Epileptic encephalopathy with continuous spikes and waves during sleep in children with shunted hydrocephalus: A study of nine cases

Authors


Address correspondence to Roberto Horacio Caraballo, M.D., Neurology Department, Hospital de Pediatría “Prof Dr Juan P Garrahan,” Combate de los Pozos 1881 CP 1245, Buenos Aires, Argentina. E-mail: rhcaraballo@arnet.com.ar

Summary

Purpose: We present a series of nine patients with early-onset hydrocephalus who had seizures and continuous spikes and waves during slow sleep (CSWS) associated with neurocognitive and motor deterioration.

Methods: Six boys and three girls aged 9–16 years (mean 11.3 years) were studied. [Correction added after online publication 12-Apr-2008: Number of girls and boys has been updated.] All patients underwent clinical examinations, electroencephalographic evaluations, neuroradiological imaging and neuropsychological assessment at first examination. Antiepileptic drugs (AEDs) were given in all cases and changed according to clinical and EEG evolution.

Results: Onset of epilepsy occurred from age 8 to 60 months (mean 19.6 months and median 14 months) with focal seizures with or without secondary generalized tonic–clonic seizures. Between ages 6 and 13 years (mean 10.4 years and median 8 years), hyperkinesia, aggressiveness, and poor socialization appeared in all nine cases. Reduced attention span, deterioration of language, and temporospatial disorientation were found in three of them. Negative myoclonus was found in two patients. The EEG showed CSWS. Response to change in treatment was good in all patients. None of the patients had relapses, seven of them have remained seizure free, and two continued having sporadic focal motor seizures during 2–5 years (mean 3 years) of follow-up.

Conclusion: In children with early-onset hydrocephalus, particularly with behavioral and language disturbances and/or motor deterioration, CSWS should be considered. Periodic EEG recordings during sleep should be done in these children.

The early identification of this particular electroclinical picture is crucial to start adequate treatment to avoid progressive cognitive deterioration.

The incidence, electroclinical features, and evolution of epilepsy in children with hydrocephalus have previously been studied (Piatt et al. 1996; Klepper et al., 1998; Bourgeois et al., 1999). The incidence of epilepsy in this group of patients is variable and ranges from 6% to 59% (Ben-Zeev et al., 2004).

There is no discussion about the fact that the etiology causing hydrocephalus and the association of additional neurologic deficits correlate directly with the presence of epilepsy and outcome, but controversy exists regarding the meaning of the shunt, shunt infections, and slit ventricule syndrome in the physiopathological mechanisms of epilepsy (Keene & Ventureyra, 1999).

Focal and secondary generalized seizures are the most frequent type of crises associated with shunted patients. They usually originate from the shunted hemisphere (Bourgeois et al., 1999; Klepper et al., 1998; Veggiotti et al., 1998). EEG abnormalities described in these patients include focal slow-wave activity accompanied by focal spikes and sharp waves predominantly over the shunted hemisphere. Generalized spikes and sharp waves, and hypsarrhythmia may occur (Piatt et al., 1996; Bourgeois et al., 1999). Studies regarding electroencephalographic findings during sleep are scarce.

According to the ILAE report, epileptic encephalopathy with continuous spikes and waves during sleep (EECSWS) is recognized as an epileptic syndrome characterized by a combination of focal seizures, neuropsychological impairment, and typical EEG findings with a pattern of diffuse spikes and waves occurring in >85% of slow sleep (Engel, 2006). It has been classified into idiopathic, cryptogenic, and symptomatic (Tassinari et al., 2000, 2002). The symptomatic cases include patients with unilateral polymicrogyria, porencephalic cyst, neonatal intraventricular hemorrhage, and—recently—early-onset hydrocephalus. CSWS in patients with congenital hydrocephalus was first described by Veggiotti et al. (1998).

Here, we present a series of nine patients with early-onset hydrocephalus who had seizures and CSWS associated with behavioral disturbances, and neurocognitive and motor deterioration.

Methods

From January 2000 to February 2006, nine children with shunted hydrocephalus in the first year of life, seizures, and CSWS were found and investigated. In the same period of time, 50 infants with shunted hydrocephalus were operated in our hospital. Follow-up ranged from 2 to 5 years (mean, 3 years) with repeated clinical examinations and EEGs. Age at onset, semiology, distribution, duration and frequency of the seizures, and neurological examinations were analyzed. Etiology of the hydrocephalus was also considered. Computed tomography (CT) scan and magnetic resonance imaging (MRI) were obtained in nine and three, respectively. All patients were psychometrically evaluated with the Wechsler Intelligence or Terman Merrill Scale. A formal psychological evaluation in the predeterioration period, at the peak of deterioration and upon resuming baseline was performed in five patients. A formal psychological evaluation at the peak of deterioration and upon resuming baseline was done in four patients. Since this study was done retrospectively, predeterioration evaluations were not performed in all the children. Antiepileptic drugs (AEDs) were given in all cases and changed according to clinical and EEG evolution.

EEG recordings were made with the 10-20 international system using 22 electrodes. Serial EEGs, performed during wakefulness and sleep, were obtained in all patients. CSWS was defined as the presence of continuous spikes and waves during slow sleep (CSWS). It must be underlined that CSWS was taken to mean the presence of continuous spikes and waves during at least 85% of slow sleep. Evolution was also analyzed.

Results

Six boys and three girls aged 9–16 years (mean 11.3 years) were studied. [Correction added after online publication 12-Apr-2008: Number of girls and boys has been updated.] All patients had undergone placement of a ventriculoperitoneal shunt in the first 6 months of life (range, 15–180 days; mean, 53 days). All patients had mild and moderate mental retardation, motor clumsiness was found in three, hemiparesis in two, and paraparesis in one.

Onset of epilepsy occurred from age 8 to 60 months (mean 19.6 months and median 14 months) with focal motor seizures with impairment of consciousness in two patients, and secondary generalized tonic–clonic seizures in two other patients. Interictal EEG showed focal abnormalities in the right hemisphere where the ventriculoperitoneal shunt was placed in all cases (Fig. 1). In three of them, occasional anterior spikes in the left hemisphere were found. Epileptic spasms with or without hypsarrhythmia were not found in any of the cases, and neither were autistic-like or psychotic manifestations.

Figure 1.


Interictal EEG recording shows right frontal spikes during sleep.

Since the onset of epilepsy until the time when changes in the electroclinical pattern appeared, patients had received different AEDs: Carbamazepine in three cases, valproic acid in two cases, phenobarbital in one case, and a combination of carbamazepine with phenobarbital in two and with valproic acid in one.

Brain CT and MRI showed ventriculomegaly in all cases associated with cortical atrophy in five patients, and periventricular leukomalacia in three patients.

Between ages 6 and 13 years (mean 10.4 years and median 8 years), an evident change in behavior, seizures, and EEG pattern occurred. A marked impairment of IQ and behavioral changes, such as hyperkinesia, aggressiveness, and poor sociability, appeared in all nine cases. Reduced attention span, deterioration of language, and temporospatial disorientation were found in three. The linguistic disturbances were characterized by expressive aphasia rather than verbal or auditory agnosia typical of Landau–Kleffner syndrome. Negative myoclonus with gait instability appeared in two patients.

Asymmetric bilateral spikes and spikes and waves dominant in the right side were seen in all cases. These discharges became more frequent during sleep in all patients and featured the bilateral, continuous symmetric or asymmetric pattern in voltage spike-wave activity during slow-wave sleep in five and four, respectively (Fig. 2). No unilateral continuous spike-wave activity during slow-wave sleep was found in any of the patients. Continuous spikes and waves were present during 85% of slow sleep in six patients and during 90% of slow sleep in the other three.

Figure 2.


EEG recording shows asymmetric continuous spike and wave activity in more than 85% of slow sleep.

The time between the first manifestation of deterioration and the first sleep EEG was 1 to 3 months (mean 1.5 months). After the appearance of behavior and language disturbances, and motor deterioration associated with CSWS, the AEDs were switched to ethosuximide in three, ethosuximide plus clobazam in three, and the association of ethosuximide and sulthiame in three.

Response to change in treatment was good in all patients with significant improvement of behavioral disturbances and CSWS (Fig. 3). CSWS as well as behavioral disturbances disappeared within 1 to 6 months after treatment initiation (mean time 3 months) and the EEG showed focal abnormalities similar to the period before the onset of CSWS (Figs. 4A, B). None of the patients had relapses, seven of them have remained seizure free, and two continued having sporadic focal motor seizures during 2 to 5 years (mean, 3 years) of follow-up.

Figure 3.


Interictal EEG recording shows frequent spike and slow wave in right frontotemporal areas during sleep.

Figure 4.


(A) Sleep EEG recording shows continuous spike and wave activity in more than 85% of the recording. (B) sleep interictal EEG recording shows bilateral spikes with higher amplitude in right hemisphere.

During periods of clinical changes and CSWS, abilities worsened in the various tests applied, but responses again reached their basal level after the behavior and language disturbances and negative myoclonus disappeared, with evident improvement in quality of life. However, cognitive impairment was not severe enough to affect activities of daily living, except in one case. In Table 1 we show the main clinical, electroencephalographic, and neuroradiological findings, treatment and evolution of our series of patients. In Table 2 we show the intelligence quotients pre-, during- and postdeterioration in our series of patients.

Table 1.  Electroclinical features, neuroradiological findings, treatment, and evolution of our series of patients
Patients123456789
GenderMFMMFMMFM
 Present age (years) 14 16 12 10 10 9 10 10 11
 Etiology of hydrocephalus Postmeningitis Postmeningitis Postmeningitis Posthemorrhagic Posthemorrhagic Posthemorrhagic Posthemorrhagic Postmeningitis Posthemorrhagic
 Age (days) at shunt placement 60 20 180 45 35 30 15 70 25
 Seizure onset before CSWS (months) 14 8 18 24 12 60 18 14 9
 Seizure type before CSWS MFS-CFS MFS MFS-SGTCS MFS MFS MFS-SGTCS MFS-CFS MFS MFS
 EEG abnormalities before CSWS Right temporal spikes Occasional left frontal spikes Right frontotemporal spikes Right frontotemporal spikes Right frontal spikes Right frontal spikes Right temporal spikes Occasional left temporal spikes Right frontotemporal spikes Right fronto-temporal spikes Right temporal spikes; Left frontal spikes
 Cognitive level and neurological examination before CSWS Mild mental retardation Hemiparesis Moderate mental retardation Mild mental retardation Paraparesis Mild mental retardation Clumsiness Mild mental retardation Mild mental retardation Clumsiness Mild mental retardation Moderate mental retardation Moderate mental retardation
 Treatment before CSWS VPA.CBZ VPA CBZ CBZ PHB CBZ.PHB CBZ CBZ.PHB VPA
 Age at onset of CSWS (years) 9 13 10 8 7 6 8 7 6
 Clinical changes with CSWS Hyperactive, aggressive, inattentive Hyperactive, aggressive, inattentive Hyperactive, aggressive, language disturbances Hyperactive, aggressive Negative myoclonus Language disturbances, hyperactive, aggressive Negative myoclonus Hyperactive, aggressive Hyperactive, aggressive Language disturbances Hyperactive, aggressive, inattentive Hyperactive, aggressive Social withdrawal
 Treatment of CSWS ESM, Sulthiame ESM, Sulthiame ESM, CLB ESM ESM, CLB ESM ESM Sulthiame ESM ESM, CLB
 Clinical treatment response Back to baseline Back to baseline Back to baseline Back to baseline Back to baseline Back to baseline Back to baseline Back to baseline Back to baseline
 EEG tretament response Right temporal spikes Right frontal spikes Bilateral spikes Right temporal spikes Right frontotemporal spikes Bilateral spikes Right temporal spikes Right frontal spikes Right fronto-temporal spikes Left frontal spikes
 Neuroradiological findings Ventriculomegaly White matter alterations Ventriculomegaly Cortical atrophy Ventriculomegaly Cortical atrophy Ventriculomegaly White matter alterations Ventriculomegaly Ventriculomegaly Cortical atrophy Ventriculomegaly Cortical atrophy Ventriculomegaly White matter alterations Ventriculomegaly Cortical atrophy
 Outcome No seizures Normal school Sporadic seizures Special school with good perfomance No seizures Special school No seizures Special school with good performance No seizures Normal school No seizures Special school with good perfomance Sporadic seizures Special school No seizures Special school Sporadic seizures Special school with good perfomance
Table 2.  Intelligence quotient before, during, and after CSWS disappeared in our series of patients using Wechsler Intelligence Scale for Children
PatientsBefore CSWSDuring CSWSAfter CSWS disappeared
 IQIQIQ
1655060
2Not tested5565
3655562
4Not tested6070
5706068
6696168
7Not tested5869
8Not tested6170
9605260

Discussion

All nine children in this study had common electroclinical features consistent with shunted hydrocephalus, mental retardation, and a particular type of epileptic syndrome. The hydrocephalus was due to various etiologies, and the patients were submitted to shunting in their first year of life. Seizures began by age 2 years, gradually evolving into the distinctive electroclinical pattern by age 8 years. Epilepsy with secondary bilateral EEG abnormalities developed with symmetrical or asymmetrical CSWS causing behavioral and language disturbances, and negative myoclonus. In our series, none of the patients had electroclinical features compatible with West syndrome in the first years of life as was published by Bourgeois et al. (1999) and Piatt et al. (1996). One of our patients with shunted hydrocephalus in the first year of life had West syndrome, and has not developed CSWS yet. Obviously, this patient was not included in this study. The language disturbances observed in three of our patients were different from the verbal or auditory agnosia described in Landau–Kleffner syndrome (Soprano et al., 1994; Caraballo et al., 1999a; Fejerman et al., 2007).

Our series of patients had electroclinical features compatible with EECSWS included in the group of epileptic encephalopathies as described in the ILAE report (Engel, 2001).

This is the fourth series of patients described with both shunted hydrocephalus and CSWS. The first, published by Veggiotti et al. (1998), identified six children with CSWS of 20 patients with congenital hydrocephalus recognized by a routine sleep study, and the clinical features associated with CSWS were recognized retrospectively. In the second series, three children with CSWS associated with shunted hydrocephalus had clinical features of Landau–Kleffner syndrome (Ben-Zeev et al., 2004). In the third series, CSWS in shunted posthemorrhagic infantile hydrocephalus was described in four children (Battaglia et al., 2005).

In 95% of the patients in Veggiotti's series and in all patients of our series, focal epileptiform EEG abnormalities were lateralized to the shunt insertion side. Our findings were consistent with previous data relating focal EEG changes in the shunt insertion region in patients with hydrocephalus (Laws & Niedermeyer, 1970; Ines & Markand, 1977). The epileptiform activity may be secondary to either local valve infection during insertion or a foreign body effect. This finding may suggest that shunting should be considered an important factor in the pathogenesis of focal EEG abnormalities. However, in Ben-Zeev's series, two children never underwent shunting, and in the remaining three patients, focal EEG abnormalities were found in the nonshunted hemisphere. This means that the shunt itself may not be the trigger of the EEG discharges and CSWS.

Genetic predisposition, a specific mechanism related to early-onset hydrocephalus, and/or a coexisting but unrelated brain pathology may be the cause of this clinical and electrographic picture (Veggiotti et al., 1998; Ben-Zeev et al., 2004).

Unilateral polymicrogyria may be important in the pathogenesis of the CSWS (Caraballo et al., 1999b). Thalamic hemorrhagic infarction and the subsequent disturbed corticothalamic interactions in the transformation of physiologic brain oscillations during slow-wave sleep into paroxysmal epileptic activity evolving into CSWS may also play a role (Monteiro et al., 2001).

According to Dalla Bernardina et al. (1989), the presence of CSWS in patients with focal epilepsies, independently of the type of focal seizures or whether the epileptic syndrome is idiopathic, cryptogenic, or symptomatic, may be considered the effect of a secondary bilateral synchrony (SBS) of the focal abnormalities occurring during wakefulness. This has been confirmed by Kobayashi et al. (1994), who also underlined the role possibly played by the corpus callosum in the generation of SBS in patients with a genetic predisposition to epilepsy.

In accordance with the different atypical evolutions recognized in children with benign childhood epilepsies (Fejerman et al., 2007), we were able to identify seven patients with typical EECSWS and two patients with what we called mixed forms of atypical evolutions including negative myoclonus, some language involvement and behavioral manifestations. In all these patients, the EEG recording during slow-sleep showed continuous, symmetrical, or asymmetrical spike-wave discharges. In this series of patients, we did not find any case with typical electroclinical features of Landau–Kleffner syndrome.

All our patients received AEDs for focal seizures before evolving into CSWS. It is very well known that all these drugs induce SBS (Caraballo et al., 1989, 1999b; Guerrini et al., 1995; Prats et al., 1998). Hence, the treatment indicated in our patients was to withdraw the previous drugs, and to add clobazam, ethosuximide, and sulthiame, alone or in combination. All patients improved significantly and the CSWS disappeared. The same therapeutical scheme was proposed by other authors (Lerman & Lerman-Sagie, 1995; Oguni et al., 1998; Capovilla et al., 1999). Similar results were found in our series of patients with unilateral polymicrogyria and SBS (Caraballo et al., 1999b, 2007).

According to our experience, unilateral polymicrogyria is the most frequent cause of symptomatic EECSWS (Caraballo et al., 1999b, 2007). Patients with prenatal or perinatal unilateral thalamic injuries associated with symptomatic CSWS have also been described (Monteiro et al., 2001; Guzzetta et al., 2005). Our group has published patients with a porencephalic cyst associated with CSWS (Caraballo et al., 2003). In Table 3, the differential diagnoses of focal epilepsies and CSWS are listed.

Table 3.  Differential diagnosis of focal epilepsies and CSWS
• Idiopathic focal epilepsies
 √ Atypical evolution of benign childhood epilepsy with
 centrotemporal spikes
 √ Atypical evolution of Panayiotopoulos syndrome
 √ Atypical evolution of childhood occipital epilepsy of Gastaut
• Symptomatic focal epilepsies
 √ Unilateral and multilobar polymicrogyria and CSWS
 √ Porencephaly and CSWS
 √ Shunted hydrocephalus and CSWS
 √ Thalamic injuries and CSWS
 √ Nonshunted hydrocephalus and CSWS
• Cryptogenic focal epilepsy and CSWS

Shunt malfunction and infection are the most common causes for behavioral, cognitive, and motor deterioration and seizures increase or aggravate in children with hydrocephalus. Once we rule out these complications, CSWS should be considered as a cause of these manifestations.

CSWS is more frequently associated with shunted hydrocephalus of early onset than in the general population. In children with early-onset hydrocephalus, particularly with behavioral and language disturbances, and/or motor deterioration, CSWS should be considered. Periodic EEG recordings during sleep should be done in these children.

The early identification of this particular electroclinical picture is crucial to start adequate treatment to avoid progressive cognitive deterioration.

We have not found any predictive factors for children with shunted hydrocephalus and epilepsy to develop CSWS. Neither have we found any differences between the group with shunted hydrocephalus who developed CSWS and the group with shunted hydrocephalus and epilepsy without CSWS. However, as we commented above, AEDs useful for focal seizures might be avoided.

A prospective study on the prevalence of epilepsy and the recognition of the different types of epileptic syndrome, especially EECSWS, in patients with shunted hydrocephalus should be done. It would also be important to identify predictive factors for patients with shunted hydrocephalus to develop CSWS.

Conflict of interest: We confirm that we have read the journal's position on issues involved in ethical publication and affirm that this report is consistent with those guidelines. None of the authors has any conflicts of interest.

Ancillary