Please note! This essay has been submitted by a student.
Epilepsy is a type of neurological disorder characterized by recurrent seizures over periods of time. These periods can either be: long periods (vigorous shaking) or brief periods. Paroxysmal discharges in neurons are a cause of seizures. These discharges are triggered either by a loss of inhibition (in GABAergic neurotransmitters) or excessive excitation (in Glutamatergic neurotransmitters). The above mentioned neurotransmitters activate or deactivate ion channels at the synapse. Thus a breakdown of ion channels is one potential cause of epilepsy which can be caused by genetic abnormalities or structural lesions.
Seizures are mainly divided into two main categories: generalized and focal seizures. Generalized seizures refer to seizures which spread over the right and left hemisphere of the brain. The main types of generalized seizures are convulsive or tonic-clonic seizures (previously known as grand mal) and absence seizure. Focal or partial seizure activity occurs in one part of the brain in a more localised manner. The main types of focal seizures are simple partial seizures and complex partial seizures.
The mean prevalence of epilepsy is higher in developing countries (1.5%). This contrasts with the lower prevalence found in Europe (0.52%) and in the USA (0.68%) (Reviewed by Strzelczyk et al., 2008). Although in Europe, epilepsy is more prevalent amongst adults (0.6%) than children (0.45-0.5%), this paper will focus on surgical preventions in children. Patients with intractable epilepsy are the best candidates for epilepsy surgery. Success rate is generally 62% to 71% in comparison to 14% of patients who do not undergo surgery (Edelvik et al., 2013; Sarkis et al., 2012). In some clinical cases, patients do not respond to surgery (seizure free) until several months later.
This paper will review the process of determining the chances of a good outcome from epilepsy surgery in children, since unsuccessful treatment in childhood can lead to chronic epilepsy in adulthood, accompanied with debilitating side effects. Different sources were used to obtain relevant research articles including Pubmed, Medscape and Google Scholar. Papers containing successful and unsuccessful outcomes from epilepsy surgery in children were obtained in order to make this critical review.
Although clinical diagnosis of focal seizures in children has traditionally been a difficult task, recent advances in magnetic resonance imaging (MRI), interictal EEG, and PET have facilitated easier detection of focal brain abnormalities. As a result, surgical interventions can now take place at early stages of epilepsy.
Several points need to be considered to decide whether surgery should take place at earlier rather than later stages. In many cases, adults undergoing resective surgery for the alleviation of debilitating seizures, have had a long-life history of seizures beginning in childhood and often treated with several antiepileptic drugs (AEDS). It was previously believed that children can eventually “grow out” of epilepsy, however advances in recognition of epileptic classification have now lead to better treatments and prognosis. Focal epilepsy is one the main drug resistant type of epilepsy. If unresolved, psychosocial morbidity is often involved with this type of epilepsy (chronic epilepsy). Temporal lobe epilepsy in youth can lead to significant impairments in ones ability to live an independent adult life (Ounsted et al., 1995). Therefore, early surgical intervention can improve the quality of life of an epileptic teenager and prevent morbidity caused by frequent seizures (Vasconcellas et al., 2001). If an epileptic patient has not responded to a minimum of three types AED’s over a three-year period, they are classified as having ‘medically intractable epilepsy’. However this is not applicable to young children, who experience recurrent seizures and developmental deficits. A greater number of drugs over a smaller time window is the approach in this case (Cross, 2001). One description of intractability in children is “a failure of adequate trials of two tolerated and appropriately chosen and used AED schedules, whether as monotherapies or in combination, to achieve sustained seizure freedom” (Kwan et al., 2010). There is a lack of good prognosis tools thus counteracting intractability in children remains an issue. Based on epidemiological studies, poor prognosis indicators lead to early commencement of seizures, unsuccessful response to anti-epileptic drugs, focal seizures, and lesions in the brain. Thus early intervention and cessation of seizures will allow the child to live a fulfilled life. Early onset of focal seizures in children needs prompt evaluation and surgical intervention.
Clinical evaluation of focal epilepsy can be a difficult process. Areas of the brain where recurrent seizures occur need to be identified followed by determining whether removal of the affected area will benefit rather than affect a child’s cognitive functions such as language, memory or movement. The main principles of pre-surgical evaluation include physical examination, accurate clinical history, neuropsychological assessment and neuroimaging investigation which may include interictal EEG, MRI and PET.
An important goal of neuropsychological assessment is to characterise lateralisation of function. Traditionally, the Wada test has been used to identify the dominant hemisphere for language before surgery. However advancements in neuroimaging techniques as well as neuropsychological assessment such as memory, intelligence, language, writing and reading tasks have led to a reduction in the use of the WADA test (Cross, 2001). In addition, evaluation using neuroimaging methods can highlight whether the seizures are generalized or focal in addition to whether the origin is extra-temporal or temporal and if there are seizure related lesions (Chungani et al., 1996).
Common surgical procedures for intractable epilepsy include temporal lobectomy, extra-temporal resection and hemispherectomy. Temproal lobectomy involves the removal of portions of the temporal lobe. Extra-temporal resection entails a more complex procedure where parts of the brain outside the temporal lobes are removed. Hemispherectomy is a more drastic procedure in which an entire hemisphere is removed. Prevalence of each of these procedures have been investigated by Wyllie and colleagues (1998) and are presented in table 1.
Table 1. Number of temporal, extratemporal and hemispherectomy surgeries perfomed on children and adolescents (Cross J, 2006).
Figure 2- MRI images demostrating abnormallities found in children with drug resistant epilepsy, the arrow indicates areas suitable for surgical intervention. A- Right Hippocampal sclerosis. B-Left temporal dysembryoplastic tumor. C-Left frontal focal cortical dysplasia (Cross J, 2006).
Since some neurocutaneous syndromes can also cause epilepsy, early surgical intervention should be considered. Other seizure-inducing conditions in which surgical intervention is suitable are Sturge-Weber syndrome and tubular sclerosis (Koh et al., 2000; Uri et al., 2000). Poor cognitive outcome was observed from these conditions (Chungani et al., 1996). However Koh and colleagues (2000) found that nine out thirteen children with tubular sclerosis were seizure free, post-surgery.
Good surgical procedure can influence a positive outcome. Surgical procedure involves two stages. First, implantation of intra-cranial electrodes (unilaterally or bilaterally) in combination with video electroencephalography in order to acquire interictal and ictal data. Signs of infection in the central nervous system are also monitored (Schevon et al., 2007). Second, removal of the electrodes followed by resection of the epileptic region. For some children, additional electrode implantation may be needed to ensure the focus is removed or to identify additional foci (adjacent or distant) (Bauman et al., 2005; Madhavan., 2007). As a result, further surgery may be necessary.
Focal resection can entail partial or whole removal of a lobe. Corpus Callosotomy is ideal for children experiencing akitenic, myoclonic and tonic attacks. Nonetheless these groups of children are highly unlikely to be seizure free after the procedure. Children with acquired epileptic aphasia are good candidates for subpial transection in conjunction with resection. Intervention in the seizure driving hemisphere is ideal for children with landau-Kleffner syndrome. However the minimal data available suggests that surgery is not always beneficial (Helen Cross, 2001).
Figure 3- Surgical proportion in chidren with epilepsy derived from different conditions (Cross J, 2006).
Englot and colleagues (2014) evaluated focal post-surgery outcome in a group of children. 76% of the children were free of disabling seizure (84 patients) (Engel Class 1 assessment). 61% of these children experience total seizure freedom (67 patients) (Engel class 1). 8% of the children were almost seizure-free (Engel class 2) and 7% of the children experienced improvement (Engel class 3). A further 8% of the children experienced no improvement (Engel class 4). In most of the Engel Class 2-4 children, seizures returned within the first six months post-surgery.
To identify factors influencing surgery success, outcomes were stratified (Table 2). The mean age of children who showed positive outcome (Engel class 1) was 13.3 ± 0.6 at the time of surgery, whereas children with unsuccessful outcome were younger (9.4 ± 1.2 years, p < 0.01). Moreover only 58% children who underwent extra-temporal resections experienced freedom from disabling seizures. Post-resection outcome was more successful in cases such as mesial temporal sclerosis (MTS) (100%, Engel Class 1) and tumour (91%, Engel Class 1). In contrast, the success rate in cases such as malformation of cortical development (MCD), a focal cortical dysplasia sub-type, was much lower (51%).
Table 2-Post-resective surgery outcome and associated factors (Englot. J et al 2014)
Qualitative analysis was further performed by Englot and colleagues (2014) to analyse whether persistent seizures post-surgery (Engel class 2-4) is influenced by other factors. Twenty five patients with persistent seizures were studied (Table 3). In 8 of the 25 children (32%) their persistent seizures were due to presence of residual epileptogenic tissue (Table 3). Five of the children had tubular sclerosis with multiple tubars affected. Two of the children had MTS with lateral temporal epileptogenic (dual pathology). In one distinctive case, a 15 month old baby had tubular sclerosis with distant epileptogenic zone. The child experienced frequent seizures despite intaking several anti-seizure tablets. An MRI showed multiple cortical tubers and subependymal nodules, as well as large tubers present in the right posterior frontal lobe. An EEG monitor showed that the seizures originated from the right frontal lobe. Frequent epileptiform activity was found. Following resection of the frontal tuber, seizure frequency lessened but the child continued experiencing debilitating epilepsy (Engel class 3).
The remaining 7 children out of the 25 cases were patients with Rasmussen’s encephalitis or hemimegalencephaly, resection was needed in these cases to avoid morbidity associated hemispherectomy. The case of 14 month old girl stood out from this group. She suffered from infantile spasm experiencing 10-12 seizures per day. MRI findings showed that there was severe damage and hemispheric cortical dysplasia and slight damage in the frontal lobe. An EEG monitor showed epileptiform activity in the frontal lobes and PET also showed hypometabolism in the left frontal lobe. A left frontal lobectomy was performed on the child. Post-surgery, a few months of seizure freedom was experienced however seizure frequency returned to original baseline levels (Engel class 4). Seizure activity was localized to the left hemisphere as detected by an EEG. A further hemispherectomy was performed 20 months later. At the age of 5 the child had 2.5 years of seizure freedom (Engel class 1A), with significant recovery of previously noticed right-sided hemiparesis.
Table 3. Qualitative analysis to analyse whether persistent seizures post-surgery is influenced by other factors (Englot et al., 2014).
Studies such as the one by Englot and colleagues (2014) nicely collate the factors associated with epileptic surgical outcomes. Out of 110 patients, favorable outcome was observed in 76% of the children. As stated earlier, there is a better probability of positive outcome with temporal lobe compared to extratemporal cases. Likewise, there is a better probability of positive outcome with MTS tumor compared to MCD or other pathologies. Factors associated with recurrence of seizures included additional epileptogenic zones distant from the resection cavity (32%), the likely presence of a hemispheric epilepsy syndrome (28%) and epileptogenic tissues adjacent to resection cavity (40%).
Incomplete resection of an epileptogenic zone is usually the cause of weak outcome after surgery but limiting the amount of excision is sometimes needed to prevent damaging of eloquent brain regions. As mentioned previously is important to balance the pros and cons of conservative resection. Greater neurological deficits can sometimes arise from more aggressive approaches. Fortunately in most cases the initial neurological deficit (post-surgery) improves gradually (De Oliveira et al., 2011). Using MRI techniques, Barbara and colleagues (2012) observed that motor function recovery is associated with re-activation of motor-related cortical regions, as found in two cases following rolandic epilepsy surgery. Thus, plasticity is a key factor in post=operative neurological recovery. Weighing the pros and cons of surgical imtervention is crucial. It is important to consider the potential complications of living with uncontrollable seizures against the risk of surgery induced neurological deficits
Hemb and colleagues (2008) investigated surgical recovery in a large paediatric epileptic centre and found that some children can benefit from more aggressive resection procedures without increased morbidity. In comparison poor outcome was observed in children receiving less aggressive approach. Thus, in young children with severe debilitating seizures, who have greater potential neural plasticity, aggressive resection approaches may be appropriate.
In some cases, unintended incomplete resection can occur. It is therefore important to gather multiple assessment evidence by using EEG, PET, MRI and if needed electrophysiological recordings. In some institutions 3-T MRI assessment is now part of the pre-operative evaluation (Chang et al., 2011; Englot et al., 2013; Yun CH et al., 2007). In 65% of epileptic patients, new lesions were detected by 3-T phased-array MRI images, in comparison to previous normal 1.5-T MRI image as observed by (Knake et al., 2005). Hopefully, 7-T MRI will become more available and enhance easier identification of subtle epileptogenic lesions (Breyer et al., 2010). However, for patients with unambiguous epileptic type, electrocorticography is more appropriate (Dario et al., 2014).
Dual pathology with MTS can cause persistent seizure after surgery. Spencer and colleagues (1995) claimed that dual pathology can cause ictogenesis. Gliosis, dysgenesis and hippocampal sclerois may induce continuos seizure even after complete resection of the temporal lobe. Some experts have suggested that more aggressive resection is needed in this focal epilepsy, including adittional amygdalohippocampectomy and anterior temporal corticectomy in order to obtain better results (Englot et al., 2012; Giulioni et al., 2009; Ogiwara et al., 2010). Thus if a child has epilepsy derived from the temporal lobe, it is very important to assess possible presence of dual pathology via intensive examination of seizure semiology and mesial temporal lobe appearance on MRI. If MTS is detected in conjunction with cortical lesion, the best approach would be large resection involving the mesial temporal structures (Kassiri et al., 2013; Kim et al., 2011).
Although the techniques of non-invasive assessment such as Ictal single-photon emission computed tomography (SPECT) (Figure 4) have improved, a small number of children often benefit from invasive assessments. This involves ECOG, either EEG monitoring with subdural grids with or without implanted electrodes or intraoperative recordings. Such children include those with focal seizures within a delicate area of the brain, and children with extra-temporal epilepsy whose data is consistent but an MRI does not show structural abnormality (Figure 4).
Table 4- Surgical outcome in relation to pathology and procedure (Cross J, 2006).
Figure 4- Effectiveness of localising seizures and giving good prognosis by using non-invasive an invasive approaches. Images of Ictal single-photon emission computed tomography (SPECT) (non-invasive assessment). A- Arrow indicating hyperperfusion found in a child with right frontal seizures with no abnormality on MRI. B-SPECT plus EEG data indicated the need to perform further invasive subdural recording. Thereafter it was decided to perform a frontal focal resection, the outcome was sucessful, the child experience freedom of seizure (Engel class 1 with no anticonvulsant medication) (Cross J, 2006).
In some cases, children experience seizure freedom for only a short period of time but earlier surgical intervention may alleviate these cases (Mathern et al., 1999). Jutila and colleagues (2001) analysed the long term outcome of those with an initially successful outcome (seizure-free). 63% remained seizure free on long term follow ups. However 5% of the patients experience seizure re-occurrence. It was found that patients with persistent seizure underwent surgical intervention at delayed stages. Thus effective prognosis and early intervention is important.
Classification of epilepsy surgery based on success and probability has been suggested by Doring and colleagues (1999). This involves division of those with well stablished prognosis and techniques (conventional temporal lobectomy), those with unclear prognosis (e.g. extratemporal resections, some temporal lobectomies, hemispherectomy for developmental lesions), and those with complex cases where surgery might help (subpial transection, callosal section, trials of partial resection of abnormal tissue). Sucessful outcome after epileptic surgery should be evaluated not only in terms of seizure-freedom (Engel class 1) but also from the perspective of impairements on behaviour, development, neuropsychology and quality of life.
In conclusion, advances in pre-surgical evaluations including clinical asessments and neuroimaging in combination with improvements in surgical techniques have led to a greater prevalence of positive outcome in epilepsy surgery. Long term follow ups and simple clinical scoring (Engel class) are used to assess postoperative outcomes. Surgical resection to focal regions of the brain are beneficial for children with localized drug resistant epilepsy but as suggested by children who experience persistent post-surgical seizures, a more aggressive approach may be necessary in some cases such as in patients with hemispheric epilepsy syndrome where hemispherectomy may be beneficial. However, assessments of the preoperative factors associated with a greater probability of positive outcome are still being investigated. Further knowledge of these factors in addition to technological advances in imaging techniques will further increase the likelihood of post-operative positive outcomes.