| Cardiovascular regulation in epilepsy with emphasis on the interictal state | ||
|---|---|---|
| Prev | Chapter 2. Review of the literature | Next |
Epilepsy may be associated with a variety of changes in the ANS function. During acute epileptic seizures, there may be autonomic symptoms present and epileptic seizures may manifest themselves as ANS dysfunction only (Marshall et al. 1983, Gilchrist 1985, Wannamaker 1985, Blumhardt et al. 1986, Schraeder & Lathers 1989, Vaughn et al. 1996, Messenheimer et al. 1997). It is also known that long term ANS abnormalities may arise in epilepsy but the clinical significance of these signs and symptoms of chronic dysregulation of ANS function is not completely understood (Kälviäinen et al. 1990, Frysinger et al. 1993, Devinsky et al. 1994, Faustmann & Ganz 1994, Massetani et al. 1997, Tomson et al. 1998, Druschky et al. 2001). In addition, means to treat or prevent them are not yet known. Although the significance of subclinical ANS dysfunction is incompletely understood, it may sometimes be associated with events that, under unfavourable circumstances, may even be fatal to a patient.
Various symptoms and signs of ANS dysfunction may be present during epileptic seizures. These common and dramatic cardiorespiratory accompaniments have raised the question of how the patients can survive such an overwhelming event. Nonetheless, almost invariably the patients awake without sequlae.
Autonomic changes during epileptic seizures have been studied during electroshock therapy (ECT) (Brown et al. 1953). In this study patients received only subparalyzing doses of curare so that the convulsions could be monitored. At the onset of the ECT response, the avarage systolic and diastolic BP decreased 20 to 50 mmHg below control and in contrast to expected tachycardia in response to fall of BP, the HR slowed by over 50 beats/min. At the same time respiratory arrest appeared and lasted an avarage of 52 seconds. As convulsions continued, both the BP and HR increased only to fall again below control levels after the convulsions ceased. This second decrease in BP was followed by yet another elevation above control level, which gradually waned. Bradycardia persisted throughout the postictal period. In a second subgroup of experiments, subconvulsive shocks failed to produce apnea, but the variations of BP and HR was observed. (Brown et al. 1953) Later, other investigators have confirmed these ictal cardiorespiratory changes, which are correlated with generalized seizures, even when clinical seizure activity is prevented by neuroblocking agents (Mosier et al. 1957, McKenna et al. 1970).
Especially temporal partial seizures are associated with autonomic dysregulation. Seizures may only manifest themselves as burping, pallor or blushing, feeling of fear or panic, rising gustatory sensation including dyspepsia and cardiac arrythmia or apnea. Based on various studies of focal temporal epileptic discharges and simultaneous ECG-recording it seems that especially tachyarrhytmias are a common phenomenon and that localized discharges in the temporal lobe and medial frontal cortex are often associated with major cardiorespiratory and cardiovascular changes similar to changes during generalized seizures (Marshall et al. 1983, Gilchrist 1985, 1996 Wannamaker 1985, Blumhardt et al. 1986, Schraeder & Lathers 1989, Vaughn et al 1996, Messenheimer et al. 1997) It is noteworthy also that accleration of HR often occurs around the time or even before the earliest scalp electroencephalographic or clinical change (Zijlmans et al. 2002).
Typical abnormal function of the ANS occurs with prolonged seizures. Hypotension develops in the course of a prolonged seizure and neurogenic pulmonary oedema is a well described phenomenon in status epilepticus (Lathers et al. 1997). Especially status epilepticus may be associated with ECG changes within 48 hours of the onset of the ictal phenomenon, and ischemic patterns and QT interval prolongation are the most typical findings. In an autopsy series of status epilepticus patients myofibrillar necrosis of the heart, a sign of high catecholamine state, was a common finding (Boggs et al. 1993).
The long term alterations of autonomic cardiovascular regulation may be subtle and only manifest themselves as changes in HR and BP regulation. Based on experimental work, it has been concluded that even minimal epileptogenic activity can be associated with altered cardiac neural discharge and arrhythmias (Lathers et al. 1997). It is possible, therefore, that even subclinical epileptogenic activity alters the function of different areas of the brain, resulting in changes in HR and BP and cardiac neural discharge as cardiovascular regulation is, in fact, a function of neuronal activity in the cerebral cortex, the amygdala and the medullary reticular formation.
There are only a few previous studies on interictal autonomic function in epilepsy. Devinsky and his colleques (Devinsky et al. 1994) found essentially normal autonomic function in patients with epilepsy using a conventional pattern of autonomic reflex testing. However, epilepsy patients showed a greater variability in BP and HR in response cardiovascular reflex test procedures than the control subjects. This was thought to be partly attributable to CBZ treatment. In a recent study, a time and frequency domain analysis of HR variation and post-ganglionic innervation of the heart by the means of [123I]metaiodobenzylguanide (MIGB)-SPECT was studied in TLE patients (Druschky et al. 2001). The results revealed predominant parasympathetic activity in patients with TLE compared to the control subjects and significantely decreased cardiac MIGB uptake, reflecting altered post-ganglionic sympathetic innervation of the study patients compared to the control group. Faustmann & Ganz studied the structure of HR dynamics by estimation of the Largest Lyapunov Exponent (LLE) in patients with idiopathic generalized epilepsies (Faustmann & Ganz 1994). HR dynamics in patients with normal EEG did not differ from that of the control subjects, whereas patients with interictal epileptiform discharges in EEG had significantely lower LLE of the HR dynamics. Frysinger and his co-workers studied spectral analysis of HRV in TLE patients prior to resective surgery (Frysinger et al. 1993). TLE patients had significantely lower low frequency bands of HRV compared to the control subjects. However, these study patients were confined to bed, whereas the control subjects continued with normal daily living during the ambulatory ECG recording, which may have affected the results. Cardiovascular reflex tests were performed in patients with progressive myoclonus epilepsy, a neurodegenerative disease characterized by generalized tonic clonic and sometimes absence seizures and progressive myoclonus (Kälviäinen et al.1990). The majority of these patients complained various symptoms suggesting ANS dysfunction. Yet, HRV during deep breathing was the only significantly altered parameter in this study. However, it was concluded that this might be an early sign of ANS dysfunction in these young adult patients.
Spectral analysis of HR variability from ambulatory ECG and after certain stimuli (e.g. supine position and passive tilt position) was studied by Massetani and his co-workers in TLE patients (Massetani et al. 1997). The results of this study suggested that patients with TLE had a significant decrease in the total HRV in the supine position, and of the low frequency/high frequency ratio in the orthostatic position. These findings seemed to be associated particularly with right sided epileptic focus in EEG. Moreover, no correlation between the results and pharmacotherapy was seen. Finally, spectral analysis of HRV from ambulatory 24-hour ECG recording was also studied by Tomson and his colleagues in patients with JME and TLE (Tomson et al. 1998). In this study, patients with TLE had significantely lower values of various measures than their control subjects. Also patients with JME had decreased value of low frequency/high frequency ratio compared to the control subjects reflecting the predominance of altered HRV in TLE. Patients with CBZ medication also had decreased values of the spectral analysis of HRV. However, most of the patients with CBZ had TLE, and therefore the authors were not able to conclude whether the observed changes were more likely due to TLE or CBZ.
Based on the above described studies altered autonomic regulation seems to be present in patients with epilepsy. However, it is not yet known how these observed changes affect the well being of an individual or whether these abnormalities progress with the continuation of epilepsy.
The increased mortality of patients with epilepsy compared to the general population is partly due to the co-morbidity in epilepsy since there is a variety of progressive diseases that lead to epilepsy and eventually to death (Hauser et al. 1993, O´Donoghue & Sander 1997). However, sudden death is substantially more common in patients with epilepsy than in the general population. Its incidence varies from 1/100 patient years in patients with severe intractable epilepsy to 1/1000 patient years in patients with well controlled epilepsy (Hirsch & Martin 1971, Leestma et al. 1984, Leestma et al. 1989, Devinsky et al. 1994, Donoghue & Sander 1997, Nashef et al. 1998, Nilsson et al. 1999, Sperling et al. 1999, Walczak et al. 2001, Langan et al. 2002). In the most comprehensive population-based study so far published the incidence of SUDEP was 0.35/1000 person-years, exceeding the expected rate of sudden death in the general population by nearly 24 times (Ficker et al. 1998). In Finland, there are no statistics available to survey the incidence of this phenomenon.
SUDEP is defined as sudden, unexpected, witnessed or unwittnessed, non-traumatic and non-drowning death of an epilepsy patient with or without evidence for seizure and excluding documented status epilepticus. Autopsy may reveal pulmonary oedema and congestion of other organs, as well as scattered microscopic injury of the heart muscle, but not the cause of death (Earnest et al. 1992, Devinsky et al. 1994, Nashef 1997, Shorvon 1997, Nashef et al. 1998, Sperling et al. 1999). The risk factors for SUDEP seem to be multiple. SUDEP has been suggested to affect mostly young patients, aged 20-40 years, to associate with alcohol abuse, psychiatric co-morbidity, related medication, noncompliance and male gender in uncontrolled studies, but later it has been shown that these risk factors are not mandatory, but a healthy, compliant patient may also die suddenly (Tennis et al. 1995, Hanna 1997, Leestma et al. 1997, Nashef 1997, Nashef et al 1998, Nilsson et al. 1999, Walzcak et al. 2001). Moreover, it has been suggested that carbamazepine may be associated with increased risk for SUDEP, but this has not yet been confirmed (Kennebäck et al. 1997, Timmings 1998). However, refractory epilepsy and polytherapy with AEDs seem to be important individual risk factors as shown in few case-control studies (Shorvon 1997, Nilsson et al. 1999).
The pathofysiology of SUDEP is not yet completely understood. It has been suggested that the utmost pathology lies on cardiorespiratory dysregulation that predisposes a patient to potentially fatal cardiac arrythmia and central hypoventilation or apnea. The few experimental studies published so far on SUDEP have not given any conclusive results. (Schraeder & Lathers 1989, Oppenheimer & Cechetto 1990, Oppenheimer et al. 1990, Oppenheimer et al. 1992, Johnston et al. 1995, Johnston et al. 1997, Lathers et al. 1997, Nashef et al. 1998)
One possible mechanism for autonomic dysfunction and possible risk of SUDEP may relate to the experimental finding in which simultaneous recordings of cardiac autonomic neural discharges and cerebral epileptiform discharges revealed a “lockstep phenomenon”, defined as the occurrence of cardiac sympathetic and vagal cardiac neural discharges intermittently synchronized with epileptogenic discharge (Lathers et al. 1983, Lathers et al. 1987). With the disappereance of constant interdischarge intervals (e.g. stable lockstep phenomenon) and the apparance of variable interspike intervals (e.g. unstable lockstep phenomenon), precipitous changes in BP and incidence of ECG changes occurred more frequently (Stauffer et al. 1989). It has been postulated that the development of abnormal rhytmic activity of the unstable lockstep phenomenon may alter neurotransmitter release and initiate autonomic dysfunction thereby possibly contributing to the occurrence of SUDEP (Lathers et al. 1997).
It is well known that neuroanatomic connections between the brain and the heart provide links that allow cardiac arrhythmias to develop in response to activation of discrete areas in the brain. The neural aberration may then initiate other biological events, such as secretion of catecholamine that may contribute to the induction of cardiac arrhythmias or damage (Lathers et al. 1997). In one experimental work, epileptiform discharge in hypothalamic and mesencephalic neurons triggered various alterations of conduction system of the heart and bundle branch blocks (Mameli et al. 1993). In humans, on the other hand, various alterations in cardiovascular and respiratory regulation have been demonstrated in association with epileptogenic discharges and the phenomenon of apnea in association with seizures is well known to occur in children, especially in association with bradycardia (Lathers et al. 1997). It can be concluded that risk factors of SUDEP appear to be multiple, and a generalized autonomic storm leading to death, will have both sympathetic and parasympathetic effects.