disclosures, and Guo-Qiang Zhang report no disclosures. Go to Neurology.org/N for full disclosures.