-
ABCDEFG
UNIVERS ITY OF OULU P.O.B . 7500 F I -90014 UNIVERS ITY OF OULU
F INLAND
A C T A U N I V E R S I T A T I S O U L U E N S I S
S E R I E S E D I T O R S
SCIENTIAE RERUM NATURALIUM
HUMANIORA
TECHNICA
MEDICA
SCIENTIAE RERUM SOCIALIUM
SCRIPTA ACADEMICA
OECONOMICA
EDITOR IN CHIEF
PUBLICATIONS EDITOR
Senior Assistant Jorma Arhippainen
Lecturer Santeri Palviainen
Professor Hannu Heusala
Professor Olli Vuolteenaho
Senior Researcher Eila Estola
Director Sinikka Eskelinen
Professor Jari Juga
Professor Olli Vuolteenaho
Publications Editor Kirsti Nurkkala
ISBN 978-951-42-9554-6 (Paperback)ISBN 978-951-42-9555-3
(PDF)ISSN 0355-3221 (Print)ISSN 1796-2234 (Online)
U N I V E R S I TAT I S O U L U E N S I S
MEDICA
ACTAD
D 1118
ACTA
Eija Suorsa
OULU 2011
D 1118
Eija Suorsa
ASSESSMENT OF HEART RATE VARIABILITY AS AN INDICATOR OF
CARDIOVASCULAR AUTONOMIC DYSREGULATION IN SUBJECTS WITH CHRONIC
EPILEPSY
UNIVERSITY OF OULU,FACULTY OF MEDICINE,INSTITUTE OF CLINICAL
MEDICINE,DEPARTMENT OF NEUROLOGY
-
A C T A U N I V E R S I T A T I S O U L U E N S I SD M e d i c a
1 1 1 8
EIJA SUORSA
ASSESSMENT OF HEART RATE VARIABILITY AS AN INDICATOR OF
CARDIOVASCULAR AUTONOMIC DYSREGULATION IN SUBJECTS WITH CHRONIC
EPILEPSY
Academic dissertation to be presented with the assent ofthe
Faculty of Medicine of the University of Oulu forpublic defence in
Auditorium 8 of Oulu UniversityHospital, on 11 November 2011, at 12
noon
UNIVERSITY OF OULU, OULU 2011
-
Copyright © 2011Acta Univ. Oul. D 1118, 2011
Supervised byDocent Jouko IsojärviDocent Juha Korpelainen
Reviewed byProfessor Torbjörn TomsonDocent Aarne Ylinen
ISBN 978-951-42-9554-6 (Paperback)ISBN 978-951-42-9555-3
(PDF)
ISSN 0355-3221 (Printed)ISSN 1796-2234 (Online)
Cover DesignRaimo Ahonen
JUVENES PRINTTAMPERE 2011
-
Suorsa, Eija, Assessment of heart rate variability as an
indicator of cardiovascularautonomic dysregulation in subjects with
chronic epilepsy. University of Oulu, Faculty of Medicine,
Institute of Clinical Medicine, Department ofNeurology, P.O. Box
5000, FI-90014 University of Oulu, FinlandActa Univ. Oul. D 1118,
2011Oulu, Finland
AbstractAutonomic dysfunction in epilepsy is widely recognized.
Both partial and generalized epilepsiesaffect autonomic functions
during interictal, ictal and postictal states. Interestingly, there
isincreasing evidence of interictal autonomic nervous system
dysfunction as evidenced by reducedheart rate (HR) variability in
patients with epilepsy. Reduced HR variation has also been
detectedin many other chronic diseases and it has been shown to be
associated with unfavourable prognosiswith an increased risk of
mortality in various heart diseases. Recently, more attention has
also beenpaid to possible association of decreased HR variability
with sudden unexpected death in epilepsy(SUDEP). However, the
clinical significance of the observed changes in cardiovascular
regulationin patients with epilepsy is still poorly outlined and
there are no long-term studies about changesin HR variation in
relation to epilepsy.
This study was designed to evaluate long-term changes in
autonomic cardiovascular regulationin patients with temporal lobe
epilepsy (TLE) and also to evaluate HR variation during vagusnerve
stimulation (VNS) treatment in patients with refractory epilepsy,
using 24-hour ambulatoryECG recordings. Special attention was paid
to changes in HR variation and to circadian HRfluctuation over
time.
The results of this study show that autonomic cardiovascular
regulation is affected both inpatients with well-controlled TLE and
in patients with refractory TLE, and that the
cardiovasculardysregulation also presents itself with changes in
circadian HR variability, with more pronouncedalterations observed
during the night time. HR variability was also found to decrease
progressivelywith time in patients with chronic refractory TLE with
uncontrolled seizures. VNS treatment wasnot observed to alter HR
variation.
Keywords: autonomic nervous system, heart rate variation, sudden
unexpected death inepilepsy, temporal lobe epilepsy, vagus nerve
stimulation
-
Suorsa, Eija, Sydämen sykevaihtelu kroonisessa epilepsiassa.
Oulun yliopisto, Lääketieteellinen tiedekunta, Kliinisen
lääketieteen laitos, Neurologia, PL5000, 90014 Oulun yliopistoActa
Univ. Oul. D 1118, 2011Oulu
TiivistelmäEpilepsiapotilailla esiintyy autonomisen hermoston
toiminnan häiriöitä. Näitä häiriöitä voidaantodeta
epilepsiakohtausten aikana, heti kohtausten jälkeen ja kohtausten
välillä sekä paikallisal-kuisissa että yleistyneissä epilepsioissa.
Viimeaikaisissa tutkimuksissa on osoitettu kardiovasku-laarisen
säätelyjärjestelmän häiriöiden voivan ilmentyä alentuneena sydämen
sykevaihtelunaepilepsiakohtausten väliaikoina. Sydänsairauksien
yhteydessä sykevaihtelun vähenemisen onosoitettu liittyvän
kohonneeseen kuolemanriskiin. Epilepsiapotilailla alentuneen
sydämen syke-vaihtelun on epäilty liittyvän epilepsiapotilailla
ilmenevien odottamattomien ja ilman selkeääsyytä tapahtuvien
äkkikuolemien (SUDEP) lisääntyneeseen riskiin. Kertyneestä tiedosta
huoli-matta alentuneen sykevaihtelun kliininen merkitys
epilepsiapotilailla on edelleen epäselvä.
Pit-käaikaisseurantatutkimuksia sydämen sykevaihtelun muutoksista
epilepsiapotilailla ei ole jul-kaistu.
Tämän tutkimuksen tarkoituksena oli selvittää
ohimolohkoepilepsiaan liittyviä pitkäaikaisiainteriktaalisia
(kohtausten välillä esiintyviä) kardiovaskulaarisia ilmentymiä.
Lisäksi haluttiintutkia vaikeahoitoisessa epilepsiassa käytetyn
hoitomuodon, vagushermostimulaation, mahdolli-sia vaikutuksia
sydämen toimintaan. Erityisesti haluttiin analysoida sykevaihtelun
vuorokausi-rytmiä.
Tulokset osoittavat autonomisen hermoston kardiovaskulaarisen
säätelyjärjestelmän toimin-nan olevan häiriintyneen sekä
vaikeahoitoisilla että hyvähoitoisilla ohimolohkoalkuista
epilepsi-aa sairastavilla potilailla. Sydämen sykevariaatio on
alentunut erityisesti yöaikaan. Lisäksi sydä-men sykevaihtelu
alenee pitkäaikaisseurannassa vaikeahoitoista epilepsiaa
sairastavilla potilail-la, joilla ilmenee toistuvia epileptisiä
kohtauksia. Vagusstimulaatio ei aiheuttanut muutoksiasyketaajuuden
vaihteluun.
Asiasanat: autonominen hermosto, ohimolohko epilepsia, sydämen
sykevaihtelu,vagushermo stimulaattori, äkkikuolema
-
To my family
-
8
-
9
Acknowledgements
The research for this thesis was carried out at the Department
of Neurology, at the University of Oulu, during the years
1999–2011.
I wish to express my warmest thanks to my supervisor, Docent
Jouko Isojärvi, for introducing me to the world of epilepsy science
with such expertise. Your positive and enthusiastic attitude has
given me inspiration and guided me over the difficult times as
well. I also want to thank my supervisor Docent Juha Korpelainen,
who provided his practical advice and shared his scientific
knowledge. You always had time for discussions and your tireless
encouragement made it possible to finish this work. Thanks to my
supervisor’s encouragement, logical thinking and supportive
responsibility distribution, I have found the world of science
fascinating and challenging. It has been a privilege to work with
you.
I wish to express my gratitude to Professor Vilho V. Myllylä,
for his guidance and sympathetic support for my study. His broad
experience of science has created an inspiring atmosphere to work
in. I am also very grateful to Professor Matti Hillbom and
Professor Kari Majamaa for providing research facilities for
scientific work.
I wish to thank Professor Heikki Huikuri for his excellence
guidance in the field of heart rate variability analysis, and
Pirkko Huikuri for her valuable technical assistance in heart rate
variability data processing. I also wish to thank Professor Esa
Heikkinen for his expertise in the field of neurosurgery with vagus
nerve stimulation. I also thank Docent Kyösti Sotaniemi for his
kind encouragement and interest in my research work.
I wish to acknowledge most sincerely Professor Torbjörn Tomson
and Docent Aarne Ylinen for their expertise and valuable comments
during the preparation of the final manuscript for this thesis. I
also feel honoured that Professor Tapani Keränen kindly agreed to
serve as my opponent. I express my appreciation to Anna Vuolteenaho
for the careful revision of the English language of the
manuscript.
I owe my thanks to the entire staff of the Departments of
Neurology and Division of Cardiology for their excellent
co-operation through the years of this study. Special thanks are
due to Ilona Huovinen for her friendly secretarial assistance
throughout the study. I would also like to thank Risto Bloigu for
his expertise with statistical analyses. I wish to express my
sincere gratitude to the patients and their families who made this
work possible.
-
10
I warmly thank my colleagues in epilepsy research, Johanna
Rättyä, Hanna Ansakorpi, Virpi Pylvänen, Eeva Löfgren, Katja Luoma,
Kirsi Mikkonen and Usko Huuskonen, for their interest in my project
as well as for your friendship and inspiring conversation. I warmly
thank Jaana Huttula and Paula Kelhä for their friendship and
assistance.
It is my pleasure to thank Peter Baumann for introducing me to
the world of clinical neurology. It was such a pleasure to work
with you.
I thank all my dear and cheerful friends, who have made life
more interesting and enjoyable for me. Henna Mustaniemi, Milla
Riski and Anne Mäntyniemi, deserves my loving thanks for friendship
and laughs. You have a special place in my heart! My
parents-in-law, Sinikka and Esko Suorsa are warmly thanked for
their endless belief that I would finish this work, and also for
practical help with the care of our children.
I thank my sister Piia Pitkänen for your love, support and
laughter that have kept me going. Your own projects, Ida and
Oskari, are my sunshine. I owe my greatest gratitude to my
wonderful parents Salme and Veikko Ronkainen for providing me with
a supportive family and for advising me when needed. Your enormous
encouragement and belief in me have made lots of things happen.
Finally, my dearest thanks are expressed to my life companion
Ville Suorsa. Not only for your work for this thesis but also your
patience and love. We have been blessed with two special girls,
Ella and Anni. Love you.
This research project was supported by grants from the Epilepsy
Research Foundation, the Research Foundation of Orion Corporation,
the Centre for Arctic Medicine and the Oulu Medical Research
Foundation. All these are warmly acknowledged.
Oulu, September 2011 Eija Suorsa
-
11
Abbreviations
α short-term scaling exponent β slope of the power-law
relationship AED antiepileptic drug ANS autonomic nervous system
ApEn approximate entropy CBZ carbamazepine CNS central nervous
system CT computerized tomography ECG electrocardiography EEG
electroencephalography GABA gamma-amino butyric acid GBP gabapentin
HF high frequency HR heart rate LF low frequency LEV levetiracetam
LTG lamotrigine MRI magnetic resonance imaging NTS nucleus tractus
solitarius OXC oxcarbazepine PHT phenytoin RR interval
R-peak-to-R-peak interval SD1 instantaneous beat-to-beat RR
interval variability SD2 long-term continuous RR interval
variability SDNN standard deviation of all RR intervals SUDEP
sudden unexpected death in epilepsy TGB tiagabine TLE temporal lobe
epilepsy TPM topiramate VGB vigabatrin VLF very low frequency VNS
vagus nerve stimulation/stimulator VPA valproate
-
12
-
13
List of original articles
This thesis is based on the following publications, which are
cited in the text by their Roman numerals:
I Ronkainen E, Ansakorpi H, Huikuri HV, Myllylä VV, Isojärvi JIT
& Korpelainen JT (2005) Suppressed circadian heart rate
dynamics in temporal lobe epilepsy. J Neurol Neurosurg Psychiatry
76(10): 1382–1386.
II Suorsa E, Korpelainen JT, Ansakorpi H, Huikuri HV, Suorsa V,
Myllylä VV & Isojärvi JIT (2011) Heart rate dynamics in
temporal lobe epilepsy – a long term follow-up study. Epilepsy
Research 93(1): 80–83.
III Suorsa E, Isojärvi JIT, Ansakorpi H, Huikuri HV, Suorsa V,
Myllylä VV & Korpelainen JT (2011) Long-term changes in
circadian heart rate variability in patients with temporal lobe
epilepsy. Manuscript
IV Ronkainen E, Korpelainen JT, Heikkinen E, Myllylä VV, Huikuri
HV & Isojärvi JIT (2006) Cardiac autonomic control in patients
with refractory epilepsy before and during vagus nerve stimulation
treatment – a one year follow-up study. Epilepsia 47(3):
556–562.
-
14
-
15
Contents
Abstract Tiivistelmä Acknowledgements 7 Abbreviations 11 List of
original articles 13 Contents 15 1 Introduction 17 2 Review of the
literature 19
2.1 General aspects of epilepsy
.....................................................................
19 2.1.1
Definition......................................................................................
19 2.1.2 Epidemiology
...............................................................................
19 2.1.3 Aetiology
......................................................................................
20 2.1.4 Classification
................................................................................
20 2.1.5 Diagnosis
......................................................................................
24 2.1.6 Prognosis
......................................................................................
25
2.2 Temporal lobe epilepsy
...........................................................................
25 2.3 Treatment of epilepsy
..............................................................................
26
2.3.1 Antiepileptic drugs
.......................................................................
26 2.3.2 Surgery
.........................................................................................
31 2.3.3 Vagus nerve stimulation
...............................................................
32
2.4 Autonomic nervous system
.....................................................................
34 2.4.1 Anatomy of the autonomic nervous system
................................. 34 2.4.2 Cardiovascular
regulation
.............................................................
35
2.5 Heart rate variability and its clinical implications
.................................. 39 2.5.1 Physiological
background of heart rate variability and
heart rate dynamics
.......................................................................
39 2.5.2 Factors affecting heart rate variability
.......................................... 41 2.5.3 Heart rate
variability in pathological conditions ..........................
42
2.6 Epilepsy and autonomic cardiovascular dysregulation
........................... 43 2.6.1 Ictal autonomic dysfunction
......................................................... 43 2.6.2
Interictal heart rate variation
........................................................ 44 2.6.3
Circadian heart rate variation
....................................................... 45 2.6.4
Effect of vagus nerve stimulation on cardiovascular
autonomic function
.......................................................................
46 2.7 Sudden unexpected death in epilepsy
..................................................... 47
-
16
2.7.1 Definition
......................................................................................
47 2.7.2 Epidemiology
...............................................................................
47 2.7.3 Aetiology
......................................................................................
48
3 Aims of the study 51 4 Subjects and methods 53
4.1 Subjects
...................................................................................................
53 4.2 Methods
...................................................................................................
56
4.2.1 Clinical examination (Studies I-IV)
.............................................. 56 4.2.2 Adjustment
and use of vagus nerve stimulator (Study IV) ........... 56 4.2.3
Analysis of heart rate behaviour (Studies I-IV)
............................ 57 4.2.4 Statistical analysis
.........................................................................
59
5 Results 61 5.1 Clinical evaluation of autonomic nervous system
function .................... 61 5.2 Cardiac regulation in temporal
lobe epilepsy .......................................... 61
5.2.1 Long-term heart rate dynamics (Study II)
.................................... 61 5.2.2 Circadian heart rate
variation (Study I) ........................................ 62
5.2.3 Long-term changes in circadian heart rate variation
(Study
III)
.................................................................................................
65 5.3 Effect of vagus nerve stimulation on heart rate dynamics
(Study
IV)
...........................................................................................................
68 6 Discussion 71
6.1 General
aspects........................................................................................
71 6.2 Clinical findings of autonomic nervous system function
in
patients with epilepsy
..............................................................................
72 6.3 Cardiac regulation in temporal lobe epilepsy
.......................................... 72
6.3.1 Long-term heart rate dynamics
..................................................... 72 6.3.2
Circadian heart rate variation
....................................................... 73 6.3.3
Long-term changes in circadian heart rate variation
.................... 75 6.3.4 Effect of antiepileptic medication on
heart rate variation ............. 76
6.4 Effect of vagus nerve stimulation on heart rate
dynamics....................... 76 6.5 Methodological
considerations
...............................................................
77
7 Conclusions 79 References 81 Original publications 101
-
17
1 Introduction
Epilepsy is a descriptive term for a large group of anatomical
and functional disorders of the brain that are characterized by
repeated seizures. Epilepsy is one of the most common serious brain
disorders. Throughout history, epilepsy has been associated with
superstition and stigma, but over the past decades scientists have
made significant advances in the understanding and treatment of
this disorder. Simultaneously, public knowledge of this disorder
has improved, and attitudes have slowly changed towards its
acceptance.
The prevalence of active epilepsy is estimated at 3–8 per 1,000
of the population according to European studies (Keränen et al.
1989, Olafsson & Hauser 1999, Rocca et al. 2001). The treatment
of epilepsy is mainly based on drug treatment. However, for those
who do not achieve seizure freedom despite adequate antiepileptic
drug (AED) treatment epilepsy surgery may be a therapeutic
alternative. With appropriate treatment, 70–75% of patients with
epilepsy become seizure-free, but the remainder will be resistant
to treatment and classified as patients with refractory epilepsy
(Sander & Shorvon 1996, Cockerell et al. 1997, Kwan &
Brodie 2000, Kwan & Sander 2004). Electrical stimulation of the
vagus nerve is a treatment for patients with refractory epilepsy
who are unsuitable candidates for resective surgery or who have
experienced insufficient benefit from such treatment (Uthman et al.
1993, Ben-Menachem 2002). Despite appropriate treatment, patients
with epilepsy have mortality rates that are 2–3 times higher
compared to general population (Olafsson et al. 1998, Tomson
2000).
Epilepsy may affect autonomic nervous system (ANS) function
during interictal (between seizures), ictal (during seizures) and
postictal (after seizures) states. Cardiac function may be altered
in patients with epilepsy as a result of autonomic dysfunction at
several different levels, including central and peripheral ANS
pathways, and also at the level of the heart. It is not clear
whether the observed imbalance of the sympathetic and
parasympathetic input to the heart in patients with epilepsy is due
to epilepsy per se or whether other factors, such as medication or
its withdrawal, may play a role as well (Isojärvi et al. 1998,
Tomson et al. 1998, Hennessy et al. 2001, Ansakorpi et al.
2002).
Traditional time and frequency domain measures of heart rate
(HR) variability along with the newer methods based on fractal and
complexity scaling of RR interval variability have both been used
as non-invasive tools for assessing autonomic cardiovascular
regulation (Task Force 1996). Using conventional
-
18
short- and long-term ECG recordings, previous studies have
reported diminished interictal HR variability in patients with
epilepsy, mainly temporal lobe epilepsy (TLE), but it has remained
unclear whether the observed reduction in cardiovascular responses
is due to the epileptic process itself or to the AEDs (Frysinger et
al. 1993, Isojärvi et al. 1998, Tomson et al. 1998, Ansakorpi et
al. 2000, Ansakorpi et al. 2002). On the other hand, low HR
variability has also been reported in a number of other
pathophysiologic conditions (Bernardi et al. 1992, Huikuri et al.
1994, Korpelainen et al. 1997), and has been shown to be a marker
of an increased risk of mortality in patients with these conditions
(Kleiger et al. 1987, Huikuri et al. 1994, Malliani et al. 1994,
Barron & Viskin 1998).
Patients with epilepsy are at increased risk for SUDEP, and a
large body of data has defined different risk factors for SUDEP,
e.g. youth, polytherapy with AEDs, lack of compliance with
treatment and poor seizure control (Devinsky et al. 1994, Nilsson
et al. 1999, Opeskin et al. 2000, Tomson et al. 2008). However,
otherwise healthy, compliant patients may also die unexpectedly
(Earnest et al. 1992, Nashef et al. 1998). Recently, more attention
has been paid to possible association between altered
cardiovascular function and SUDEP. It has been suggested that
reduced HR variability may play a role in the pathophysiology of
sudden unexpected death in epilepsy (SUDEP) in patients with
chronic epilepsy (Tomson et al. 1998). However, according to one
recent study no HR variability parameter was associated with SUDEP,
suggesting that HR variability parameters are not clear-cut
predictors for SUDEP (Surges et al. 2009a).
There are no previous longitudinal studies on changes in
autonomic cardiovascular regulation in patients with TLE over time.
There is also limited information on the circadian HR variation in
this patient group. Furthermore, there are only few published
short-term studies on the effects of VNS treatment on
cardiovascular regulation despite the close interaction between VNS
and the centres controlling cardiovascular regulation.
The present study was designed to evaluate long-term
cardiovascular autonomic regulation in patients with TLE and in
patients with VNS treatment by using 24-hour ECG recordings.
-
19
2 Review of the literature
2.1 General aspects of epilepsy
2.1.1 Definition
Epilepsy is the name of a brain disorder characterized
predominantly by recurrent and unpredictable interruptions of
normal brain function, called epileptic seizures. Epileptic
seizures can affect sensory, motor, and autonomic function, which
are characterized by various clinical manifestations such as
decrease of consciousness, abnormal sensory phenomena, increased
autonomic activity and involuntary movements. Epilepsy is not a
singular disease entity but a dynamic process, which reflects
complex functional changes occurring in the anatomy and physiology
of the brain in the presence of environmental and genetic factors.
(Waltimo 1983, Fisher et al. 2005)
2.1.2 Epidemiology
Epilepsy is the most common chronic disorder of the central
nervous system (CNS). Numerous studies have been published on the
epidemiology of epilepsy. However, the study designs and the
definition of epilepsy and seizure types differ from one study to
another, which makes comparison between different studies
difficult.
Prospective, population-based studies indicate that in general
population there is an 8–10% lifetime risk of one seizure (Hauser
et al. 1990) and a 3% chance of epilepsy (Hauser et al. 1993). The
incidence of epilepsy is 24–53 /100,000 person years in developed
countries (Keränen et al. 1989, Olafsson et al. 1996, Zarrelli et
al. 1999, MacDonald et al. 2000), whereas in developing countries
it is considered to be higher. The incidence is high in childhood
and increases again in elderly people (Sillanpää 1973, Keränen et
al. 1989, Hauser et al. 1993, Forsgren et al. 1996, Olafsson et al.
1996, Beilmann et al. 1999). Interestingly, a recent
population-based study found that the incidence of epilepsy in
Finland has declined significantly in both children and adults with
a concurrent increase in incidence among the elderly (Sillanpää et
al. 2006), but the reasons for the changes in incidence remained
unclear.
The prevalence of active epilepsy, often defined as patients
with epilepsy who have had at least one seizure during the last 5
years, is 3.3–7.8 per 1000
-
20
inhabitants in studies performed in European countries (Granieri
et al. 1983, Keränen et al. 1989, Olafsson & Hauser 1999, Rocca
et al. 2001). In most previous studies the prevalence of epilepsy
has been higher in males than females (Granieri et al. 1983, Hauser
1997, Olafsson & Hauser 1999, Rocca et al. 2001). However, the
prevalence difference between genders has rarely been shown to be
statistically significant (Granieri et al. 1983, Keränen et al.
1989). According to the Finnish Social Insurance Institution,
58,594 patients out of a total population of approximately 5
million received reimbursement for antiepileptic medication in
Finland in 2009. Approximately 9,000 of these patients suffer from
medically intractable epilepsy. (Social Insurance Institution
2006)
2.1.3 Aetiology
Almost any cerebral pathology may be associated with epilepsy.
In the adult population the cause of epilepsy is unknown in the
majority of patients (Beghi 2004). The most common causes of
epilepsy are cerebrovascular diseases, head trauma, intracranial
haemorrhages, cerebral tumours and neurodegenerative diseases.
(Sander et al. 1990, Forsgren et al. 1996, Olafsson et al. 1996,
Oun et al. 2003). In children metabolic defects, congenital
malformations, infections and genetic diseases are among common
aetiologies (Beghi 2004). It seems that almost everyone may
experience a seizure in a particular set of circumstances, but some
people seem to have a lower seizure threshold than others.
2.1.4 Classification
Classification of epileptic seizures
Epileptic seizures can be classified in several different ways.
Their electroclinical features identify them as either partial or
generalized. Partial seizures start in a circumscribed set of nerve
cells (the “epileptic focus”) in one hemisphere of the brain and
spread from there. Generalized seizures involve both sides of the
brain from the onset, although they may sometimes involve only a
small part of the two hemispheres in a symmetrical manner. (ILAE
1981) Table 1 presents 1981 ILAE classification of seizures, which
was used in the present thesis. However, new terminology and
classification has been established since the start of this
research (ILAE 2005–2009) (Berg et al. 2010).
-
21
According to new classification of epilepsies, the diagnostic
scheme provides the basis for a standardized description of
individual patients, and consists of five levels. The levels are
organized to facilitate a logical clinical approach to the
development of hypotheses necessary to determine the diagnostic
studies that should be performed in individual patients. The levels
are 1) ictal phenomenology 2) epileptic seizure type/types 3)
epileptic syndrome 4) aetiology and 5) the degree of impairment
caused by the epileptic condition.
Table 1. International classification of seizures (Commission on
Classification and Terminology, ILAE, 1981).
Class Classification
I Partial (focal) seizures
A Simple partial seizures (consciousness not impaired)
1. With motor signs
Focal motor without march
Focal motor with march (Jacksonian)
Versive
Postural
Phonatory (vocalization or arrest of speech)
2. With somatory or special-sensory symptoms (simple
hallucinations, e.g. tingling, light
flashes, buzzing)
Somatosensory
Visual
Auditory
Olfactory
Gustatory
Vertiginous
3. With autonomic symptoms or signs (including epigastric
sensation, pallor, sweating,
flushing, piloerection, and pupillary dilatation)
4. With psychic symptoms (disturbances of higher cerebral
functions); these symptoms rarely
occur without impairment of consciousness and are much more
commonly experienced as
complex partial seizures
Dysphasic
Dynamic (e.g. déjà vu)
Cognitive (e.g. dreamy states, distortions of time sense)
Affective (fear, anger, etc.)
Illusions (e.g. macropsia)
Structured hallucinations (e.g. music, scenes)
B. Complex partial seizures (with impairment of consciousness;
may sometimes begin with
simple symptomatology)
1. Simple partial onset followed by impairment of
consciousness
-
22
Class Classification
With simple partial features (A.1. - A.4.) followed by impaired
consciousness
With automatisms
2. With impairment of consciousness at onset
With impairment of consciousness only
With automatisms
C. Partial seizures evolving to secondary generalized seizures
(may be generalized tonic-
clonic, tonic, or clonic)
1. Simple partial seizures (A) evolving to generalized
seizures
2. Complex partial seizures (B) evolving to generalized
seizures
3. Simple partial seizures evolving to complex partial seizures
evolving to generalized seizures
II Generalized seizures (convulsive or non-convulsive)
A.
1. Absence seizures
Impairment of consciousness only
With mild clonic components
With atonic components
With tonic components
With automatisms
With autonomic components
2. Atypical absence
May have:
Changes in tone that are more pronounced than in A.1.
Onset and/or cessation that is not abrupt
B. Myoclonic seizures
Myoclonic jerks (single or multiple)
C. Clonic seizures
D. Tonic seizures
E. Tonic-clonic seizures
F. Atonic seizures
III Unclassified seizures
Includes all seizures that cannot be classified because of
inadequate or incomplete data.
Classification of epilepsies and epileptic syndromes
Epilepsies and epileptic syndromes have been classified
according to aetiology and anatomic origin of seizures. Symptomatic
epilepsies are due to a recognizable insult to the brain (e.g.
resulting from a malformation, trauma, or tumour). In idiopathic
epilepsies, a known cause cannot be identified, but they are
commonly caused by a genetic defect. If a symptomatic aetiology is
suspected but cannot be demonstrated, the condition is called
cryptogenic epilepsy. (Shorvon et al. 2004)
-
23
Table 2 presents the 1989 ILAE classification of epilepsies and
epileptic syndromes, which is used in this thesis. According to new
proposed terminology and classification (ILAE 2005–2009) genetic,
structural-metabolic and unknown represent modified concepts to
replace idiopathic, symptomatic and cryptogenic. Effective
classification of seizures and syndromes is indispensable for
adequate therapy and prognosis.
Table 2. International classification of epilepsies and
epileptic syndromes (Commission on Classification and Terminology,
ILAE, 1989).
Class Classification
1. Localization-related (focal, local, partial)
1.1. Idiopathic (with age-related onset)
Benign childhood epilepsy with centrotemporal spikes
Childhood epilepsy with occipital paroxysm
Primary reading epilepsy
1.2. Symptomatic
Chronic progressive epilepsia partialis continua of
childhood
1.3. Cryptogenic
The symptomatic and cryptogenic categories comprise syndromes of
great individual
variability that are based on:
Seizure types (according to the International classification of
Epileptic Seizures)
Anatomic localization: Temporal, frontal, parietal, and
occipital lobe epilepsies
Bi- and multilobar epilepsies
Aetiology (in symptomatic epilepsies)
Specific modes of precipitation
2. Generalized
2.1. Idiopathic (with age-related onset, in order of age)
Benign neonatal familial convulsions
Benign neonatal convulsions
Benign myoclonic epilepsy of infancy
Childhood absence epilepsy (pyknolepsy)
Juvenile absence epilepsy
Juvenile myoclonus epilepsy (impulsive petit mal)
Epilepsy with grand mal (GTC) seizures on awaking
Other idiopathic generalized epilepsies not defined above
Epilepsies with seizure precipitated by specific modes of
activation
2.2. Cryptogenic or symptomatic (in order of age)
West syndrome (infantile spasms, Blitz-Nick-Salaam-Krämpfe)
Lennox-Gastaut syndrome
Epilepsy with myoclonic-astatic seizures
Epilepsy with myoclonic absences
-
24
Class Classification
2.3. Symptomatic
2.3.1. Non-specific aetiology
Early myoclonic encephalopathy
Early infantile epileptic encephalopathy with
suppression-burst
Other symptomatic generalized epilepsies not defined above
2.3.2 Specific syndromes (see the original reference)
3. Epilepsies and syndromes undetermined whether focal or
generalized
3.1. With both generalized and focal seizures
Neonatal seizures
Severe myoclonic epilepsy of infancy
Epilepsy with continuous spike-waves during sleep
Acquired epileptic aphasia (Landau-Kleffner syndrome)
Other undetermined epilepsies not defined above
3.2. Without unequivocal generalized or focal features (e.g.
many cases of sleep-grand
mal)
4. Special syndromes
4.1 Situation-related seizures (Gelegenheitsanfälle)
Febrile convulsions
Isolated seizures or isolated status epilepticus
Seizures due to acute metabolic or toxic factors such as
alcohol, drugs, eclampsia
2.1.5 Diagnosis
Since a variety of conditions can cause episodes of transiently
disturbed consciousness or function, the identification of an
epileptic seizure is important. If a patient has had one unprovoked
seizure with epileptiform discharges in electroencephalography
(EEG) or two or more unprovoked seizures, the diagnosis of epilepsy
can be made. The diagnosis of epilepsy is based on medical history
and a detailed description of events that occurred before, during
and after a suspected epileptic seizure. Eyewitness information is
often essential. Detailed clinical examination, focusing on
neurological and cardiovascular evaluation is needed, although the
clinical examination does not often reveal abnormal findings in
cases of new onset epilepsy. Laboratory tests add little to the
diagnostics of epilepsy. EEG recording is usually conducted to
detect possible spike or sharp wave discharges. Electrocardiography
(ECG) is obtained to exclude cardiac abnormalities. Cerebral
computerized tomography (CT) or magnetic resonance imaging (MRI)
scan is done to visualize possible focal structural pathology in
the brain. Because the examinations often yield normal results, the
recognition and
-
25
diagnosis of an epileptic seizure is almost entirely based on
medical history, and a detailed description of the clinical
features of the seizure is essential for the diagnosis. (Shorvon et
al. 2004, Elger & Schmidt 2008)
2.1.6 Prognosis
The prognosis of epilepsy varies greatly depending on the
aetiology and type of epileptic syndrome but the overall prognosis
for patients with epilepsy is good in terms of seizure remission.
There are differences among published studies regarding definitions
of remission, duration of the follow-up, and the number of seizures
the patients have experienced, which makes comparison of different
studies difficult.
In epilepsy, three prognostic groups are generally considered:
(1) spontaneous remission (20–30% of the over all epilepsy patient
population) as seen e.g. in benign epilepsy with centrotemporal
spikes or childhood absences; (2) seizure remission achieved with
AEDs (20–30%), which occurs in most focal epilepsies and juvenile
myoclonic epilepsy syndromes; (3) persistent clinical seizures
despite AEDs (30–40%), i.e. refractory epilepsy. (Kwan & Sander
2004)
Despite an overall good prognosis for seizure control, epilepsy
is a potentially life-threatening condition associated with
increased mortality (Cockerell et al. 1997). In population-based
studies, the mortality rates have been approximately 2–3 times
higher in patients with epilepsy than in the general population
(Olafsson et al. 1998, Tomson 2000). Increased mortality rates are
due to deaths that are unrelated to epilepsy, e.g. ischaemic heart
disease, neoplasm outside the CNS, and to deaths in which epilepsy
itself is the cause of death, e.g., suicides, accidents and status
epilepticus. (Hauser et al. 1980, Cockerell et al. 1994) SUDEP is
the most important epilepsy-related cause of death, reported to be
responsible for 18–25% of deaths among patients with
medically-refractory epilepsy (Walczak et al. 2001). SUDEP is
discussed in detail in section 2.6.
2.2 Temporal lobe epilepsy
TLE is the most common epileptic syndrome. Anatomically TLE can
be subclassified according to seizure origin to those with seizures
originating in the mesial temporal structures (MTLE) and those with
seizures beginning elsewhere in the temporal lobe (e.g. neocortical
temporal lobe epilepsy). (Shorvon et al. 2004)
-
26
In MTLE seizures originate from the limbic areas of the mesial
temporal lobe such as the hippocampus, amygdala and the
parahippocampal gyrus, which are the most epileptogenic regions of
the brain. The main pathology associated with MTLE is hippocampal
sclerosis, which may result from previous status epilepticus,
complicated febrile convulsions, encephalitis or ischaemic insult
(Heinemann 2004). However, other mesial temporal pathologies, e.g.
tumours and congenital pathologies, may also result in MTLE. Recent
studies have also implicated a strong genetic role in the
development of hippocampal sclerosis (Kobayashi et al. 2001).
In MTLE, the seizures usually start in early childhood (Janszky
et al. 2004), but there may be a long seizure-free period from the
primary insult before unprovoked seizures develop. The seizures
often respond well to AEDs at the beginning, but when the seizures
return in adolescence they become refractory to medical treatment
(Berg 2008). It has been widely accepted that there is a strong
correlation between hippocampal sclerosis and the severity of
epilepsy. Patients with TLE who have hippocampal sclerosis will
become seizure-free in only 10–30% of the cases with the use of
adequate AED treatment, and these patients should therefore be
considered for surgical evaluation. (Shorvon et al. 2004)
2.3 Treatment of epilepsy
2.3.1 Antiepileptic drugs
The treatment of epilepsy is mainly based on drug treatment. In
recent decades, several new AEDs have been developed, and there are
currently more than 20 AEDs available, making it challenging for
physicians to master the optimal use of these agents (Prunetti
& Perucca 2011).
The goal of the treatment with AEDs is to achieve complete
seizure freedom with as few side effects as possible (Duncan et al.
2006). The treatment is usually started after two or more
unprovoked epileptic seizures, but may also be considered in
patients after a single seizure if specific prognostic factors
indicate a high risk of recurrence, e.g. in patients with an
underlying brain disorder, when the EEG shows interictal
epileptiform abnormalities, or in patients who have a high-risk
epilepsy syndrome, such as juvenile myoclonic epilepsy. (Kälviäinen
& Keränen 2001, Shorvon et al. 2004, Brodie 2005) The decision
on whether treatment is appropriate requires careful consideration
of the individual risk-benefit ratio related to treatment, based on
such issues as the type and frequency
-
27
of seizures, type of epilepsy syndrome, age and sex of the
patient, presence of associated medical conditions and the possible
side effects of the AED chosen. (Adult Epilepsy:Current Care
summary,2008) In a few specific epilepsy syndromes, such as benign
Rolandic epilepsy, pharmacotherapy may be unneccessary (Brodie
2005).
Monotherapy is preferred, and about 60–70% of patients with
recent onset epilepsy respond to AED treatment (Cockerell et al.
1995, Kwan & Brodie 2001). Those who do not experience
satisfactory seizure control with monotherapy often require
polytherapy (combination of two or more AEDs). If the seizures
continue at the maximally tolerated dose of the first appropriately
chosen AED, and no underlying aetiology for the seizures is
identified, it may indicate refractoriness of epilepsy (Kwan &
Brodie 2000). Refractory epilepsy is usually defined as a scenario
where a patient continues to have seizures despite adequate use of
two well-tolerated AEDs (Kwan et al. 2010). There is currently no
evidence to suggest whether switching to monotherapy with another
AED or adding another AED is more effective in the treatment of
seizures in subjects who fail their AED (Beghi et al. 2003). It has
been suggested that combining drugs with different mechanisms of
action is beneficial (Brodie 2005). However, the classification of
a patient’s epilepsy as drug resistant at a given point in time is
valid only at the time of assessment and does not necessarily imply
that the patient will never become seizure-free on further
manipulation of AED therapy (Callaghan et al. 2007, Schiller &
Najjar 2008). Since many epilepsies are prone to undergo remission,
the possibility of discontinuation of the AED should be considered
after 3–5 years’ seizure freedom (Schmidt & Gram 1996).
Combining AEDs requires an understanding of their pharmacology,
in particular their mechanism of action (Rogawski 2002, Elger &
Schmidt 2008). Although, the mechanisms of action of all AEDs are
not fully understood, they fall into a number of general
categories. The molecular targets and clinical efficacy of
different AEDs are presented in Table 3.
-
28
Table 3. Molecular targets and the spectrum of clinical efficacy
of antiepileptic drugs.
Drug Na+
Channels
Ca2+
Channels
GABAA
Receptor
GABA
Transaminase
GABA
Transporter
GABAB
Receptor
NMDA
Receptor
Clinical
Efficacy
Carbamazepine + Partial,
GTC
Oxcarbazepine + Partial,
GTC
Lacosamide + Partial
Phenytoin + Partial,
GTC
Lamotrigine + + Broad
Spectrumb
Zonisamidea + + Partial,
GTC,
myoclonicb
Ethosuximide + + Absence
Phenobarbital + + Broad
Spectrum
Benzodiazepines + Broad
spectrum
Vigabatrin + Partialb
Tiagabine + Partial
Gabapentin + + Partial
Felbamate + + + Broad
spectrum
Topiramatea + + Broad
spectrumb
Valproic acid + + + Broad
spectrum
Levetirecetam + Partial
GTC aZonisamide and topimarate are weak carbonic anhydrase
inhibitors bVigabatrin, lamotrigine, zonisamide and topimarate may
be useful in treating infantile spasm
GTC Generalized tonic-clonic
Modified from Rogawski 2002 and Elger & Schmidt 2008
Sodium channel blockers
Carbamazepine (CBZ) is first-line or adjunctive therapy in
partial seizures with or without secondary generalization and is
not recommended in the absence or myoclonic seizures. CBZ is
generally well tolerated but CNS side effects are
-
29
fairly common when the serum concentration is high. CBZ is
highly (70–80%) bound to serum proteins and metabolized almost
entirely by the liver, and induces the cytochrome P450 enzyme
system in the liver, which may increase or decrease the metabolism
of other drugs, including other AEDs. (Brodie 1992, Sillanpää 2004)
The main mechanism of action of CBZ is on neuronal sodium channel
conductance by reducing high-frequency repetitive firing of action
potentials (Sillanpää 2004), and CBZ slows the conduction velocity
of both central and peripheral nerves (Traccis et al. 1983,
Mervaala et al. 1987). CBZ also blocks N-acetyl-D-aspartate
receptors (MacDonald 2002).
Due to the mechanism of action, CBZ is associated with a delay
in atrioventricular conduction and may induce brady-arrhythmias
(Ladefoged & Mogelvang 1982, Boesen et al. 1983, Kasarskis et
al. 1992), although sinus tachycardia has also been observed
(Kasarskis et al. 1992). However, there is evidence that CBZ, at
therapeutic levels, has no or minimal effects on the heart
conduction system in the vast majority of patients (Kennebäck et
al. 1992). In addition, CBZ has been shown to increase the
sympathetic tone in the autonomic nervous system and to suppress
both parasympathetic and sympathetic function (Isojärvi et al.
1998). CBZ treatment has also been suggested to be associated with
an increased risk of SUDEP, but the findings are controversial
(Kennebäck et al. 1997, Timmings 1998, Nilsson et al. 1999, Walczak
et al. 2001, Hesdorffer et al. 2011).
Oxcarbazepine (OXC) is a 10-keto analogue of CBZ and its active
monohydroxy derivative limits the firing of sodium-dependent action
potentials at lower concentrations than CBZ. OXC has less potential
for drug interactions than CBZ since it is not metabolized by
cytochrome P450-dependent enzymes. OXC is used as monotherapy or
adjunctive therapy in the treatment of partial seizures with or
without generalization. (Faught 2004) The most common side effects
early in the treatment that often lead to discontinuation of OXC
affect the CNS. These side effects include headache, dizziness,
ataxia, and nausea (Faught 2004). Furthermore, hyponatraemia is a
common side effect of OXC therapy, especially in the elderly and
female patients (Pendlebury et al. 1989, Isojärvi et al. 2001).
Lamotrigine (LTG) acts by blocking voltage-dependent sodium and
calcium channels in the neural membranes of the brain, thus
reducing the epileptic activity (Meldrum 1996). LTG also inhibits
gamma-amino butyric acid (GABA) release. Furthermore, LTG has been
shown to inhibit the cardiac rapid delayed rectifier potassium ion
current (Ikr) (Danielsson et al. 2005). Ikrblocking drugs are
-
30
generally considered to be associated with an increased risk of
cardiac arrhythmia and SUDEP (Danielsson et al. 2005).
LTG is indicated as an adjunctive or monotherapy in partial and
generalized epilepsies as well as in Lennox-Gastaut syndrome.
Headache, skin rash and nausea are common side-effects of LTG
(Brodie et al. 1995). No ECG abnormalities have been associated
with the use of LTG (Betts et al. 1991).
Phenytoin (PHT) is used for partial and tonic-clonic seizures
especially when the seizures are generalized. PHT, as i.v.
formulation, is also used to treat status epilepticus. The
mechanism of action is thought to be based on the drug’s capacity
to bind to and prolong the inactivation of voltage-dependent sodium
channels in neuronal cell membranes. (Eadie 2004) PHT has a large
number of interactions with AEDs and other drugs. PHT is highly
protein bound (90%) and metabolized by hepatic P450 enzymes, which
may increase or decrease the metabolism of other drugs, including
other AEDs. PHT also has antiarrhythmic properties and it depresses
the hyperactivity of cardiac sympathetic nerves. (Eadie 2004) The
main adverse cardiac effect of PHT reported is bradyarrhythmias,
although mostly in i.v. administration of the drug (Earnest et al.
1983). There is also a case report of complete atrioventricular
block with ventricular asystole in a patient receiving i.v. PHT
(Randazzo et al. 1995). The main advantage of PHT over other AEDs
is good efficacy with low cost, but long-term side effects make it
less attractive as a long-term treatment for the majority of
epilepsy patients.
Gabaergic drugs
Tiagabine (TGB) is a GABA uptake inhibitor and it prolongs the
duration of the peak inhibitory postsynaptic current, consistent
with temporarily sustained levels of endogenously released GABA in
the synapse. TGB has proven effective as add-on therapy in patients
with refractory partial seizures with or without secondary
generalization. (Schachter 1999)
Vigabatrin (VGB) is used as adjunctive therapy in partial and
secondarily generalized epilepsy (Dichter & Brodie 1996), and
also for infantile spasms and Lennox-Gastaut syndrome (Appleton et
al. 1999). VGB inhibits presynaptic GABA degradation by selective,
enzyme-activated irreversible blockade of the mitochondrial enzyme
GABA transaminase. VGB is not widely used because its use is
associated with persistent peripheral visual field defects
(Kälviäinen et al. 1999).
-
31
Drugs with other mechanisms of action
Ethosuximide is the first-line or adjunctive therapy in
generalized absence seizures. The mechanism of action against
absence seizures is the reduction of low threshold T-type calcium
currents in thalamic neurons. (MacDonald & Kelly 1995)
Gabapentin (GBP) is used as monotherapy in adults with partial
or secondarily generalized epilepsy. It is well tolerated and does
not have any significant drug interactions, and is therefore, is
easy to use. GBP is thought to inhibit high voltage-activated
calcium channels (α2δ subunit) (Elger & Schmidt 2008).
Valproate (VPA) differs structurally from other AEDs and its
mechanism of action has remained undefined until now. Different
studies have suggested that VPA increases GABA concentrations
through the activation of the GABA-synthesizing enzyme glutamic
acid decarboxylase. However, VPA also blocks voltage-dependent
sodium channels and affects calcium (T) conductance. (Arroyo 2004)
VPA is the drug of choice for primary generalized epilepsies and
useful in a wide spectrum of other epilepsies (Perucca 2002). VPA
has been shown to be associated with obesity and hyperinsulinaemia,
which may promote hyperandrogenism, anovulatory cycles, amenorrhoea
and polycystic ovary syndrome that are well-known side effects of
VPA (Isojärvi et al. 1993, Rättyä et al. 1999, Mikkonen et al.
2004).
Topiramate (TPM) is a broad spectrum AED with multiple
mechanisms of action. It enhances GABA action, but it also inhibits
sodium conduction, AMPA subtype glutamate receptors and L-type
high-voltage-activated calcium channels. TPM is used in partial and
generalized epilepsies as adjunctive therapy and as monotherapy if
epilepsy is r efractory to commonly used AEDs. Its use is limited
by cognitive side effects. (Glauser 1999)
Levetiracetam (LEV) is well tolerated and effective against
partial and generalized seizures. The exact mechanism of action is
not fully confirmed, but it is thought that LEV binds to a synaptic
vesicle protein, SV2A, and acts by modulating its function (Elger
& Schmidt 2008).
2.3.2 Surgery
Epilepsy surgery is defined as any neurosurgical intervention
with the primary goal of eliminating epileptic seizures. About 30
per cent of patients do not
-
32
achieve seizure freedom despite adequate AED treatment (Kwan
& Brodie 2001, Kwan & Sperling 2009), and for these
patients epilepsy surgery may be a therapeutic alternative.
Recently, drug-refractory epilepsy has been defined as the failure
of adequate trials of two well-tolerated and appropriately chosen
and used AEDs to achieve sustained freedom from seizures (Kwan
& Sperling 2009). Trials can usually be completed during 2–3
years (Shorvon & Luciano 2007). If epilepsy is not controlled
within this time interval, it is unlikely to ever be completely
controlled with AEDs alone (Kwan & Brodie 2000) and the
evaluation of surgery is appropriate at that time.
Resective epilepsy surgery can be a curative therapy when the
epileptic zone can be identified, and results in complete seizure
control in about 60–80% of this type of patients with intractable
epilepsy (Wiebe et al. 2001, Jutila et al. 2002, Spencer & Huh
2008). Candidates for epilepsy surgery need to go through a careful
comprehensive pre-surgical evaluation.
2.3.3 Vagus nerve stimulation
General aspects
VNS is a non-pharmacological antiepileptic therapy for patients
with refractory seizures over the age of 12 years who are not
candidates for resective surgery or who have had resective surgery
with unsatisfactory results (Uthman et al. 1993, Ben-Menachem
2002). Absolute contraindications for implantation of VNS are
limited to previous left or bilateral cervical vagotomy. Since the
approval of the VNS Therapy System™ (Cyberonics Inc.) by the Food
and Drug Administration in the US in 1997, over 20,000 patients
with epilepsy have been treated with VNS therapy worldwide
(Schachter 2006).
VNS is normally implanted below the left clavicle and the
stimulating electrodes are placed around the left vagus nerve
distal to the branching of the recurrent laryngeal nerve, thereby
conveying the electrical signal produced by the generator to the
vagus nerve. The bipolar lead has two connector pins at one end,
which are plugged into the generator, and two separate helical
silicone coils at the other end. Each helix has three turns, with a
platinum ribbon electrode within the middle turn. (Schachter 2004)
The implantation procedure is done with a standardized methodology
(Reid 1990).
The generator is individually programmed to stimulate the nerve
automatically. The output current, pulse width and frequency and
the duration of
-
33
each stimulus can be adjusted according to the patient’s needs.
In addition, extra stimulation can be activated at pre-programmed
settings through a magnet passed over the generator in case of aura
or a seizure. When the magnet is left in place over the generator,
the device is inactivated. This can be used to suspend stimulation
at times when side effects would be inconvenient. (Schachter
2004)
The anatomy of the vagus nerve is discussed detail in section
2.4.1.
Mechanism of action
Despite extensive experimental studies and some human data, the
precise mechanism of action of VNS is unknown. However, in recent
years much progress has been made through neurophysiological,
neuroanatomical, neurochemical and cerebral blood flow studies in
understanding the underlying mechanisms of action of VNS. In early
animal studies VNS stimulation was shown to induce increased EEG
synchronization in non-epileptic animals, depending on the
frequency of stimulation (Zanchetti et al. 1952, Chase et al.
1967). The frequencies and output currents used in different
studies vary (Woodbury & Woodbury 1990, Zabara 1992). Naritoku
et al. were the first to identify some key structures in the
neuronal network between the brainstem and forebrain during VNS
stimulation (Naritoku et al. 1995). They showed that VNS alters
multiregional neuronal activities of the brainstem and cortex,
especially the amygdala, a highly epileptogenic region that also
plays a role in the generalization of seizures. Another study
showed that an increase in GABA or a decrease of glutamate
transmission in rats’ nucleus tractus solitarius (NTS) reduces the
severity of limbic seizures and provides a potential mechanism for
the seizure protection obtained with vagal stimulation (Walker et
al. 1999). Furthermore, in electrically kindled cats, a model for
chronic epilepsy, it has been shown that VNS delays the development
of seizures induced by electrical kindling in the amygdala
suggesting a possible preventative effect of VNS on epileptogenesis
(Fernández-Guardiola et al. 1999).
In human studies, VNS has been observed to increase cerebral
blood flow in the brain, e.g. the thalamus, the right posterior
temporal cortex, the left putamen and the left inferior cerebellum
(Ko et al. 1996, Henry et al. 1999). The increased blood flow in
the thalamus has been shown to have significant correlation with
long-term seizure control (Henry et al. 1999). Furthermore, VNS is
thought to affect A and B myelinated fibres, which may contribute
to the antiseizure effect (Handforth et al. 1998, Banzett et al.
1999, DeGiorgio et al. 2000).
-
34
Chronic VNS also appears to have an effect on various amino
acids pools in the brain. A cerebrospinal fluid study showed a
significant increase in GABA after 3 to 4 months of VNS
(Ben-Menachem et al. 1995), which may be related to the
anti-seizure effect of VNS.
Efficacy, safety and tolerability of VNS
VNS treatment has been shown to be safe, well tolerated and
effective in seizure reduction (Amar et al. 1999, Morris, 3rd &
Mueller 1999, DeGiorgio et al. 2000, Schachter 2004, Wheeler et al.
2011) Furthermore, long-term follow-up studies have shown improved
seizure control over time (DeGiorgio et al. 2000, Ben- Menachem
2002, Uthman et al. 2004). However, even after long-term treatment,
up to 25% of patients do not experience any positive effect of VNS
(Ben-Menachem 2002).
The most frequently encountered adverse effects are
stimulation-related, such as throat pain, coughing, and hoarseness,
which are usually mild to moderate in severity and resolve with
reduction of the intensity of the current or spontaneously over
time (Handforth et al. 1998, Schachter & Saper 1998, Morris,
3rd & Mueller 1999, Boon et al. 2001). Lack of the typical CNS
side effects seen with most of the commonly used AEDs is one of the
advantages of VNS treatment compared to AEDs.
2.4 Autonomic nervous system
2.4.1 Anatomy of the autonomic nervous system
ANS is responsible for visceral functions by a complex
reflectory system that provides an effective mechanism in
maintaining homeostasis and adapting to the demands of changing
external and internal conditions. Therefore, reactions of ANS are
linked to almost all the physiological and pathological conditions
of the human body, for example cardiovascular, gastrointestinal,
urinary, sexual and thermal functions. (Appenzeller 1990, Loewy
1990b)
The ANS is anatomically and functionally divided into three
distinct interacting divisions: the sympathetic, parasympathetic
and entric nervous systems (Appenzeller 1990, Iversen et al. 2000).
The sympathetic and parasympathetic nervous systems maintain
balance in the tonic activities of many visceral structures and
organs (Appenzeller 1990), while the entric nervous
-
35
system in the wall of the gastrointestinal tract is responsible
for the reflex activity involved in peristalsis and segmentation
during the passage of food through the bowel (Jänig & McLachlan
1999). Both sympathetic and parasympathetic nervous systems consist
of a chain of two neurons, which are separated from each other by
the ganglio dividing the chain into pre- and postganglionic parts.
Acetylcholine is the neurotransmitter for both sympathetic and
parasympathetic preganglionic neurons as well as postganglionic
parasympathetic neurons. Sympathetic postganglionic neurons use
noradrenaline as neurotransmitter, with the exception of the
neurons innervating sweat glands, which use acetylcholine. (Loewy
1990b)
The sympathetic preganglionic neuron cell bodies lie in the
spinal cord to form the intermediolateral cell column of the
thoracic and upper lumbar spinal cord (T1-L4). The preganglionic
neurons synapse with the paravertebral ganglio located laterally to
the spinal cord, called truncus sympathicus. The parasympathetic
preganglionic neurons arise either from nuclei in the brain steam
or from the intermediolateral cell column of the sacral spinal cord
(S1-S3). They leave the CNS via distinct cranial nerves, sacral
ventral roots and pelvic splanchnic nerves, projecting their axons
directly to the organs they supply. The postganglionic neurons are
located in small ganglia just outside or even within the wall of
the target organ. (Loewy 1990b)
The vagus nerve is the main parasympathetic efferent nerve
regulating autonomic functions. It provides parasympathetic control
of the heart, smooth muscle (pharynx, oesophagus, larynx), glands
of the viscera of the neck, the lungs (bronchial constriction and
pulmonary secretion), and the gastrointestinal system (increased
peristalsis and secretions). However, the vagus nerve is also a
mixed nerve composed of about 80% of afferent sensory fibres
carrying information arising from the head, neck and abdomen to the
brain (Loewy 1990b). Somata of the efferent fibres are located in
the dorsal motor nucleus and nucleus ambicuus. Afferent fibres have
their origin in the nodose ganglion and primarily project to the
nucleus of the solitary tract (NTS). (Boon et al. 2001)
2.4.2 Cardiovascular regulation
Under normal conditions, the sinus node is the cardiac
pacemaker. Although, the heart possesses an inherited ability for
spontaneous, rhythmic initiation of the cardiac excitation impulse,
sinus node activity is also regulated by the ANS whereas the
autonomic activity is regulated in the brainstem where the
integration
-
36
of information from higher cortical centres and the periphery is
analysed. (Benarroch E.E. 1997)
The balance of parasympathetic and sympathetic influences is
critical for control of cardiac function, including HR,
excitability and contractility. It is known that sympathetic nerve
fibres innervate the entire heart, including the sinus node,
atrioventricular conduction pathways and the arterial and
ventricular myocardium, while the vagus nerve innervates the sinus
node, the atrioventricular pathways and the atrial muscle (Kamath
& Fallen 1993). Parasympathetic activation decelerates the HR
where as sympathetic activation increases it. Furthermore, HR is
mostly influenced by the right vagus nerve that has dense
projections primarily to the atria of the heart.
The central autonomic network controls autonomic functions in a
tonic, reflexive and adaptive manner and integrates autonomic with
hormonal, behavioural, immunomodulatory and pain-controlling
responses to internal or external environmental challenges. The
central autonomic network is composed of several interconnected
areas distributed throughout the neuraxis including central nucleus
of amygdala, several nuclei of the hypothalamus and NTS. (Benarroch
E.E. 1997)
NTS is the major visceral sensory relay cell group in the brain
and it receives inputs from all the major organs of the body.
Afferents from cardiac receptors, pulmonary receptors, and
gastrointestinal receptors project to specific areas in the NTS.
Furthermore, there are specific areas in the NTS to gain
information from the carotid sinus and aortic depressor nerves that
transmit high-pressure baroreceptor and chemoreceptor afferent
information. The information is processed in the NTS and used to
affect a number of autonomic, neuroendocrine and behavioural
functions. (Loewy 1990a)
The NTS has widespread projections to numerous areas in the
forebrain as well as the brainstem including important areas for
epileptogenesis such as the amygdala and the thalamus. There are
direct neural projections into the raphe nucleus, which is the
major source of serotonergic neurons and A5 nuclei that contain
noradrenergic neurons. Furthermore, there are numerous diffuse
cortical connections. (Rutecki 1990, McLachlan 1993)
Cardiovagal motoneurons are located in the nucleus ambiguus and
dorsal vagal nucleus (Kalia 1981) and both regions receive inputs
from the NTS, the site of termination of cardiovascular and
respiratory afferents involved in cardiorespiratory reflexes
(Benarroch E.E. 1997). Nucleus ambicuus and dorsal
-
37
vagal nucleus also receive projections from the hypothalamus,
amygdala and insular cortex.
Hippocampal structures, especially the amygdala, are among the
centres at the highest level of cardiovascular autonomic control
(Frysinger & Harper 1990). Central nucleus of the amygdala
receives inputs from the NTS and from the parabrachial nucleus, and
a number of other areas and fibres from the amygdala project to the
hypothalamic area, medial parabrachial nuclei, locus coeruleus and
raphe nuclei and to the NTS and to the dorsal motor nucleus of the
vagus nerve, being thus directly involved in the autonomic
modulation of HR.
-
38
Fig. 1. Graph illustrating autonomic cardiovascular regulation.
AV=atrioventricular node, CAN=central autonomic network, DVN=dorsal
vagal nucleus, NA=nucleus ambiguous, NTS=nucleus tractus
solitarius, SN=sinus node. (Modified after Loewy 1990a and
Benarroch 1997).
NTS
Gastrointestinal Receptors
Cardiovascular Receptors
Respiratory Receptors
PulmonaryReceptors
NA, DVN=cardiovagal motoneuron
Parasympathetic nervous system Sympathetic neurvous system
HEARTSNAV
Central integration
Reflexes
VISCERAL AFFERENTS
CAN
Amygdala
Hypothalamus
HR ↓ HR ↑
-
39
2.5 Heart rate variability and its clinical implications
2.5.1 Physiological background of heart rate variability and
heart rate dynamics
HR variability is a term that is used to describe the variations
in beat-to-beat fluctuations around the mean HR. It gives
information about the sympathetic and parasympathetic autonomic
balance and other physiological control mechanisms on cardiac
function. A high variability in HR is a sign of good adaptability,
implying a healthy individual with well-functioning autonomic
control mechanism. (Task Force 1996)
Measurement of HR variability has become a widely used tool for
assessing cardiovascular autonomic function in various
physiological and pathological conditions (Lipsitz et al. 1990,
Huikuri et al. 1993, Korpelainen et al. 1996, Tulppo et al. 1996,
Tomson et al. 1998, Ansakorpi et al. 2000, Pikkujämsä et al. 2001).
Long-term, usually 24-hour ECG recording can be used to assess
autonomic nervous responses during normal daily activities in
healthy subjects, subjects with disease and in response to
therapeutic interventions, e.g. exercise or drugs. Furthermore,
24-hour ECG recordings are a non-invasive and easy approach to gain
information on cardiovascular function and HR variability, and the
measurements have good reproducibility if used under standardized
conditions (Kleiger et al. 1991).
The physiological mechanisms underlying the various measures of
HR variability differ from each other. The average HR and the
standard deviation of all normal-to-normal RR intervals over an
entire recording (SDNN) as the time domain measures of HR
variability have been found to reflect well both sympathetic and
parasympathetic influences on HR variability (Bigger, Jr. et al.
1989, Kleiger et al. 1992). Careful editing to exclude ectopic
beats, artifacts and missed beats is required to calculate SDNN
accurately (Kleiger et al. 2005).
Power spectrum analysis reflects the amplitude of HR
fluctuations present at different oscillation frequencies
(frequency domain measures of HR variability). Very low frequency
(VLF) is found at frequency 0.005–0.04 Hz. The exact physiologic
mechanism of VLF is not understood in detail, but it is suggested
that VLF power reflects the thermoregulation or vasomotor activity
(van Ravenswaaij-Arts et al. 1993). Furthermore, VLF power is
reduced by angiotensin-converting enzyme inhibition, suggesting
that it reflects the activity of the renin-aldosterone system
(Bonaduce et al. 1994).
-
40
The low frequency (LF) component is observed around 0.04–0.15 Hz
and is modulated by baroreflexes with a combination of sympathetic
and parasympathetic efferent nerve traffic to the sinus node
(Kamath & Fallen 1993, Taylor et al. 1998). An increase in LF
power has been proposed as being a marker of sympathetic activation
(Kamath & Fallen 1993), although, the parasympathetic
regulation has been reported to have an influence on the LF power
(Akselrod et al. 1985, Pomeranz et al. 1985).
High frequency (HF) is found around 0.15–0.4 Hz and it reflects
ventilatory modulation of RR intervals (respiratory sinus
arrhythmia) with the efferent impulses on the cardiac vagus nerve,
which is considered to be a marker of parasympathetic activation
(Pomeranz et al. 1985, Pagani et al. 1997).
Geometrical methods are techniques in which RR intervals are
converted into various geometrical forms. In Poincaré
scatterograms, each RR interval is plotted as a function of the
previous one. The plots can be interpreted either visually or
quantitatively (Huikuri et al. 1996, Tulppo et al. 1996). The
standard deviation of the longitudinal axis of the plot (SD2) is a
marker of long-term HR variability, while the standard deviation of
the vertical axis is a marker of short-term beat-to-beat (SD1)
(Huikuri et al. 1996). The former reflects partly physical activity
of the patients in addition to autonomic regulation of HR (Tulppo
& Huikuri 2004). The latter index is a more direct measure of
cardiac vagal outflow. In fact, SD1 is a more reliable index of
cardiac vagal activity than the HF spectral component of HR
variability, when measured from ambulatory recordings (Tulppo et
al. 1998). One advantage of the Poincaré method over spectral
analysis techniques is that it is not sensitive to stationary
irregularities and trends in RR intervals, therefore being more
suitable for HR variability analyses using ambulatory ECG
recordings (Tulppo et al. 1996).
Physical activity, emotional stimuli or reflexes of various
kinds can cause non-periodic changes to RR interval time series.
Newer non-linear dynamic methods based on chaos system theory have
been developed to detect these changes. These newer methods offer
information on the quality properties of HR fluctuation.
Detrended fractal scaling exponent, also referred to as
short-term scaling exponent α, is computed from detrended
fluctuation analysis and is a measure of the degree to which the RR
interval pattern is random at one extreme, or correlated at the
other on a scale of 3–11 beats (Kleiger et al. 2005). Short-term
scaling exponent α, has been shown to predict sudden cardiac death
in the random population of elderly subjects (Mäkikallio et al.
2001). In addition, abnormal
-
41
short-term scaling measure α has been reported to be associated
with life threatening arrhythmias in patients with myocardial
infarction (Mäkikallio et al. 1997, Mäkikallio et al. 1999b).
The slope β of the power-law relationship of HR variability
reflects the distribution of spectral characteristics of RR
interval oscillations (Saul et al. 1988, Bigger, Jr. et al. 1996).
The physiological background for the spectral distribution is not
known exactly, but the observation of a significantly steeper slope
in denervated hearts (Bigger, Jr. et al. 1996) suggests that the
autonomic nervous system has an important role in determining
long-term HR variability. Previous studies have shown that altered
long-term variability measurements predict mortality in patients
with impaired left ventricular function (Mäkikallio et al. 1999a),
after stroke (Mäkikallio et al. 2004) as well as in the elderly
(Huikuri et al. 1998).
Approximate entropy (ApEn) is a measure that quantifies the
predictability or regularity of time series data. A low value
indicates that the signal is deterministic while a high value
indicates randomness. In previous studies, physiological ageing has
been associated with a loss of ApEn (Lipsitz & Goldberger 1992,
Pikkujämsä et al. 1999) but the clinical significance has, yet,
remained unclear.
2.5.2 Factors affecting heart rate variability
Previous studies with healthy people have shown that HR
variability decreases with normal ageing being lower in elderly
people compared to middle-aged or young subjects, and these
age-related changes seem to be modified by gender (Hayano et al.
1991, Pikkujämsä et al. 1999, Fukusaki et al. 2000, Jokinen et al.
2005). It has been shown that HR variability is lower in women than
in men (Ramaekers et al. 1998, Bonnemeier et al. 2003). The gender
difference in HR variability is most pronounced in subjects younger
than 30 years, disappearing with age by approximately 50 years of
age (Umetani et al. 1998, Antelmi et al. 2004).
Many medications act directly or indirectly on the ANS and
affect HR variability measurements. Thus, the influence of
medication needs to be taken into account in the analysis of HR
variation. Atropine has been observed to abolish respiratory sinus
arrhythmia (Akselrod et al. 1985). There are also studies on the
effects of drugs on HR variability performed with antiarrhythmic
drugs, anaesthetics, sedatives and chemotherapeutic agents (Task
Force 1996). However,
-
42
further studies are needed to assess possible effect of
different drugs on HR variability.
Smoking and alcohol use reduces HR variability (Malpas et al.
1991, Lucini et al. 1996) while exercise is shown to increase it
(Tulppo et al. 1998, Hautala et al. 2009).
2.5.3 Heart rate variability in pathological conditions
The analysis of HR variability has been used widely in
quantifying risk in both cardiac and non-cardiac diseases, e.g.
stroke, diabetes mellitus, multiple sclerosis, ischaemic heart
disease, cardiomyopathy and congestive heart disease (Kleiger et
al. 1987, Ewing 1991, Bigger, Jr. et al. 1996, Mäkikallio et al.
2004). Reduced HR variability has been associated with increased
cardiac arrhythmogenic mortality in patients with various heart
diseases and overall mortality in the elderly (Kleiger et al. 1987,
Huikuri et al. 1994, Malliani et al. 1994, Mäkikallio 1996). Some
studies have also associated low HR variability with sudden cardiac
death (Barron & Viskin 1998). Reduction of HR fluctuation has
also been reported in various neurological diseases, including
stroke (Korpelainen et al. 1997), and Parkinson’s disease
(Pursiainen et al. 2002), but its clinical significance in these
settings is still undefined.
Analysis based on non-linear dynamics of HR fluctuation seems to
provide prognostic information among patients with various cardiac
diseases and reveal alterations in HR dynamics not detectable with
conventional analysis methods (Peng et al. 1995, Mäkikallio et al.
1999a, Huikuri et al. 2000). One study showed that altered
short-term fractal scaling properties of HR indicate an increased
risk for cardiac mortality, particularly sudden cardiac death, in
the random population of elderly subjects (Mäkikallio et al. 2001).
Furthermore, it has been suggested that abnormal long-term HR
dynamics predict post-stroke mortality (Mäkikallio et al. 2004).
Reduction in HR variability has been shown to be associated with
increased risk of mortality in septic patients as well (Garrard et
al. 1993, Buchman et al. 2002). However, there is currently no
consensus about the best available index of HR variability for
clinical use or risk stratification.
-
43
2.6 Epilepsy and autonomic cardiovascular dysregulation
2.6.1 Ictal autonomic dysfunction
Epileptic seizures can provoke a variety of autonomic responses
such as cardiovascular, respiratory, gastrointestinal, cutaneous,
urinary, and genital manifestations, and also emotional and sexual
feelings (Devinsky 2004). The effects of seizure discharges on ANS
are thought to be mediated through the cortical, limbic and
hypothalamic systems, and thus the seizures that arise from or
spread to areas in the central autonomic network can mimic
stimulation of autonomic afferents or modify autonomic expression
(Goodman et al. 1990, Oppenheimer et al. 1992).
Episodes of tachycardias are considered to be the most common
ECG changes during seizures (Smith et al. 1989, Leutmezer et al.
2003, Rugg-Gunn et al. 2004). Bradycardias are thought to be a rare
event (Devinsky 2004), and asystole episodes even less frequent
(Schuele et al. 2007). However, it has been suggested that ictal
bradycardias with or without asystoles are currently underestimated
(Nashef et al. 1996, Tinuper et al. 2001, Rugg-Gunn et al.
2004).
According to MRI and CT scan findings, it seems that these ECG
findings can occur in the absence of any obvious structural brain
abnormality and the presence of particular changes in HR does not
express side of the seizure focus (Reeves et al. 1996, Britton et
al. 2006). It has also been shown that seizures arising from the
temporal lobe, especially mesial lobe onset seizures, are more
prone to elicit HR changes (Galimberti et al. 1996, Tinuper et al.
2001, Garcia et al. 2001, Leutmezer et al. 2003, Britton et al.
2006). Given that in TLE seizures arise near the centres
controlling cardiovascular regulation, the high incidence of HR
changes during seizures with TLE does not surprise.
In one study, there was a higher risk of ictal ECG abnormalities
when seizures arose from sleep or from the left hemisphere when MRI
showed evidence of hippocampal sclerosis (Opherk et al. 2002).
Furthermore, in patients with epilepsy different abnormalities in
ECG morphologies during or immediately after seizures have been
observed, e.g. ST-segment depression (Opherk et al. 2002, Tigaran
et al. 2003) or elevation (Nei et al. 2000, Nei et al. 2004),
life-threatening asystoles (Liedholm & Gudjonsson 1992,
Devinsky et al. 1997, Rugg-Gunn et al. 2000, Mascia et al. 2005)
and total atrioventricular block (Tigaran et al. 2002).
-
44
There have been attempts to identify seizures by analysing HR
and HR variability measurement in patients with epilepsy. Novak et
al. found that there are detectable HR variability changes minutes
prior to the clinical onset of complex partial seizures (Novak et
al. 1999). Similarly, other studies have confirmed these findings
that seizures can be predicted using HR or HR variability analysis
(Kerem & Geva 2005, van Elmpt et al. 2006). However, these
methods are not currently in clinical use.
2.6.2 Interictal heart rate variation
During the interictal state the prevalence of cardiac
arrhythmias has been noted to be similar in patients with epilepsy
to that in the healthy population (Blumhardt et al. 1986, Massetani
et al. 1997). However, increased interictal HR has been observed in
some patients with various types of epilepsy (Evrengul et al. 2005,
Harnod et al. 2008). These changes in HR have been suggested to
occur due to alteration of autonomic cardiac function.
Using conventional short- and long-term HR variability analysis
methods, it has been shown that patients with chronic epilepsy have
dysfunction of both parasympathetic and sympathetic nervous systems
during the interictal state (Frysinger et al. 1993, Massetani et
al. 1997, Isojärvi et al. 1998, Tomson et al. 1998, Ansakorpi et
al. 2000, Harnod et al. 2008), but the clinical significance of
these findings has not been established. Although the mechanisms
leading to such autonomic dysfunction are not yet clearly
understood, it has been suggested that diminished interictal HR
variability in patients with TLE is due to the epileptic process
itself, rather than any specific AED regimen (Ansakorpi et al.
2002), but the opposite has also been proposed (Devinsky et al.
1994, Tomson et al. 1998). Overall, it is difficult to
differentiate with certainty the influence of epilepsy itself and
that of AEDs.
Previous studies indicate that antiarrhythmic drugs with sodium
channel blocking properties are associated with increased mortality
and reduced HR variability. Regarding the AEDs, CBZ which is also a
sodium channel blocker has most often been suggested to be
associated with the altered ANS function observed in patients with
epilepsy (Devinsky et al. 1994, Isojärvi et al. 1998, Tomson et al.
1998, Ansakorpi et al. 2000, Persson et al. 2003). In addition, one
study in patients with medically intractable seizures reported
increased cardiac sympathetic activity during sleep induced by
sudden discontinuation of CBZ (Hennessy et al. 2001). However, in
contrast to previous findings, one study
-
45
found that sympathetic autonomic dysfunction was less severe in
patients using CBZ or OXC compared with patients not using these
drugs (Koseoglu et al. 2009).
In TLE seizures arise from the mesial temporal structures, i.e.
the amygdala and hippocampus, or neocortical regions, and damage in
those areas may result in abnormalities manifested by altered HR
variation. This is supported by observations that hippocampus
sclerosis may be associated with decreased interictal HR variation
(Ansakorpi et al. 2004, Koseoglu et al. 2009). One previous study
found that epilepsy surgery does not affect HR variability (Persson
et al. 2006), although HR variability was reduced in epilepsy
surgery candidates before surgery (Persson et al. 2005). Indeed,
these autonomic alterations seem not to exist or are minor in early
stages of epilepsy, and evolve with time along with the epileptic
process itself. However, the role of structural brain lesions and
chronic epilepsy in autonomic dysfunction is difficult to
assess.
Impaired autonomic cardiac control in patients with epilepsy is
of particular interest considering that reduced HR variability has
been shown to predict mortality and sudden death in other
conditions than epilepsy (Binder et al. 1992). It is also
interesting to note that chronic TLE is associated with reduced
interictal HR variability (Massetani et al. 1997, Tomson et al.
1998, Ansakorpi et al. 2002, Mukherjee et al. 2009), since patients
with chronic TLE seems to be at greater risk for SUDEP. However,
there are no prospective studies regarding progression of changes
in HR variation in patients with epilepsy in the long term.
2.6.3 Circadian heart rate variation
In animal studies clear diurnal patterns of seizures have been
observed in various epilepsy models (Quigg et al. 1998). It is also
well known that night time seizures are common in some frontal lobe
epilepsies, e.g. autosomal dominant nocturnal frontal lobe
epilepsy, and myoclonic seizures in juvenile myoclonic epilepsy
occur predominantly after awakening in the morning. Furthermore,
many physiological functions, such as thermoregulation, wakefulness
and sleep, show diurnal variation (Hastings et al. 2007), and there
is also a significant decrease in arterial pressure in healthy
subjects during sleep (Millar-Craig et al. 1978).
Previous studies have mostly analysed HR variability from full
24-hour ECG recordings, although it is well established that HR and
HR variability have a circadian rhythm (Lombardi et al. 1992,
Huikuri et al. 1994). Reduction of circadian HR fluctuation has
been reported in various cardiovascular and
-
46
neurological diseases, including stroke (Korpelainen et al.
1997), diabetes mellitus (Bernardi et al. 1992), coronary artery
disease (Huikuri et al. 1994), hypertension (Chakko et al. 1993)
and Parkinson’s disease (Pursiainen et al. 2002). After an acute
myocardial infarct, suppressed circadian fluctuation seems to be
related to lethal arrhythmic events (Huikuri et al. 1992), and
infants at risk of sudden infant death syndrome show a significant
reduction in HR variability during sleep (Harper et al. 1978,
Eiselt et al. 1993).
There is one study (Ferri et al. 2002) concerning changes in HR
variation during sleep in children with partial epilepsy. The
results showed that during sleep, patients with epilepsy tended to
have an overall lower HR variability in both time- and
frequency-domain parameters, which was most evident for HF absolute
power. Therefore, LF/HF ratio was higher in patients than in normal
controls. In accordance with the previous finding, one study
reported that patients who later died of SUDEP were found to differ
from other patients with refractory epilepsy in that seizures
during sleep induced a more pronounced increase in HR than seizures
during wakefulness (Nei et al. 2004). In this regard, one previous
study observed that diminution of circadian HR variation during the
night may, in fact, be a marker for an increased risk for SUDEP
(Eppinger et al. 2004), and thus the relationship between HR
variability and altered HR variation during the day and night time
is interesting.
2.6.4 Effect of vagus nerve stimulation on cardiovascular
autonomic function
Previous studies have only reported cardiac arrhythmias during
lead tests upon VNS device implantation (Tatum et al. 1999, Ali et
al. 2004) However, some studies have also reported cardiac rhythm
changes during chronic VNS treatment (Frei & Osorio 2001, Amark
et al. 2007).
Despite the close interaction between the vagus nerve and the
heart, there are only few studies concerning the effects of VNS on
HR variability. The most important previous studies were performed
only during wakefulness and after short-term VNS treatment. One
previous study found a significant increase in HF component (Kamath
et al. 1992) while other studies suggested that VNS does not affect
HR variation (Handforth et al. 1998, Setty et al. 1998). It has
been suggested that the cardiac rhythm does not change with the
stimulation of the vagal nerve during sleep (Murray et al.
2001).
-
47
Only one study has tried to explore the long-term effects of VNS
on cardiac vagal tone (Galli et al. 2003). In that study, long-term
VNS therapy appeared to have some effects on cardiac autonomic
function, with a reduction of the HF component during the night and
a flattening of sympathovagal circadian changes.
During VNS treatment, the left vagus nerve is stimulated below
the cardiac branches of the vagus nerve, and this may explain why
the cardiac function is unaffected by routine VNS treatment.
However, further studies are needed to clarify the effect of VNS on
cardiac autonomic control.
2.7 Sudden unexpected death in epilepsy
2.7.1 Definition
There has been a lack of consensus regarding the definition of
SUDEP. The most widely used definition has defined SUDEP as a
“sudden, unexpected, witnessed or unwitnessed, non-traumatic and
non-drowning death in a patient with epilepsy, with or without
evidence of a seizure and excluding documented status epilepticus,
in which postmortem examination does not reveal a toxicologic or
anatomic cause of death”. (Nashef 1997) Due to differences in SUDEP
definitions and methodologies it is challenging to compare the
results of different studies on SUDEP.
2.7.2 Epidemiology
The risk of SUDEP is increased in the general epilepsy
population, but the reported incidence varies widely depending on
criteria and definitions, study methods, and in particular on the
type of epilepsy population under study. According to
community-based studies, the incidence of SUDEP ranges from 0.09 to
2.3 per 1,000 person-years (Terrence, Jr. et al. 1975, Leestma et
al. 1989, Ficker et al. 1998, Langan et al. 1998, Lhatoo et al.
2001). Furthermore, SUDEP i