The epileptic catastrophic encephalopathies are a group of conditions characterized by epileptic seizures that are often, but not always, intractable to treatment and associated with significant neurodevelopmental delay [1,2] The relationship between epilepsy and sleep disorders is well known and their mutual effects are widely documented [3,4]. Night seizures, but not day seizures, significantly reduced sleep efficiency [5]. In particular, SBD induce a chronic modification of sleep architecture and a state of intermittent oxygen desaturation, resulting in poor control of seizures and excessive daytime sleepiness [6]. Obstructive apnea is common in medically refractory epilepsy adult patients [7].

The comorbidity between SBD and epileptic catastrophic encephalopathies in the pediatric population is very poorly investigated in literature and rarely considered in general practice. This study aims to assess the presence of SBD in a cohort of children with catastrophic epileptic encephalopathy, characterized by seizures refractory to standard antiepileptic drugs.

Patients were enrolled in the study with respect of the following criteria: age less than 12 years; clinical and electroencephalographic diagnosis of partial or generalized crypto-symptomatic catastrophic epileptic encephalopathies characterized by drug-resistant seizures, occurring both in sleeping and waking, mental retardation and cerebral palsy; medication regimens optimized; treatment failure with at least three antiepileptic drugs given alone or in combination.

Exclusion criteria were: seizure secondary to infection, cerebral neoplasia and/or haemorrhage, significant history of medical disease (e.g. liver or kidney failure, metabolic illness) or progressive degenerative disease; non epileptic spells, alone or in combinations with epileptic seizures; another primary sleep disorder requiring medication like sedatives or hypnotics, that should affect the results of study; poor compliance to the PSG study by parents-care givers and/or children; previously diagnosis of OSAS.

Data including age, sex, body mass index (BMI), Mallampati index were collected. Computed tomography-magnetic resonance brain study, chest X-ray study and ECG were performed through all population.

Before the study, each patient underwent a detailed interview about personal history, a general clinical examination and a somatometric data collection. Parent’s or care-given were asked about patients sleep characteristic, such total nocturnal sleep time, history of excessive daytime sleepiness, insomnia, nocturnal snoring, awakenings during sleep and completed also the Brouillette questionnaire on symptoms of OSAS. According the questionnaire, OSAS was no predicted for scores less than -1, inconclusive from -1 to 3,5 and predicted >3,5. No change in antiepileptic therapy was performed during the study, or in the four previous months.

Characteristics of epilepsy like frequency of epileptic seizure, circadian rhythm of seizure, previous and current AED use were reviewed. The diagnosis of epilepsy was made by an epileptologist based on clinical history of recurrent epileptic seizures and supporting EEG data. Seizure classification and drug resistant epilepsy definition were based using International League against Epilepsy (ILAE) criteria. Drug resistant epilepsy was defined as any recurrent complex partial or generalized seizure in the past six months despite adequate use and compliance whit at least two tolerated and appropriately chosen AED in the past.

Polysomnographic recording

The polysomnograms were performed at the A.O. Ospedali dei Colli - A.O.R.N. Monaldi, sleep laboratory. Overnight sleep studies were carried out in all subject using a digital polysomnography equipment (SOMNOlab V 2.01.0001, Weinmann Medical technology, Germany). The subject was prepared about 1 h prior to the recording by attaching electrodes using average 12-18 channels. The recording time ranged from 4 to 8 h, started at the patients’ usual bedtime, and continued until spontaneous awakening. Patients have not taken sedative or hypnotic medication before the study.

All signals were digitally recorded and were reproduced on paper. At all times, patients were under supervision of a qualified technician and raw data were manually scored by 30-second epoch, according to published standards, prior to final interpretation. The PSG recording included a pressure transducer for nasal air-flow and a thermistor for nasal-oral air-flow, piezoelectric chest and abdominal belts, pulse oximetry monitoring, a snore microphone, two anterior tibial channels, submental electromyogram, electro-oculogramm, electroencephalogram (C3/A2-C4/A1 O2/A1,O1A2 of the 10-20 international placement electrode system), electrocardiogram (modified V2 lead), a body position sensor and simultaneous video monitoring.

The following sleep architecture parameters were assessed: total sleep time (TST, time from sleep onset to the end of the final sleep epoch minus time awake), time in bed (TIB), sleep latency (SL), sleep efficiency (TST/time in bed*100), percentage of each stage in total sleep time, snoring percentage. Sleep stages were visually scored following standard criteria. The respiratory parameters evaluated were: oxygen desaturation index (ODI), average oxygen saturation, nadir oxygen saturation, oxygen saturation less than 90% for more than 30% of the registration period (T 90%) and apnoea-hypopnea index (AHI).

The polysomnographic diagnosis of patients was diagnosed according to American Academy of Sleep Medicine guidelines [8]. An event was defined as obstructive apnoea when the duration was more than two breaths, the amplitude reduction in the thermal sensor was 90% or more, and the respiratory effort was continued or increased throughout the entire period of decrease airflow. An event was defined as hypopnea when the duration was more than two breath, the amplitude reduction in the nasal pressure was 50% or more, and the associated oxygen desaturation was 3% or more, or there was presence of arousal or awakening.

Apnoea-hypopnea index was defined as the number of obstructive apnoea and hypopnea per hour of the total sleep time. Obstructive sleep apnoea was diagnosed if the obstructive index was more than one on polysomnography. The severity of obstructive sleep apnoea increase with increasing obstructive index: apnoea-hypopnea index scores were classified as mild (1-5/hour), moderate (5-10/hour), and severe (>o=10/hours). Sleep hypoventilation syndrome (SHVS) was diagnosed in presence of sustained hypoxemia with arterial oxygen saturation <90% for more than 30% of the registration period, during sleep, not related to apnea or hypopnea episodes. Primary snoring was diagnosed if there are not polysomnographic evidence of sleep apnoea or hypoventilation in a patient with history of snoring.

Demographic data

The cohort of the patients were composed by eleven Caucasian subjects. Nine of them were males and two were females. There were no obese children (Table 1).

Table 1: Clinical characteristics of the 11 study patients. (M: male; F: Female, LSG: Lennox-Gastaud syndrome; PSGE: partial secondary generalised epilepsy; TSC: tuberous sclerosis complex).

Clinical results

Nine patients presented hypersomnia, morning headaches, nicturia and impaired concentration, (Brouillette questionnaire >3,5). Of these, five patients complained awakenings and two patients presented choking during sleep. Two patients showed only daytime sleepiness. Nine subjects were referred for snoring or sleep-related breathing problems.

The clinical profile of the patients including diagnosis of epileptic syndrome associated neurological symptoms. All patients had both diurnal and nocturnal complex seizures and were in therapy with a various combination of Lamotrigine, Carbamazepine, Phenobarbital, Phenytoin, Valproate and Levetiracetam.

EEG results

The epileptic syndromes included generalized encephalopathy (five), Lennox Gastaut syndrome (LGS) (one), partial secondary generalized epilepsy (PSGE) (three), tuberous sclerosis complex (TSC) (one), Malignant migrating partial seizures (one) (Table 1). The seizure type, often combined in a given patient, included spasms and tonic seizure (three), polymorphous seizures (four), tonic-clonic seizures (seven), and partial with or without secondary generalization (five). All patients presented seizures during the recording night. Seizures occurred more frequently during the stage 3 of non-REM sleep.

Polysomnographic results

All patients presented Sleep apnoea syndrome (OSAS). In particular, seven patients presented severe OSAS, two presented moderate OSAS and two patients were found to have mild OSAS. Sleep hypoventilation syndrome (SHVS) was diagnosed in five patients. Mallampati score have a strong linear correlation with AHI and hence OSAS (Table 2). Very often the episodes of apnea and hypopnea occurred during the epileptic crisis.

Table 2: The respiratory polysomnographic parameters of the 11 studypatients. (AHI: Apnoea-Hypopnea Index; Average SpO₂: AverageOxygen Saturation; Nadir SpO2%: Nadir Oxygen Saturation ODI:Oxygen Desaturation Index; Nadir SpO₂%: Nadir Oxygen Saturation;T90%: oxygen saturation less than 90% for more than 30% of theregistration period).

Seven patients showed a significant reduction of the total time in bed and of the total sleep time. All patients presented a reduction of REM sleep percentage. Apnoea/Hypopnea index and average oxygen saturation are negatively correlated with REM sleep percentage (p<0.05). Nadir oxygen saturation negatively correlates with the efficiency of sleep (p<0.05). No other significant correlation with polysomnographic data were detected.

Our study shows that paediatric patients with drug-resistant epilepsy, cerebral palsy and mental retardation are frequently affected by obstructive sleep apnoea. The severity of OSAS strongly influence the quality of sleep. Epilepsy and OSAS, when coexist, profoundly exacerbate each other’s [9].

On the one hand, OSAS has been hypnotized to exacerbate seizure frequency by disrupting REM sleep [10], on the other hand, children with epilepsy report sleep deprivation as a significant seizuresprecipitating factor and drug poor control.

In fact, a high prevalence of OSAS has been reported in drugresistant epilepsy patients [11]. In addiction, sleep deprivation in children might be associate with cognitive and behavioural dysfunction, and consequent downgrading of the quality of life [12].

Nevertheless, is very difficult to recognize symptoms associated with sleep breathing disorders in these patients, despite the significant abnormalities of the sleep architecture and the serious impairment of respiratory parameters, characterized by frequent episodes of apneas and severe nocturnal oxygen desaturation [13].

Consequently, sleep breathing disorders are underdiagnosed in patients with epilepsy and even more in children with catastrophic epileptic encephalopathy [14]. One possibility is that antiepileptic drugs increases daytime sleepiness with the influence of the sleep architecture, acting as a confounding factor. AEDs can also induce insomnia [15]. Antiepileptic medications may also contribute to sleep disorders. For example, in a patient predisposed to obstructive sleep apnea (OSAS), barbiturates and benzodiazepines may worsen the frequency of apneas and hypopneas by decreasing upper airway resistance or arousal mechanisms [16,17]. In this case, become very useful to look at the Mallampati index as demonstrated in this study. In fact, in children, a high Mallampati score is more predictive of the presence of OSAS than in adults because the main cause of the obstruction is pharyngeal [18].

Several studies have also demonstrated that an improvement in seizure frequency in refractory epileptic patients is seen whit the treatment of OSAS [19,20]. The screening for SBD to recognize an altered nocturnal respiratory function in the medically refractory epilepsy population is particularly outstanding to set up early intervention strategies [21-23]. This may lead to overall improved night oxygen saturation, seizure control, sleep quality, daytime functioning and quality of life.