Epilepsy as a Late Complication

Fig. 18.1

MR scans of a 28-year-old woman with a history of learning difficulties at childhood and recent episodic diffuse headache that had lasted for almost 3 months as well as additional generalized seizure 1 month prior to admission revealed a chronic (no active) hydrocephalus. The patient was treated conservatively by antiepileptic drug (valproate) without any ventriculoperitoneal shunt insertion and had a good outcome during 15 months follow-up (Courtesy of Ali Akhaddar MD, Rabat, Morocco)

18.3 Influence of Etiology of Hydrocephalus Upon Epilepsy

In the current literature, there is general agreement that the etiology of hydrocephalus may play a critical role in the development of epilepsy, albeit the results are conflicting [4, 22, 24, 36, 38]. Piatt and Carlson [38] suggested that the cause of the hydrocephalus was correlated with the risk of development of epilepsy. Etiological categories of hydrocephalus are summarized as follows: hemorrhage, infection, intracranial tumor, myelomeningocele, other congenital malformations such as aqueduct stenosis, arachnoidal cyst, Dandy–Walker malformation, and idiopathic [20] (Fig. 18.2).

Fig. 18.2

(a, b) MR scans of a 19-year-old young lady with bilateral shunts and partial agenesis of corpus callosum with recent memory changes, decreased school performance, and automatisms. EEG identified temporal lobe epilepsy (Courtesy of Jogi V. Pattisapu, MD, Orlando, FL)

The highest incidence of epilepsy was found in cases who had posthemorrhagic and postinfectious hydrocephalus, etiologies known to be associated with complex brain pathology and low functional status. In 1974, Hosking [22] reported that hydrocephalus was developed secondary to either neonatal intracranial hemorrhage or meningitis. Afterwards, Blaauw [4] found that seizures associated with hydrocephalus were more frequent in the patients with shunt infections.

Notably, hydrocephalus associated with various congenital anomalies including myelomeningocele or arachnoidal cysts carried a far higher incidence of epilepsy [28]. In 1978, Lorber et al. [34] reported that 49 % of the hydrocephalic patients associated with a morphological lesion of the central nervous system (CNS) had epilepsy and they suggested that epileptic seizures were frequently seen in patients with physical or mental disabilities. Then, Noetzel and Blake [36] found that, in a long-term follow-up on 68 hydrocephalic patients, mental retardation and malformations of the CNS correlated with seizure occurrence. Also, Keene and Ventureyra [27] reported that the risk of seizures in hydrocephalic children associated with motor or intellectual impairment was increased due to underlying brain abnormalities.

In a series of ten hydrocephalic children with tuberous sclerosis and intraventricular subependymal giant cell astrocytomas, Di Rocco et al. [14] reported that seven patients underwent direct surgical excision of the lesion, but the remaining three patients underwent a ventriculoperitoneal (VP) shunting and then removal of the intraventricular tumor. In addition, Di Rocco et al. [14] found that the surgical tumor removal was followed by a significant improvement in the epilepsy and they concluded that the surgical removal of the intraventricular tumors is the most appropriate treatment in patients with tuberous sclerosis and associated hydrocephalus.

Interestingly, as a cause of neonatal epilepsy with hydrocephalus, β-mannosidosis, which results from a deficiency of β-mannosidase, is an extremely rare disorder in humans [6]. Broomfield et al. [6] suggested that it should be considered in the differential diagnosis of neonatal seizures and subsequent hydrocephalus during follow-up, whereas others reported that there was no association between occurrence of epileptic seizures and the underlying etiology of the hydrocephalus [12, 27, 42]. In these patients, clinical findings such as altered skull morphology and intractable seizures develop in the neonatal period.

18.4 Influence of Intracranial Shunting Procedure Upon Development of Epilepsy in Hydrocephalic Patients

In neurosurgery, various shunting techniques known as ventriculo-atrial (VA) and VP are the standard treatment for hydrocephalus in both children and adults. Some authors reported that patients undergoing shunt surgery are at high risk of developing epilepsy as a surgical complication, but relation of the hydrocephalus and the shunting operation with the development of epilepsy is still controversial [28, 33, 40].

Today, it is accepted that there is an increased incidence of epilepsy risk after placement of the ventricular catheter, ranging from 5 to 58 % [5, 10, 12, 23, 24, 27, 28, 43]. To date, many authors reported large clinical series of seizure disorder following intracranial shunt insertion for hydrocephalus. In 1986, Stellman et al. [43] studied a total of 202 shunted hydrocephalic children, congenital or acquired origin, and they found an incidence of seizure disorder as 39 %. Of the 207 shunted hydrocephalic patients reported by Dan and Wade [12], 9.4 % had epilepsy. Besides, Johnson et al. [24] found that 38 % of the 817 children with shunted hydrocephalus had epilepsy. In a review of 182 patients, Klepper et al. [28] reported that shunt-related epilepsy was developed in 12 % patients. In a retrospective review of 197 patients with shunted hydrocephalus, Keene and Ventureyra [27] found that 17 % of hydrocephalic patients developed seizures.

Several authors investigated the role of shunting procedure upon the development of epilepsy in hydrocephalic patients. In a review of 92 patients with hydrocephalus, Ines and Markand [23] found that the incidence of epilepsy was high in the shunted group (65 % in the shunted group, while 18 % in the nonshunted group). Retrospectively, Venes and Deuser [47] found that 24 patients of 93 patients with hydrocephalus had epileptic seizures before the shunting procedure, but epilepsy developed following the procedure in only 5 patients. Afterwards, in a study of 168 shunt-treated hydrocephalic children, Saukkonen et al. [42] found that 48 % of the patients had epileptic seizures: 22 % of patients had epilepsy prior to the shunting procedure, and 26 % had epilepsy following the shunting procedure. Moreover, Klepper et al. [28] found that 37 (20 %) of the 182 patients developed epilepsy, 15 patients (8 %) before shunt insertion, and 22 patients (12 %) after intracranial shunting.

From an etiologic point of view, some authors investigated the effect of hydrocephalus upon the development of epilepsy in shunt-treated patients. In a retrospective study of 315 shunted hydrocephalic children, Lorber et al. [34] found that only 4 hydrocephalic patients with congenital etiology had seizures before the shunt placement, while seizures were related to the shunt device in 15 patients. Then, Klepper et al. [28] reported that epilepsy developed in 37 (20 %) of 182 patients with shunt insertion for hydrocephalus due to various etiologies including posthemorrhagic (5 %), postinfectious (4 %), myelomeningocele (2 %), and aqueduct stenosis (0 %). In a retrospective study of 802 children with hydrocephalus, Bourgeois et al. [5] reported that 32 % of the patients had epileptic seizure, possibly owing to such episodes of raised ICP or the presence of a shunt device as an epileptogenic focus. Further, Kao et al. [26] found that postmeningitis hydrocephalic patients showed the highest incidence of epilepsy as 40 %, possibly due to its high shunt revision rate.

In clinical practice, the findings confirming the effect of shunting in development of epilepsy are: (a) the development of epilepsy following surgery; (b) focal discharges at the site of the shunt in electroencephalography (EEG); and (c) the existence of contralateral seizures [10, 23, 42, 47]. Besides, Ines and Markand [23] reported that all of the shunted patients who had epilepsy developed them after the shunting procedure and left-sided focal epilepsy was the most frequent focal motor seizures in the patients with shunt placement on the right side [23]. Based on their observation upon seizures involving the body side contralateral to the shunt placement, Copeland et al. [10] noted that development of seizures was possibly due to the surgical shunting procedure.

18.5 Causative Factors for Epilepsy in Patients Who Underwent Shunting Procedure for Hydrocephalus

Especially in shunted hydrocephalic children, it is commonly recognized that epileptic seizures occur as a result of shunting procedures, surgical complications due to these procedures, or the hydrocephalus itself. In this section, we will focus on increased risk of epilepsy following placement of shunt device, possibly related with the sex of patients and the age of patient at time of shunt placement, number of shunt revision procedures, shunt location (frontal, parietal) and shunt systems used, shunt malfunction, shunt infection, slit ventricle syndrome (SLVS), cortical malformation, intracranial hemorrhage, hyponatremia due to abdominal pseudocyst, episodes of raised intracranial pressure (ICP), and intracranial hypotension related with body posture in detail (Table 18.1).

Table 18.1

Causative factors for epilepsy in shunted hydrocephalic patients

Cause of epilepsy
Gender and age of patient
Number of shunt revision
Shunt location and shunt systems used
Shunt malfunction
Shunt infection
Slit ventricle syndrome
Cortical malformation
Intracranial malformation
Hyponatremia due to abdominal pseudocyst
Episodes of raised intracranial pressure
Intracranial hypotension related with body posture

18.5.1 Sex and Age of Patient at Time of Shunt Surgery

Studies suggested that there was no link between gender of the patients and occurrence of epilepsy in hydrocephalic patients [28]. On the other hand, there is now compelling evidence that age of the patient at the time of shunting procedure may be an important factor. It has been shown that children younger than 2 years of age have a high risk of epilepsy in contrast to older ones, possibly due to an increased risk of shunt malfunction [40]. As expected, early shunting as a well-known determinant of risk in cases with shunt obstruction was associated with a higher risk for epilepsy [10, 28, 40]. Accordingly, Dan and Wade [12] also found that postshunt seizures developed in 9 % of 207 patients with ventricular shunts, ranging from 15 % in infants to 7 % in patients over 50 years of age. Based on the results of their retrospective series, Noetzel and Blake [36] noted that risk factors for development of epileptic seizure in patients with shunted hydrocephalus included age at time of shunting. However, there was no correlation between the occurrence of epileptic seizures and the age of the patient at the time of initial shunt procedure [27, 38].

18.5.2 Number of Shunt Revisions

In a previous study, it has been reported that epilepsy developed in 24 % patients with shunt revision, in contrast to 6 % patients without shunt revision [12]. According to the results of a retrospective study, Noetzel and Blake [36] noted that risk factors for development of epilepsy in patients with shunted hydrocephalus included the total number of shunt revisions. Then, Johnson et al. [24] reported that a shunt revision was done in 3 % of admissions to the emergency unit of the hospital for epilepsy, and 1 % of shunt revisions was complicated with epilepsy.

In the existence of multiple shunt revisions, epilepsy is more common owing to traumatic injury to the brain tissue during the intracranial shunting for hydrocephalus [4, 10, 12, 22, 23, 24, 34, 36, 38, 40, 42, 43, 47]. Importantly, Heinsbergen et al. [20] found that patients with more than two shunt revisions have a high incidence of epileptic seizure. Especially in patients with postmeningitis hydrocephalus, higher shunt revision rates were reported compared with those due to other etiological types of hydrocephalus [26].

Nevertheless, others suggested that there was no correlation between risk of development of epileptic seizures and the number of shunt revisions [27, 28, 38, 42]. In a review of 168 shunt-treated children for hydrocephalus, Saukkonen et al. [42] found that there was no correlation between epileptic seizure and number of shunt revisions. They agree that multiple shunt revisions had no influence on the incidence of epilepsy and thus the total number of shunt revisions did not differ between the epileptic and nonepileptic groups [27, 28, 38, 42].

18.5.3 Shunt Location and Shunt Systems Used

Numerous studies have underlined that anatomic location of shunt insertion is important for the development of epilepsy [10, 12, 22, 23, 34, 36, 38, 40, 42, 43, 47]. In 1986, Dan and Wade [12] reported that 6 % of 168 cases who had shunt placement in the parietal region had epilepsy, in contrast to 55 % of 11 patients who had undergone shunting procedure in the frontal region. Nevertheless, others found that the location of the burr hole for the shunt insertion and shunt device, frontal and parietal areas, did not correlate with the occurrence of focal or generalized seizures [24, 27, 43, 47].

Further studies investigating the role of the shunt systems as a foreign body upon development of epilepsy revealed that there is no difference between the epileptic and nonepileptic groups [28, 40].

18.5.4 Shunt Malfunction

It is logical to suggest that shunt malfunction may be related with epileptic seizure (Fig. 18.3). In a review of 200 hydrocephalic children, 10 patients had a seizure due to a blocked shunt device [22]. Faillace and Canady [18] retrospectively reviewed 15 patients with hydrocephalus who had an epileptic seizure at the time of shunt malfunction and they found that there had been no history of epilepsy in 8 patients. They suggest that as a rule, shunt malfunction should be considered, if a new or recurrent epileptic seizure develops after shunt insertion for hydrocephalus [18].

Fig. 18.3

CT scan of a 4-year-old girl with acute symptoms of headaches and generalized seizure presenting with enlarged ventricles due to shunt malfunction. The last surgical procedure was the original shunt insertion which was performed 3 years earlier (Courtesy of Jogi V. Pattisapu, MD, Orlando, FL)

On the other hand, epilepsy was not generally associated with shunt malfunction in some series [4, 19, 27, 34, 36, 38, 42, 43]. Thus, they concluded that the existence of epileptic seizure alone was not a reliable indicator of a shunt malfunction [19].

18.5.5 Shunt Infection

Numerous retrospective studies reported that the risk of development of epileptic seizures was significantly increased in cases with shunt and/or cerebrospinal fluid (CSF) infection [10, 22, 24, 36, 37, 40, 43, 45]. In a review of 168 shunt-treated hydrocephalic children, however, Saukkonen et al. [42] found that there was no link between epileptic seizure and existence of shunt infection. Likewise, Piatt and Carlson [38] reported that there was no correlation between risk of development of epileptic seizures and a history of shunt infection. No matter in what way, shunt infection should be considered as a general rule, if an epileptic seizure develops after shunt insertion for hydrocephalus.

18.5.6 Slit Ventricle Syndrome

Typically, SLVS, which is characterized by very small (“slit-like”) ventricles in computed tomography (CT) or magnetic response imaging (MRI), occurs as a result of collapse of the ventricles due to overdrainage of the CSF in minority of patients after shunt placement or revision (Fig. 18.4). As a cause of epilepsy after shunting, it was observed in only three of 182 patients with hydrocephalus, corresponding with the 0.9–3.3 % incidence in the current literature [41, 42]. After shunting procedure, epilepsy developed in 44 % of patients in the SLVS group, in contrast to 6 % of those in the non-SLVS group [41]. Out of 141 hydrocephalic patients treated with shunting, epilepsy developed in 31 those with SLVS, but 7 those with normal or dilated ventricles during the follow-up period [42]. More importantly, the same authors found that epilepsy decreased in patients with the SLVS after treatment [41]. Thus, one may suggest that serial EEG evaluation is useful in the follow-up of the patients after shunting.