Obstructive and central sleep apnoea in a patient with medically intractable epilepsy

Abstract

A woman in her 30s with medically intractable epilepsy and Lennox-Gastaut Syndrome on multiple antiseizure medications and with a deep brain stimulator presented to the epilepsy monitoring unit with increased seizure frequency. She was noted to have periods of apparent apnoea time linked to bursts of epileptiform activity on continuous video EEG monitoring. Once the clinical seizures were controlled, she was discharged to the sleep laboratory. She was noted to have obstructive and central sleep apnoea, which improved with the use of positive airway pressure. Central sleep apnoeas were time linked to electrographic seizures. Ictal central apnoea can easily be overlooked and is likely more common than currently recognised in patients with epilepsy. Ictal central apnoea may be a biomarker for sudden unexpected death in epilepsy.

Background

Sleep-disordered breathing commonly occurs in patients with epilepsy. Obstructive sleep apnoea (OSA) is two times more common in people with epilepsy as compared with the general population¹ and central sleep apnoea (CSA) is likely underdiagnosed in this population. Ictal central apnoea has been noted in 40%–48% of non-generalising focal seizures.^2–4 These ictal apnoeas are noted in patients with chest and/or abdominal plethysmography, which may have otherwise been missed. Long-term monitoring electroencephalography (EEG) in the epilepsy monitoring unit (EMU) rarely involves the use of chest and abdominal plethysmography, meaning that ictal central apnoea can easily be overlooked.

Paroxysmal asphyxia was first described by Jackson⁵ in which he noted patients and monkeys ‘turning blue’ during their seizures, despite absence of stiffness and followed by a long breath. He further recommended careful observation of these patients, including chest wall and abdomen movements. The apnoeas seen by Jackson in these monkeys may have been central in origin. Central apnoeas occur exclusively in focal epilepsy^{3 4 6} and have even been the main manifestation of focal seizures.^7–11 This is of particular importance as ictal apnoeas may be a biomarker for sudden unexpected death in epilepsy (SUDEP).

Case presentation

A woman in her 30s with a medical history significant for medically intractable epilepsy and Lennox Gastaut Syndrome status post vagal nerve stimulator, no longer active and deep brain stimulator (DBS) on multiple antiseizure medications, chronic hyponatraemia due to syndrome of inappropriate antidiuretic hormone from chronic oxcarbamazepine (OXC) use partially responsive to fluid restriction (now with low normal sodium levels), and hypothyroidism well managed on levothyroxine was admitted to the EMU due to increased seizure frequency. Her mother reported thirty to forty ‘drop’ seizures at home over the prior 6 days as compared with her baseline one to two per day. There was no apparent cause for this increase in seizure frequency. She had been consistently taking her medications (confirmed by optimal blood levels of her antiseizure medications) and had no new sleep issues, no infectious symptoms, and no alcohol or illicit drug consumption.

The patient had her first seizure as a teenager. Seizures were with focal onset (staring), followed by secondary generalisation. After multiple medication titrations, she was seizure free for several years on OXC and levetiracetam. She was weaned off of OXC several years later and had a generalised motor seizure. She was placed back on the OXC with no further occurrence of the generalised motor seizures. She continued having generalised non-motor seizures consisting of her eyes rolling back, moaning and head drop followed by postictal confusion up to 5–6 times per day. Over the years, she developed generalised motor atonic seizures as well. She was trialled on several additional antiseizure medications, including valproic acid (VPA), carbamazepine, topiramate, clonazepam, lamotrigine, phenytoin, zonisamide, lacosamide, felbamate, rufinamide and the ketogenic diet. She was also implanted with a DBS. When she presented to the hospital, she was on clobazam, brivaracetam and oxcarbazepine as well as depo provera as she was noted to have a catamenial component to her seizures.

Investigations

She was admitted to the EMU for closed circuit video EEG (vEEG) and was found to have multiple clinical seizures while awake consisting of tonic upward eye deviation. Electrographically, the seizures were associated with a generalised desynchronisation of the background followed by paroxysmal fast activity seen maximally in the left parietal chain. The seizures lasted 10–20 s. In addition, innumerable ‘subclinical’ seizures were recorded during sleep, characterised by bursts of generalised paroxysmal fast activity followed by generalised rhythmic delta activity lasting 2–30 s. After extensive review of the video, it was discovered that during the paroxysmal fast activity, there were apparent episodes of apnoea as evidenced by no chest rise on the video monitoring. During this time, she also had no other evidence of muscle artefact on EEG or clinical seizure activity, including no stiffening of limbs, making the apnoea unlikely to be secondary to a tonic seizure. There was no associated desaturation during these episodes.

Once her clinical seizures were controlled with the addition of VPA, the patient was transferred to the sleep laboratory for further evaluation of suspected sleep disordered breathing with video polysomnography (PSG). Snoring was noted and OSA was suspected. Her body mass index at the time of admission was 27.1 kg/m² and her Mallampati score was 1. Notably, she had previously been evaluated for this several years prior at a lower weight and was noted to have an Apnoea–Hyponoea Index (AHI) of 24/hour, but positive airway pressure (PAP) was not indicated at that time due to all events being CSAs associated with epileptiform activity on video-EEG. At that time a decision was made by the sleep medicine team to refer back to the epilepsy team for optimisation of her seizure disorder, which proved to be difficult despite the patient’s adherence to treatment as determined by optimal drug blood levels on monitoring and close follow-up to all appointments. This time on repeat PSG, severe OSA and CSA were noted with an AHI of 40.5/hour, indicating severe OSA and CSA. Continuous positive airway pressure was started with resolution of obstructive events. However, CSA persisted, which was linked to ‘epileptiform bursts’ as can be seen in figure 1. On careful review of these bursts, these were the same ‘subclinical seizures’ seen in her vEEG. Clinically, she had absence of flow and absence of chest or abdominal excursions on plethysmography for greater than 10 s following the paroxysmal fast activity as seen in figure 2. No other clinical or electrographic seizures were observed.

Figure 1.

An example of the patient’s central apnoea. Shortly after the onset of the paroxysmal fast activity, there is absence of flow, and notably there is also absence of chest or abdominal excursions, indicating a central apnoea.

Figure 2.

An example of the patient’s ‘subclinical’ seizure. Electrographically these consisted of generalised paroxysmal fast activity followed by rhythmic delta activity on the left. Clinically, there is absence of apparent chest rise on vEEG timelocked to the paroxysmal fast activity. vEEG, video EEG.

Treatment

Given the challenges, a decision was made to treat with a PAP mode effective for both obstructive and central apnoeas, despite lack of data on the use of adaptive servo ventilation (ASV) specifically for treatment of ictal central apnoeas. The patient was effectively titrated. A few months after discharge from both the EMU and the sleep laboratory, she was started on cannabidiol. The VPA and clobazam were weaned down. She currently continues on clobazam, cannabidiol and OXC.

Outcome and follow-up

The patient remains seizure free for several years now. The treatment intervention involved a combination of medication titration and effective treatment of OSA and ictal central apnoeas.

Discussion

Ictal central apnoea can easily be overlooked and is likely more common than is currently recognised in patients with epilepsy. Ictal central apnoea has been noted in 40%–48% of non-generalising focal seizures.^2–4 This is noted in studies using chest and/or abdominal plethysmography and these ictal apnoeas may not have otherwise been noted. It has also been the main manifestation of focal seizures in case reports.^7–11

Multiple studies have demonstrated that the amygdala is implicated in ictal central apnoea. Nobis et al looked at seven patients with temporal lobe epilepsy, implanted in the amgydala among other locations for surgical stereotactic EEG evaluation. The amygdala was stimulated during functional mapping and induced apnoea in all seven patients. Apnoea was not induced with stimulation of other sites already implanted for functional mapping.¹² In a similar study, Dlouhy et al looked at three patients using stereotactic EEG monitoring and noted that all patiets had apnoea during stimulation of the amygdala with associated desaturations and all of the patients were amnestic to the apnoea with no feeling of dyspnoea. This is of particular interest, as voluntary breath holding for the same duration caused severe dyspnoea indicating that these apnoeas are occurring without consciousness.¹³ An apnoea of this duration would result in discomfort and ultimately causing the patient to take a breath in a conscious patient as a result of an increase in carbon dioxide and decrease in the oxygen saturation. Therefore, patient report of apnoeas is not reliable, so the clinician must be cognizant of the possibility of apnoeas and be looking for them. Lacuey at al also noted that electrically stimulating multiple cortical structures in patients undergoing stereotactic EEG before epilepsy surgery elicited a transient central apnoea in 13/19 patients with stimulation of limbic and paralimbic structures (amygdala, hippocampus head and body, anterior hippocampal gyrus and anteromesial fusiform gyrus). The presence of apnoea in the rest of the patients was difficult to determine due to presence excessive artefact.^{14 15}

Ictal central apnoea is of particular interest as it may be a biomarker for SUDEP. In 2013, there was a systematic retrospective survey of 147 EMUs across the world looking at SUDEP in EMUs. There were 16 reported cases of SUDEP, 9 near SUDEP and 4 deaths from other causes noted. Cardiorespiratory data was available for 10 of the cases, demonstrating a consistent pattern of a generalised tonic clonic seizure, followed by tachypnoea. Within 3 min, the patient would develop either transient or terminal apnoea. In the cases of transient apnoea, the apnoea recured within 11 min. Terminal apnoea was followed by cardiac arrest and death. The conclusion of this study was that ictal apnoea may be a biomarker for SUDEP.¹⁶

In this case, the patient was treated successfully with antiseizure medication adjustments as well as ASV, leading to seizure freedom and significant improvement in quality of life. This is the first report of use of ASV for treatment of ictal central apnoeas in a patient with refractory seizure disorder. Additional studies are needed to assess impact on patient-centric outcomes including SUDEP.

Patient’s perspective.

The patient is not able to articulate her perspective well, but her mother was able to provide updates, and this is what she had to say. ‘What I can say is that the biggest change I have noticed since she started using the bi-pap machine is that she has been able to skip the afternoon nap, which for years had been a necessity. This has allowed her to be able to participate in activities and actively participate in life again. She wakes up in the morning and the first thing she tells me is that she slept good. So, I do think the bi-pap machine has had a positive effect towards her general health.’

Learning points.

• Ictal central apnoea can easily be overlooked and is likely more common than is currently recognised in patients with epilepsy.

• Long term EEG monitoring should involve chest and abdominal plethysmography.

• Patient report is unreliable as patients are amnestic to the event with no perceived dyspnoea.

• Ictal apnoea may be a biomarker for sudden unexpected death in epilepsy.

Ethics statements

Patient consent for publication

Consent obtained from parent(s)/guardian(s).