Electrocardiographic and oximetric changes during partial complex and generalized seizures

Electrocardiographic and oximetric changes during partial complex and generalized seizures

Epilepsy Research (2011) 95, 237—245 journal homepage: www.elsevier.com/locate/epilepsyres Electrocardiographic and oximetric changes during partial...

292KB Sizes 0 Downloads 6 Views

Epilepsy Research (2011) 95, 237—245

journal homepage: www.elsevier.com/locate/epilepsyres

Electrocardiographic and oximetric changes during partial complex and generalized seizures Brian D. Moseley a,∗, Elaine C. Wirrell b,1, Katherine Nickels b,2, Jonathan N. Johnson c,3, Michael J. Ackerman c,d,e,4, Jeffrey Britton f,5 a

Department of Neurology, Mayo Clinic, 200 First Street SW, Rochester, MN 55905, USA Division of Child and Adolescent Neurology, Department of Neurology, Mayo Clinic, 200 First Street SW, Rochester, MN 55905, USA c Division of Pediatric Cardiology, Department of Pediatrics, Mayo Clinic, 200 First Street SW, Rochester, MN 55905, USA d Division of Cardiovascular Diseases, Department of Medicine, Mayo Clinic, 200 First Street SW, Rochester, MN 55905, USA e Department of Molecular Pharmacology & Experimental Therapeutics, Mayo Clinic, 200 First Street SW, Rochester, MN 55905, USA f Division of Epilepsy, Department of Neurology, Mayo Clinic, 200 First Street SW, Rochester, MN 55905, USA b

Received 6 October 2010; received in revised form 4 April 2011; accepted 12 April 2011 Available online 10 May 2011

KEYWORDS Epilepsy; Hypoxemia; Tachycardia; QTc; SUDEP

Summary Significant autonomic changes occur during seizures and may be related to sudden unexplained death in epilepsy (SUDEP). Accordingly, we performed a study to determine the prevalence of heart rate, QTc, and oximetric changes during seizures and analyzed their association with SUDEP risk factors. We analyzed 218 seizures from 76 patients. Ictal sinus tachycardia occurred in 57% of seizures and was associated with ≥3 failed AEDs (p = 0.001), generalized seizures (p < 0.001), and normal brain MRI (p = 0.04). Ictal sinus bradycardia was rare, occurring in 2% of seizures. Ictal bradycardia was associated with seizure clustering (p = 0.028) and reported history of ≥50 seizures/month (p = 0.01). Depending on the correction formula utilized for calculating QTc, clinically significant ictal QTc prolongation (≥460 ms for children



Corresponding author. Tel.: +1 507 202 6014; fax: +1 507 266 4752. E-mail addresses: [email protected] (B.D. Moseley), [email protected] (E.C. Wirrell), [email protected] (K. Nickels), [email protected] (J.N. Johnson), [email protected] (M.J. Ackerman), [email protected].edu (J. Britton). 1 Tel.: +1 507 266 0774; fax: +1 507 284 0727. 2 Tel.: +1 507 266 0774; fax: +1 507 284 0727. 3 Tel.: +1 507 266 0676; fax: +1 507 284 3968. 4 Tel.: +1 507 284 8612; fax: +1 507 284 3757. 5 Tel.: +1 507 774 4458; fax: +1 507 266 4752. 0920-1211/$ — see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.eplepsyres.2011.04.005

238

B.D. Moseley et al. ≤13 years, ≥470 ms for males age >13, and ≥480 ms for females age >13) occurred in 4.8—16.2% of seizures, while ictal QTc prolongation ≥500 ms occurred in 2.9—16.2%. Ictal QTc shortening ≤340 ms was observed in 3.8—4.8% of seizures. Ictal hypoxemia occurred in 25% of seizures and was associated with normal MRI (p = 0.01), longer seizure duration (p = 0.049), and ictal tachycardia (p = 0.003). Such findings may have implications for understanding the pathogenesis of SUDEP. © 2011 Elsevier B.V. All rights reserved.

Introduction SUDEP has an estimated incidence of 0.35—1.8 per 1000 patient-years (Ficker et al., 1998; Hughes, 2009). Higher incidences (3—9/1000 patient-years) have been reported in patients with intractable epilepsy (Tomson et al., 2005). Potential risk factors for SUDEP include intractable seizures (particularly generalized tonic-clonic), early age of seizure onset, prolonged duration of epilepsy, high seizure frequency, use of multiple antiepileptic drugs (AEDs), subtherapeutic AED levels, mental retardation, young age, and normal neuroimaging (Asadi-Pooya and Sperling, 2009; Degiorgio et al., 2010; Nashef et al., 2007; Nei and Hays, 2010; Surges et al., 2010a; Vlooswijk et al., 2007; Walczak et al., 2001). The underlying mechanisms of SUDEP are unknown, although cardiac and respiratory mechanisms are considered most likely. Previously reported ictal electrocardiographic abnormalities include tachycardia, repolarization anomalies, atrial fibrillation, and bradycardia (Keilson et al., 1989; Nei et al., 2000; Opherk et al., 2002; Surges et al., 2009). More recently, it has been theorized that ictal QT interval changes could precipitate lethal cardiac arrhythmias (Brotherstone et al., 2010; Johnson and Ackerman, 2009; Surges et al., 2009). Lengthening of the QT interval is a known risk factor for the development of torsades de pointes (a polymorphic form of ventricular tachycardia), while shortening of the QT interval may facilitate a reentrant ventricular tachycardia (Johnson and Ackerman, 2009; Surges et al., 2009). Potentially fatal decreases in cerebral oxygen could result from ictal hypoxemia and respiratory suppression (Bateman et al., 2008; Seyal and Bateman, 2009). The purpose of our study was to determine the prevalence and risk factors for ictal tachycardia, bradycardia, QTc changes, and hypoxemia in monitored patients with complex partial and generalized convulsive seizures, and to correlate those changes with identified SUDEP risk factors.

Methods Patients The medical records of all children (age 1 month to 18 years) and adults admitted to the pediatric and adult Epilepsy Monitoring Units (EMUs) at Mayo Clinic Rochester between 1/1/2009 and 2/12/2010 were evaluated for eligibility. Inclusion criteria included clinically suspected partial complex or generalized convulsive seizures. A total of 102 adults and 55 children were recruited prospectively. This number was limited secondary to the small number of specialized heart rate and pulse oximetry monitors at our disposal.

Informed consent was obtained from all recruited patients. For children under age 13 and older patients with cognitive impairment, consent was obtained from a parent or guardian. Written assent was also obtained from children cognitively mature enough to understand the nature of our study. The protocol was approved by the Mayo Clinic IRB.

Study design All recruited patients were evaluated with computerassisted continuous 30-channel scalp EEG. The International 10—20 system was used for electrode placement. The heart rate and digital pulse oximetry values for all study recruits were recorded utilizing IntelliVue Patient Monitors (Philips Healthcare, Boeblingen, Germany). The IntelliVue Patient Monitors recorded these data points every 60 s. A standard precordial single channel electrocardiogram (ECG) was also recorded continuously on the EEG. Children admitted to our Pediatric EMU were evaluated with additional pulse oximetry which is displayed every second on the EEG. Patients were excluded if they did not experience any clinical events during their EMU admission, experienced only psychogenic nonepileptic spells, or experienced only seizure types other than complex partial and generalized convulsive seizures. Patients missing both ECG and oximetry data secondary to excessive artifact obscuring interpretation and/or dislodgement of their ECG leads and pulse oximeters during all recorded seizures were also excluded. For patients with analyzable ECG and/or pulse oximetry data, clinical and electrophysiologic data were reviewed by the authors. All patients were screened for potential SUDEP risk factors identified in previous studies. These included prolonged duration of epilepsy (≥30 years), higher number of failed AEDs (≥3), mental retardation/profound developmental delay, primary/secondarily generalized seizures, seizure clustering (≥3 seizures per 24 h of EMU monitoring), and normal neuroimaging (Beran et al., 2004; Bird et al., 1997; Dasheiff and Dickinson, 1986; Degiorgio et al., 2010; Hitiris et al., 2007; Jick et al., 1992; Langan et al., 2005; McKee and Bodfish, 2000; Nei and Hays, 2010; Nei et al., 2004; Nilsson et al., 1999; Opeskin et al., 2000; So et al., 2000; Surges et al., 2010a, 2009; Timmings, 1993; Walczak et al., 2001). We also recorded all AEDs prescribed at the time of monitoring. We monitored for the use of medications known to have electrocardiographic effects including beta blockers and drugs with definite, possible, or conditional risk of prolonging QTc (Centre for Education and Research on Therapeutics, 2010). The usage of selective serotonin reuptake inhibitors (SSRIs) was examined, as these medications have recently been shown to correlate with the reduced severity of ictal hypoxemia in patients with partial seizures (Bateman et al., 2010a). To investigate potential

Electrocardiographic and oximetric changes during seizures Table 1

Abnormal MRI brain findings in our cohort; n = 49.

Imaging finding

Frequency

Mesial temporal sclerosis 24 (49%) Focal cortical dysplasia 5 (10%) Encephalomalacia 4 (8%) Gliosis 2 (4%) Ganglioglioma 2 (4%) Schizencephaly 1 (2%) Polymicrogyria 1 (2%) Angioblastoma 1 (2%) Dysembryoplastic neuroepithial tumor 1 (2%) Hypothalamic hamartoma 1 (2%) Medulloblastoma 1 (2%) Amygdala cyst 1 (2%) Middle cerebral artery aneurysm 1 (2%) Atrophic corpus callosum 1 (2%) Corpus callosotomy 1 (2%) Abnormal hemosiderin deposition 1 (2%) Focal T2 hyperintensity of undetermined etiology1 (2%)

239 the Bazett formula and at least one other formula were recorded. Ictal QTc shortening by ≥60 ms was calculated by subtracting the minimum recorded ictal QTc value from the minimum recorded pre-ictal QTc value using the Bazett formula and one other formula. Clinically significant QT prolongation pre-ictally or during seizures was defined as a QTc ≥ 460 ms for children age 13 years or younger, ≥470 ms for males age >13 years, and ≥480 ms for females age >13 years using the Bazett formula and at least one other formula (Johnson and Ackerman, 2009). Seizures associated with marked QT prolongation (QTc ≥ 500 ms), markedly short QT intervals (QTc ≤ 340 ms), and profoundly abbreviated QT intervals (QTc ≤ 300 ms) as calculated by the Bazett formula and at least one other formula were also recorded. To ensure reliability of QT interval measurement, the preictal and ictal QTc values for ten seizures from different patients were calculated by a cardiologist blinded to the first author’s measurement. Comparison of these values yielded an intraclass correlation coefficient of 0.77, indicating satisfactory agreement between observers.

Statistical analysis cardiac co-morbidities acting as cofounding variables, we documented histories of hypertension, coronary artery disease, past myocardial infarctions, congestive heart failure, atrial fibrillation, other cardiac arrhythmias, diabetes mellitus (types 1 and 2), hypercholesterolemia, past/present smoking, and past/present alcoholism. For patients having greater than 10 recorded seizures, only the first 10 seizures were assessed. The EEG recordings were reviewed by the authors. Seizure duration (defined as time in seconds from electrographic seizure onset to offset), onset localization (temporal, extratemporal, or generalized) and lateralization (right, left, or indeterminate) were recorded. Oxygen saturation was documented using the lowest value recorded in the 2 min interval prior to electrographic onset of seizures and the lowest during the seizure. Oxygen saturation data was recorded every second for pediatric patients and at minute intervals for adults. Preictal heart rate data was calculated utilizing the mean of three sequential R—R intervals measured at minute intervals during the 2 min period prior to electrographic onset. Ictal heart rates were calculated at seizure onset, 15 s, 30 s, 45 s, 60 s, 90 s, 120 s, and minute intervals thereafter until offset. We defined ictal tachycardia as a heart rate greater than the 98th percentile for age and ictal bradycardia as a heart rate less than the 2nd percentile for age (Rowe, 1987). Ictal bradycardia was not diagnosed if bradycardia was also seen in the pre-ictal period. QTc values were calculated at minute intervals for the two minutes preceding electrographic onset. Ictal QTc values were recorded at identical time intervals as heart rate values. The mean of three sequential QT intervals and their preceding R—R intervals were used to calculate ictal and pre-ictal QTc values using the Bazett-, Hodges-, Fridericia, and Framingham-heart rate correction formulas (Bazett, 1920; Funck-Brentano and Jaillon, 1993; Milic et al., 2006; Prasad et al., 2007). Ictal QTc lengthening was measured by subtracting the mean maximum ictal QTc value from the maximum pre-ictal QTc value. Ictal QTc values that were lengthened by more than 60 ms over the pre-ictal QTc by

Data entry and statistical analysis were performed using SPSS Version 16.0 (SPSS Inc., Chicago, IL, U.S.A.). To evaluate if variables (including the SUDEP risk factors identified previously) were correlated with heart rate, QTc, and/or oximetric changes, we utilized chi-square analysis (Pearson’s chi-square) for categorical data and independentsamples Student’s t-test (2 tailed) for continuous data (unless otherwise specified). P-values <0.05 were considered statistically significant. Variables with significant correlation were subsequently entered into a linear regression model (provided more than one variable had a significant correlation with the electrocardiographic or oximetric disturbance studied).

Results At least one seizure with heart rate, QTc, and/or pulse oximetry data was available in 76 patients (51 adults, 25 children). The mean number of current AEDs on admission was 2.3 ± 1.0 (range 0—6). The majority of patients (67/76, 88%) had failed at least 3 AEDs prior to admission. Forty nine of 74 patients (66%) had imaging abnormalities on head MRI (see Table 1). A minority (17/76 or 22.4%) had neurological examination abnormalities. The characteristics of the 15 patients with primary generalized seizures (including primary generalized tonic clonic and generalized tonic seizures) differed slightly from those of the 61 patients with partial onset seizures (with or without secondary generalization). The patients in our cohort with primary generalized seizures were more likely to have mental retardation/developmental delay (7/15, 47% versus 10/61, 16%, p = 0.012), neurologic examination abnormalities (7/15, 47% versus 10/61, 16%, p = 0.012), and ≥50 seizures per month (7/15, 47% versus 10/61, 16%, p = 0.012). The remaining patient characteristics are noted in Table 2. A total of 218 seizures were analyzed. The mean number of seizures with analyzable heart rate, QTc, and/or oximetry data per subject was 2.9 ± 2.4 (range 1—10). Details of

240 Table 2

B.D. Moseley et al. Cohort demographics and characteristic. 26.9 ± 15.6 (range 2—61) 59.2% male, 40.8 female 80.3% localization related, 19.7% generalized 67.1% symptomatic, 32.9% idiopathic/cryptogenic 37% daily, 40% less than daily to weekly, 17% less than weekly to monthly, 7% less than monthly 7.9%

Age (years) Gender Epilepsy syndrome

Epilepsy etiology

Seizure frequency

Previous/current use of ketogenic diet Previous VNS implantation Previous epilepsy (resective) surgery

14.5% 5.3%

seizure incidence, duration, lateralization, and localization are summarized in Table 3.

Ictal tachycardia Ictal ECG data was available in 75/76 patients and 217/218 seizures. Ictal tachycardia was noted in 124/217 (57%) seizures in 57/75 (76%) patients. Ictal tachycardia was documented in 66/138 (48%) partial complex seizures, 33/41 (81%) secondary generalized tonic-clonic (GTC) seizures, 10/12 (83%) primary GTC seizures, and 15/26 (58%) generalized tonic seizures. Ictal tachycardia occurred more often with generalized seizures than non-generalized complex partial seizures (58/79 generalized, 73% versus 66/138 complex partial, 48%; p < 0.001). The mean number of failed AEDs was higher in the ictal tachycardia group (6.1 ± 3.0 versus 4.4 ± 2.5; p = 0.025). Patients who had failed ≥3 AEDs were significantly more likely to experience ictal tachycardia (54/66 who failed ≥3 AEDs, 82% versus 3/9 who failed <3 AEDs, 33%; p = 0.001). Ictal tachycardia was recorded more

Table 3

frequently in patients with normal versus abnormal neuroimaging (23/25, 92% versus 34/48, 71%; p = 0.04). There was a trend towards increased incidence of ictal tachycardia and seizure clustering (30/35 with seizure clustering, 86% versus 27/40 without seizure clustering, 68%, p = 0.065). There were no significant correlations between ictal tachycardia and mental retardation (p = 0.55) or duration of epilepsy ≥30 years (p = 0.21). Ictal tachycardia was associated with temporal localization, but only in partial seizures without secondary generalization (55/107, 51% versus 7/25, 28% extratemporal; p = 0.035). This was no longer significant when including partial seizures which secondarily generalized. There was no association between ictal tachycardia and seizure lateralization. When entered into a linear regression model, ictal tachycardia continued to be significantly associated with ≥3 failed AEDs (beta 0.406, 95% confidence interval (CI) 0.238—0.783, p < 0.001) and seizure generalization (beta −0.242, 95% CI −0.43 to −0.111, p = 0.001). However, the actual number of failed AEDs (beta −0.088, p = 0.48) and absence of MRI abnormalities (beta −0.206, p = 0.058) were no longer predictive of ictal tachycardia.

Ictal bradycardia Ictal bradycardia was rare, occurring in 4/75 patients (5%) and 4/217 seizures (2%). Ictal bradycardia was associated with beta blocker usage (1/3, 33% versus 3/72, 4%; p = 0.028), drugs with the potential to prolong QT (3/21, 14% versus 1/54, 2%; p = 0.031), seizure clustering (4/35, 11.4% versus 0/40, 0%; p = 0.028), and reported history of ≥50 seizures/month (3/17, 18% versus 1/58, 2%; p = 0.01). Ictal bradycardia was not significantly associated with a particular localization or lateralization, mental retardation, duration of epilepsy ≥30 years, ≥3 failed AEDs, or normal brain MRI. Although ictal bradycardia was observed more frequently in complex partial seizures without generalization, this finding was not significant. When entered into a linear regression model, beta blocker usage (beta 0.29, 95% CI 0.087—0.580, p = 0.009) and ≥50 seizures/month (beta 0.329, 95% CI 0.061—0.292, p = 0.003) remained significantly associated with ictal bradycardia. However, usage of drugs known to prolong QT (beta 0.194, p = 0.074) and

Seizure localization, lateralization, and duration.

Seizure type

Number of subjects with seizure type

Mean seizure duration (s)

Seizure lateralization

Seizure regional localization

Partial complex

139/218 (64%)

126.9 ± 238.5 (range 8—1981)

60.2% right, 39.8% left

Secondarily GTC

41/218 (19%)

145.7 ± 155.8 (range 57—1013)

45.9% right, 54.1% left

Primary GTC

12/218 (6%)

N/A

Generalized Tonic

26/218 (12%)

103.1 ± 93.2 (range 9—356) 41.8 ± 38.9 (range 3—151)

81.2% temporal, 18.8% extratemporal 52.5% temporal, 47.5% extratemporal N/A

N/A

N/A

Electrocardiographic and oximetric changes during seizures seizure clustering (beta 0.167, p = 0.152) were no longer predictive.

Seizures and cardiac repolarization (ictal QTc changes) Analyzable pre-ictal and ictal QTc data were available in 43/76 patients and 105/218 recorded seizures. The mean maximum ictal QTc was significantly longer than the mean maximum pre-ictal QTc utilizing three of the correction formulas (Bazett maximum ictal 452 ± 56 ms, maximum preictal 429 ± 48 ms, Paired samples T test 2 tailed p < 0.001; Fridericia 412 ± 53 ms versus 405 ± 51 ms, p = 0.003; Hodges 431 ± 44 ms versus 411 ± 44 ms, p < 0.001). There was a trend towards QTc lengthening using the Framingham formula, but this did not reach statistical significance (408 ± 47 ms versus 405 ± 47 ms, p = 0.14). Ictal QTc lengthening by ≥60 ms was observed in 14/105 seizures (13%) using the Bazett formula alone, and in 3/105 seizures (2.9%) using the criterion of ≥60 ms in both the Bazett formula and at least one of the other 3 heart rate correction formulas. All three seizures occurred in partial complex seizures of right hemispheric onset, but this was not statistically significant. Ictal-associated clinically significant QTc prolongation was noted in 17/105 seizures (16.2%) using Bazett’s formula, and in 5/105 seizures (4.8%) using the criterion of Bazett’s plus at least one other formula. All 5 of these seizures occurred in adults and were non-generalized partial complex seizures of temporal lobe onset. Four of these seizures were lateralized to the right. However, these associations were not statistically significant. New ictal-associated QTc prolongation ≥500 ms was observed in 17/105 seizures (16.2%) using Bazett’s formula and 3/105 seizures (2.9%) using the criterion of Bazett’s plus at least one other formula. Ictal QTc prolongation ≥500 ms occurred only in adult patients with partial onset seizures of temporal lobe origin; however, this was not statistically significant. The only SUDEP risk factor significantly associated with QTc lengthening was seizure clustering; this was associated with ictal QTc lengthening ≥60 ms (3/19, 16% versus 0/24, 0%; p = 0.044). Ictal QTc shortening ≥60 ms was observed in 9/105 (8.6%) seizures using Bazett’s formula, and 8/105 (7.6%) using the criterion of Bazett’s plus at least one other formula. This was only associated with female gender (5/15 female patients versus 2/28 male patients, p = 0.027). New ictal markedly short QT intervals (QTc ≤ 340 ms) were observed in 5/105 (4.8%) seizures using Bazett’s formula, and 4/105 (3.8%) seizures using the criterion of Bazett’s plus at least one other formula. Neither ictal QTc shortening ≥60 ms nor new ictal markedly short QT intervals were associated with seizure lateralization, localization, or examined SUDEP risk factors. No seizures with profoundly attenuated QT intervals (QTc ≤ 300 ms) were recorded using the Bazett formula. New ictal QTc intervals ≤340 ms were associated with usage of Rufinamide (1/2, 50% versus 2/41, 5%; p = 0.014). However, ictal QTc shortening ≥60 ms was not associated with usage of Rufinamide (p = 0.523). No QTc changes (shortening or prolongation) were associated with concurrent usage of other drugs known to affect the QT interval.

241

Ictal hypoxemia Pulse oximetry data was available in 51/76 patients and 121/218 seizures. The remaining patients/seizures lacked usable pulse oximetry data secondary to excess movement artifact and dislodgement of oximeters. Ictal hypoxemia was seen in 18/51 patients (35%) and 30/121 seizures (25%). Of seizures with ictal desaturation, a lowest recorded oxygen saturation in the 80—89.9% range was seen in 20/30 seizures (67%), 70—79.9% in 8/30 (27%), 60—69.9% in 1/30 (3%) and less than 60% in 1/30 (3%). Ictal hypoxemia was significantly more common in patients who also had ictal tachycardia (17/34, 50% versus 1/16, 6% without ictal tachycardia; p = 0.003). Ictal hypoxemia was correlated with longer seizure duration (mean duration of seizures with ictal hypoxemia was 173 ± 162 s compared to 101 ± 196 s in seizures without; p = 0.049). Patients without brain MRI abnormalities were more likely to experience ictal hypoxemia than those with abnormalities (11/18, 61% versus 7/31, 23%; p = 0.01). Ictal hypoxemia was not associated with seizure localization, lateralization, usage of selective serotonin reuptake inhibitors, or other examined SUDEP risk factors. When entered into a linear regression model, ictal tachycardia (beta 0.316, 95% CI 0.042—0.631, p = 0.026) and lack of MRI abnormalities (beta −0.281, 95% CI −0.577 to −0.004, p = 0.047) remained statistically significant. However, prolonged seizure duration was no longer predictive (beta −0.173, p = 0.093).

Ictal electrocardiographic and oximetric changes in adults versus children When adulthood was defined as an age ≥18 years, no significant differences were noted in the frequencies of ictal autonomic disturbances in adults versus children. This included ictal tachycardia (p = 0.566), ictal bradycardia (p = 0.467), ictal QTc shortening ≥60 ms (p = 0.702), and new ictal markedly short QT intervals (p = 0.231). Although all forms of QTc lengthening/prolongation were only observed in adults, this did not reach statistical significance. This included ictal QTc lengthening ≥60 ms (3/3, 100% patients with such lengthening were adults versus 25/40, 63% patients without such lengthening, p = 0.189), clinically significant QTc prolongation (4/4, 100% versus 24/39, 62%, p = 0.124), and QTc prolongation ≥500 ms (3/3, 100% versus 25/40, 63%, p = 0.189). When adulthood was expanded to include teenagers (age ≥ 13 years), a difference was noted in the frequency of ictal hypoxemia. A significantly larger percentage of patients with ictal hypoxemia (17/18, 94%) were ≥13 years of age versus those without (23/33, 70%, p = 0.04). Although new ictal markedly short QT intervals were more frequently encountered in patients <13 years old (1/3, 33% with such shortening were ≥13 years versus 32/40, 80% without, p = 0.065), this did not reach statistical significance. There were no significant differences in the frequencies of ictal tachycardia (p = 0.255), ictal bradycardia (p = 0.738), ictal QTc lengthening ≥60 ms (p = 0.323), clinically significant ictal QTc prolongation (p = 0.248), ictal QTc prolongation ≥500 ms (p = 0.323), and ictal QTc shortening ≥60 ms (p = 0.716) based on age ≥13 years.

242

The effects of cardiac co-morbidities on ictal electrocardiographic and oximetric changes The minority of our cohort (23/76 patients, 30%) had a history of one or more cardiac co-morbidities. This included 7/76 patients (9%) with hypertension, 2/76 (2.6%) with hypercholesterolemia, 1/76 (1%) with diabetes mellitus, 15/76 (20%) with past/present smoking, and 5/76 (6.6%) with past/present alcoholism. No patients in our cohort had documented histories of coronary artery disease, past myocardial infarction, congestive heart failure, atrial fibrillation, or other cardiac arrhythmia. When examined individually, none of the above cardiac co-morbidities were significantly associated with ictal autonomic changes. When compiled, a history of one or more cardiac co-morbidities failed to confer an increased risk for ictal tachycardia (p = 0.76), ictal bradycardia (p = 0.389), ictal hypoxemia (p = 0.223), or any examined ictal lengthening/shortening of QTc.

The effects of surgical pathology on ictal electrocardiographic and oximetric changes Following EMU admission, 26/76 patients in our cohort (34.2%) underwent surgical resections for treatment of refractory seizures. The majority of these patients (22/26, 84.6%) had abnormalities noted on their preadmission neuroimaging. Patients with abnormal MRIs were significantly more likely to have identifiable pathology other than nonspecific gliosis (p = 0.007). Of the 22 patients with abnormal neuroimaging, only 3 (13.6%) were noted to have nonspecific gliosis. The remaining pathologic diagnoses in these patients included mesial temporal sclerosis (13/22, 59.1%), encephalomalacia (3/22, 13.6%), ganglioglioma (WHO grade I, 2/22, 9.1%), and low grade fibrillary astrocytoma (WHO grade II, 1/22, 4.5%). Of the 4 surgical patients with normal neuroimaging, 3 (75%) had nonspecific gliosis. Only 1 (25%) was noted to have differing pathology (a focal cortical dysplasia, Taylor-type IIB). Given the low occurrence of nonspecific gliosis in our surgical cohort, we could not make significant correlations between this pathologic diagnosis and the occurrence of ictal autonomic dysfunction. However, there was a trend towards increased incidence of ictal hypoxemia in those with only gliosis (p = 0.087).

Conclusions Significant cardiac rhythm and oxygen saturation changes were common during recorded seizures in our cohort. Ictal QTc lengthening was more common than expected, and was observed more often than ictal shortening of the QT interval. Ictal tachycardia was the most common autonomic disturbance, present in a majority of recorded seizures. In contrast, ictal bradycardia was rare. This is consistent with previous literature, which has reported bradycardia in a small proportion (<1—3.7%) of recorded seizures (Britton et al., 2006; Moseley et al., 2010; Rugg-Gunn et al., 2004). Ictal hypoxemia was recorded during 25% of seizures in our cohort; this is also similar to the results

B.D. Moseley et al. of previous studies (Bateman et al., 2008; Moseley et al., 2010). Ictal tachycardia and/or hypoxemia were associated with SUDEP risk factors, including generalized seizures, larger number of failed medications (a potential marker of intractable epilepsy), and normal neuroimaging. Such findings, although indirect, suggest a potential association between the autonomic responses that result from seizures and the pathogenesis of SUDEP. Clinically significant ictal prolongation of QTc was not rare in our cohort, occurring in 4.8—16.2% of seizures depending on the correction formula utilized. It was more common in our cohort than ictal QTc shortening, which has been reported previously with generalized seizures (Surges et al., 2010b). Ictal QTc prolongation could be secondary to seizure-related cerebral deregulation, cardiorespiratory interactions from hypoxia and hypercapnia, and the release of stress hormones (Surges et al., 2010c). Pathogenic dysregulation of cardiac depolarization/repolarization has been documented in two victims whose sudden death was classified as SUDEP; one had a mutation in the RYR2-encoded cardiac ryanodine receptor/calcium release channel, while the other had a mutation in the SCN5A-encoded cardiac Nav1.5 sodium channel (Aurlien et al., 2009; Johnson et al., 2010). However, caution must be utilized when attempting to link all cases of SUDEP to cardiac depolarization/repolarization anomalies. Only one case of ventricular fibrillation progressing to asystole and death has been documented following a complex partial seizure in a monitored setting (Dasheiff and Dickinson, 1986). It is more likely such cardiac dysfunction accounts for only some cases of SUDEP, with the remainder caused by other autonomic phenomena. Although not statistically significant, the trend favoring an association between QTc lengthening and temporal lobe seizures is intriguing. It has been shown previously that patients with type 2 long QT (LQT2) syndrome secondary to mutations in the KCNH2-encoded Kv11.1 potassium channel have a predilection for epilepsy. Notably, both KCNH2 and its encoded protein are expressed in hippocampal astrocytes (Johnson et al., 2009). Given this relationship between LQT2 and epilepsy, it is not unreasonable to envision a possible ictal predisposition to torsades de pointes and resulting SUDEP. Further research into this possibility is warranted, as the risk of SUDEP in such patients could potentially be reduced with antiarrhythmic medications and/or pacemaker-defibrillator implantation (Surges et al., 2010c). Interestingly, we found that patients with normal brain MRIs had higher rates of ictal autonomic disturbances, including ictal tachycardia and hypoxemia. In a recent case control study, Surges et al. found that abnormal MRIs were associated with a lower risk of SUDEP (Surges et al., 2010a). It was hypothesized that those with lesional MRIs would be more likely to proceed to surgery and have a better outcome, as has previously been demonstrated in anterior temporal lobectomy patients (Khoury et al., 2005). In contrast, our data suggests there may be an inherent difference between lesional and non-lesional epilepsy with regards to autonomic disturbances. Such a difference could be secondary to ion channelopathies which predispose affected patients to seizures and/or cardiac rhythm disturbances. Given that such dysfunction could be amenable to treatment, further study of autonomic disturbances in non-lesional patients and their potential association with

Electrocardiographic and oximetric changes during seizures

243

channelopathies is warranted. However, caution must be taken when attempting to apply our data to directly answer this question, as we did not screen for genetic mutations nor inquire about family histories of LQT/other hereditary arrhythmias. Our study is not the first to suggest a potential contribution of ictal tachycardia to SUDEP. When comparing maximal ictal heart rates in patients who later died of definite/probable SUDEP to controls with refractory partial epilepsy, previous studies have documented significantly higher values in the SUDEP group (Nei et al., 2004). ST segment depression has been associated with higher maximal heart rate during seizures, and has been suggested as a marker for cardiac ischemia (Tigaran et al., 2003). In a recent case—control study, a greater incidence of fibrotic changes in the deep and subendocardial myocardium of SUDEP cases at autopsy versus controls (40% versus 6.6%) was reported. This fibrosis was hypothesized to be the consequence of myocardial ischemia from repetitive seizures (P-Codrea Tigaran et al., 2005). Our study is also not the first to query whether ictal hypoxemia could be a cause of sudden death given its association with SUDEP risk factors. Ictal respiratory suppression (and ultimately apnea) may occur secondary to propagation of seizure activity or as the result of an ictally mediated disturbance of central autonomic centers involved in the modulation of medullary respiratory nuclei. Ictally mediated ventilation-perfusion mismatch from pulmonary shunting and/or neurogenic pulmonary edema have also been suggested (Seyal et al., 2010). Moderate to severe pulmonary edema documented in a majority of SUDEP cases at autopsy support such a hypothesis (Leestma et al., 1989; Terrence et al., 1981). In a review of witnessed cases of SUDEP, 12 (80%) were characterized by respiratory difficulty (Langan et al., 2000). In two cases of SUDEP in a monitored setting, ictal/postictal hypoventilation was theorized to have led to hypoxemia and acidosis, ultimately resulting in cardiac failure (Bateman et al., 2010b). Such cases, in addition to our findings, support making pulse oximetry a standard for EMU evaluation. Our findings also support further research into strategies to reduce ictal oxygen desaturations. Recently, Bateman et al. reported that SSRIs are protective against ictal hypoxemia, possibly secondary to enhancement of brainstem respiratory center excitability (Bateman et al., 2010a). Although we were unable to reproduce this finding, the relatively low number of patients on SSRIs in our study (n = 10) likely limited our ability to make such a correlation. In our study, patients with documented ictal tachycardia were significantly more likely to have documented ictal hypoxemia than those without elevated heart rates. Such findings could provide further support for the fatal coincidence (‘‘perfect storm’’) hypothesis of SUDEP. It has been argued that sudden death is likely the result of several precipitating factors coming together during the peri-ictal period, a hypothesis supported by animal models (Surges et al., 2009). In hemispherectomized rats, it was demonstrated that cardiac arrhythmias only became life threatening when they occurred in the setting of hypoxia and hypercapnia (Mameli et al., 2006). In identifying patients at greatest risk for SUDEP, it may ultimately prove more use-

ful to look at the concurrence of multiple autonomic events rather than any single disturbance. Our study was not without limitations. Our study population was comprised of those adults and children referred to a tertiary center for specialized video EEG monitoring. The inclusion of children limited our ability to comment more definitively on the effects of prolonged duration of epilepsy (≥30 years) on ictal autonomic disturbances. Given the intermittent nature of how we recorded heart rate, pulse oximetry, and QTc data using available technology, it is possible we underreported the incidence of autonomic disturbances in our cohort. This was particularly true for adults and hypoxemia. Given the limitations of the IntelliVue Patient Monitors (and the lack of continuous pulse oximetry in our adult EMU), we could only record oxygen saturations at minute intervals for adults. Given that we did not measure chest movement or end-tidal CO2 , we were unable to distinguish between central and obstructive apnea. When our electroencephalogram files are archived, EEG/ECG data is only saved from 2 min prior to electrographic seizure onset to 2 min following seizure offset. Therefore, it was impossible to obtain baseline heart rate and QTc values for a preictal time interval greater than 2 min. It is possible our preictal measurements may have been early ictal measurements for some patients, as true seizure onset can occur prior to visible scalp electrographic onset. If present, this may have resulted in underestimating QTc prolongation in our cohort. The usage of a single precordial ECG lead prevented reliable assessment of ST elevation/depression. The percentages of ictal QTc prolongation reported using only the Bazett formula could be overestimates; previous studies have shown that this particular heart rate correction formula can overestimate QTc at rapid heart rates (Benatar and Decraene, 2001; Funck-Brentano and Jaillon, 1993; Milic et al., 2006; Prasad et al., 2007). However, our reporting of ictal QTc abnormalities calculated with both Bazett’s and at least one other correction formula compensated for this. It is possible such reporting actually underestimated the frequency of QTc prolongation, as one of the other formulas (Fridericia) has been associated with such error (Funck-Brentano and Jaillon, 1993).

Acknowledgements The authors would like to express their utmost gratitude to Teresa Peters in our Department of Engineering. Without her help with the IntelliVue Patient Monitors and RDE Viewer software, this study would not have been possible. The authors would also like to thank our EEG technicians (Susan Senjem, Cindy Nelson, Jean Varner, Randy Berge, Eric Marshall, Charlene Harstad, Jeffrey Goihl, and Judith Johnson) for their constant help with patient recruitment, clipping of EEG files, and retrieval of archived files.

References Asadi-Pooya, A.A., Sperling, M.R., 2009. Clinical features of sudden unexpected death in epilepsy. J. Clin. Neurophysiol. 26, 297—301.

244 Aurlien, D., Leren, T.P., Tauboll, E., Gjerstad, L., 2009. New SCN5A mutation in a SUDEP victim with idiopathic epilepsy. Seizure 18, 158—160. Bateman, L.M., Li, C.S., Lin, T.C., Seyal, M., 2010a. Serotonin reuptake inhibitors are associated with reduced severity of ictal hypoxemia in medically refractory partial epilepsy. Epilepsia, doi:10.1111/j.1528-1167.2010.02594.x, published online: 14-May-2010. Bateman, L.M., Li, C.S., Seyal, M., 2008. Ictal hypoxemia in localization-related epilepsy: analysis of incidence, severity and risk factors. Brain 131, 3239—3245. Bateman, L.M., Spitz, M., Seyal, M., 2010b. Ictal hypoventilation contributes to cardiac arrhythmia and SUDEP: report on two deaths in video-EEG-monitored patients. Epilepsia 51, 916—920. Bazett, H., 1920. An analysis of the time-relations of electrocardiograms. Heart, 353—370. Benatar, A., Decraene, T., 2001. Comparison of formulae for heart rate correction of QT interval in exercise ECGs from healthy children. Heart 86, 199—202. Beran, R.G., Weber, S., Sungaran, R., Venn, N., Hung, A., 2004. Review of the legal obligations of the doctor to discuss Sudden Unexplained Death in Epilepsy (SUDEP) — a cohort controlled comparative cross-matched study in an outpatient epilepsy clinic. Seizure 13, 523—528. Bird, J., Dembny, K.A.T., Sandeman, D., Butler, S., 1997. Sudden unexplained death in epilepsy: an intracranially monitored case. Epilepsia 38, S52—S56. Britton, J.W., Ghearing, G.R., Benarroch, E.E., Cascino, G.D., 2006. The ictal bradycardia syndrome: localization and lateralization. Epilepsia 47, 737—744. Brotherstone, R., Blackhall, B., McLellan, A., 2010. Lengthening of corrected QT during epileptic seizures. Epilepsia 51, 221—232. Centre for Education and Research on Therapeutics, 2010. QT drug lists by risk groups. http://www.azcert.org/medical-pros/druglists/drug-lists.cfm (accessed 13.01.10). Dasheiff, R.M., Dickinson, L.J., 1986. Sudden unexpected death of epileptic patient due to cardiac arrhythmia after seizure. Arch. Neurol. 43, 194—196. Degiorgio, C.M., Miller, P., Meymandi, S., Chin, A., Epps, J., Gordon, S., Gornbein, J., Harper, R.M., 2010. RMSSD, a measure of vagusmediated heart rate variability, is associated with risk factors for SUDEP: the SUDEP-7 Inventory. Epilepsy Behav. 19 (1), 78—81. Ficker, D.M., So, E.L., Shen, W.K., Annegers, J.F., O’Brien, P.C., Cascino, G.D., Belau, P.G., 1998. Population-based study of the incidence of sudden unexplained death in epilepsy. Neurology 51, 1270—1274. Funck-Brentano, C., Jaillon, P., 1993. Rate-corrected QT interval: techniques and limitations. Am. J. Cardiol. 72, 17B—22B. Hitiris, N., Suratman, S., Kelly, K., Stephen, L.J., Sills, G.J., Brodie, M.J., 2007. Sudden unexpected death in epilepsy: a search for risk factors. Epilepsy Behav. 10, 138—141. Hughes, J.R., 2009. A review of sudden unexpected death in epilepsy: prediction of patients at risk. Epilepsy Behav. 14, 280—287. Jick, S.S., Cole, T.B., Mesher, R.A., Tennis, P., Jick, H., 1992. Sudden unexplained death in young persons with primary epilepsy. Pharmacoepidemiol. Drug Safety 1, 59—64. Johnson, J.N., Ackerman, M.J., 2009. QTc: how long is too long? Br. J. Sports Med. 43, 657—662. Johnson, J.N., Hofman, N., Haglund, C.M., Cascino, G.D., Wilde, A.A., Ackerman, M.J., 2009. Identification of a possible pathogenic link between congenital long QT syndrome and epilepsy. Neurology 72, 224—231. Johnson, J.N., Tester, D.J., Bass, N.E., Ackerman, M.J., 2010. Cardiac channel molecular autopsy for sudden unexpected death in epilepsy. J. Child Neurol. 25, 916—921.

B.D. Moseley et al. Keilson, M.J., Hauser, W.A., Magrill, J.P., 1989. Electrocardiographic changes during electrographic seizures. Arch. Neurol. 46, 1169—1170. Khoury, J.S., Winokur, R.S., Tracy, J.I., Sperling, M.R., 2005. Predicting seizure frequency after epilepsy surgery. Epilepsy Res. 67, 89—99. Langan, Y., Nashef, L., Sander, J.W., 2000. Sudden unexpected death in epilepsy: a series of witnessed deaths. J. Neurol. Neurosurg. Psychiatry 68, 211—213. Langan, Y., Nashef, L., Sander, J.W., 2005. Case—control study of SUDEP. Neurology 64, 1131—1133. Leestma, J.E., Walczak, T., Hughes, J.R., Kalelkar, M.B., Teas, S.S., 1989. A prospective study on sudden unexpected death in epilepsy. Ann. Neurol. 26, 195—203. Mameli, O., Caria, M.A., Pintus, A., Padua, G., Mameli, S., 2006. Sudden death in epilepsy: an experimental animal model. Seizure 15, 275—287. McKee, J.R., Bodfish, J.W., 2000. Sudden unexpected death in epilepsy in adults with mental retardation. Am. J. Ment. Retard. 105, 229—235. Milic, M., Bao, X., Rizos, D., Liu, F., Ziegler, M.G., 2006. Literature review and pilot studies of the effect of QT correction formulas on reported beta2-agonist-induced QTc prolongation. Clin. Ther. 28, 582—590. Moseley, B.D., Nickels, K., Britton, J., Wirrell, E., 2010. How common is ictal hypoxemia and bradycardia in children with partial complex and generalized convulsive seizures? Epilepsia 51, 1219—1224. Nashef, L., Hindocha, N., Makoff, A., 2007. Risk factors in sudden death in epilepsy (SUDEP): the quest for mechanisms. Epilepsia 48, 859—871. Nei, M., Hays, R., 2010. Sudden unexpected death in epilepsy. Curr. Neurol. Neurosci. Rep. 10, 319—326. Nei, M., Ho, R.T., Abou-Khalil, B.W., Drislane, F.W., Liporace, J., Romeo, A., Sperling, M.R., 2004. EEG and ECG in sudden unexplained death in epilepsy. Epilepsia 45, 338—345. Nei, M., Ho, R.T., Sperling, M.R., 2000. EKG abnormalities during partial seizures in refractory epilepsy. Epilepsia 41, 542—548. Nilsson, L., Farahmand, B.Y., Persson, P.G., Thiblin, I., Tomson, T., 1999. Risk factors for sudden unexpected death in epilepsy: a case—control study. Lancet 353, 888—893. Opeskin, K., Harvey, A.S., Cordner, S.M., Berkovic, S.F., 2000. Sudden unexpected death in epilepsy in Victoria. J. Clin. Neurosci. 7, 34—37. Opherk, C., Coromilas, J., Hirsch, L.J., 2002. Heart rate and EKG changes in 102 seizures: analysis of influencing factors. Epilepsy Res. 52, 117—127. P-Codrea Tigaran, S., Dalager-Pedersen, S., Baandrup, U., Dam, M., Vesterby-Charles, A., 2005. Sudden unexpected death in epilepsy: is death by seizures a cardiac disease? Am. J. Forensic Med. Pathol. 26, 99—105. Prasad, S., Furr, A.J., Zhang, S., Ball, S., Allen, A.J., 2007. Baseline values from the electrocardiograms of children and adolescents with ADHD. Child Adolesc. Psychiatry Ment. Health 1, 11. Rowe, P.C., 1987. The Harriet Lane Handbook. Year Book Medical Publishers, Chicago. Rugg-Gunn, F.J., Simister, R.J., Squirrell, M., Holdright, D.R., Duncan, J.S., 2004. Cardiac arrhythmias in focal epilepsy: a prospective long-term study. Lancet 364, 2212—2219. Seyal, M., Bateman, L.M., 2009. Ictal apnea linked to contralateral spread of temporal lobe seizures: intracranial EEG recordings in refractory temporal lobe epilepsy. Epilepsia 50, 2557—2562. Seyal, M., Bateman, L.M., Albertson, T.E., Lin, T.C., Li, C.S., 2010. Respiratory changes with seizures in localization-related epilepsy: analysis of periictal hypercapnia and airflow patterns. Epilepsia, doi:10.1111/j.1528-1167.2009.02518.x, published online: 12-February-2010.

Electrocardiographic and oximetric changes during seizures

245

So, E.L., Sam, M.C., Lagerlund, T.L., 2000. Postictal central apnea as a cause of SUDEP: evidence from near-SUDEP incident. Epilepsia 41, 1494—1497. Surges, R., Adjei, P., Kallis, C., Erhuero, J., Scott, C.A., Bell, G.S., Sander, J.W., Walker, M.C., 2010a. Pathologic cardiac repolarization in pharmacoresistant epilepsy and its potential role in sudden unexpected death in epilepsy: a case—control study. Epilepsia 51, 233—242. Surges, R., Scott, C.A., Walker, M.C., 2010b. Enhanced QT shortening and persistent tachycardia after generalized seizures. Neurology 74, 421—426. Surges, R., Taggart, P., Sander, J.W., Walker, M.C., 2010c. Too long or too short? New insights into abnormal cardiac repolarization in people with chronic epilepsy and its potential role in sudden unexpected death. Epilepsia 51, 738—744. Surges, R., Thijs, R.D., Tan, H.L., Sander, J.W., 2009. Sudden unexpected death in epilepsy: risk factors and potential pathomechanisms. Nat. Rev. Neurol. 5, 492—504.

Terrence, C.F., Rao, G.R., Perper, J.A., 1981. Neurogenic pulmonary edema in unexpected, unexplained death of epileptic patients. Ann. Neurol. 9, 458—464. Tigaran, S., Molgaard, H., McClelland, R., Dam, M., Jaffe, A.S., 2003. Evidence of cardiac ischemia during seizures in drug refractory epilepsy patients. Neurology 60, 492—495. Timmings, P.L., 1993. Sudden unexpected death in epilepsy: a local audit. Seizure 2, 287—290. Tomson, T., Walczak, T., Sillanpaa, M., Sander, J.W., 2005. Sudden unexpected death in epilepsy: a review of incidence and risk factors. Epilepsia 46 (suppl. 11), 54—61. Vlooswijk, M.C., Majoie, H.J., De Krom, M.C., Tan, I.Y., Aldenkamp, A.P., 2007. SUDEP in the Netherlands: a retrospective study in a tertiary referral center. Seizure 16, 153—159. Walczak, T.S., Leppik, I.E., D’Amelio, M., Rarick, J., So, E., Ahman, P., Ruggles, K., Cascino, G.D., Annegers, J.F., Hauser, W.A., 2001. Incidence and risk factors in sudden unexpected death in epilepsy: a prospective cohort study. Neurology 56, 519—525.