Central sleep apnea in patients with pontine tegmental cap dysplasia treated with supplemental oxygen: a case report

Abstract

This is the first case report of sleep-disordered breathing in patients with pontine tegmental cap dysplasia, a very rare neurological disorder characterized by an anatomic malformation in the pons. Patients present with hypotonia, cognitive dysfunction, and cranial nerve palsies (eg, hearing loss, trigeminal anesthesia, and swallow dysfunction). Extensive studies have demonstrated the relevance of different pontine neuronal nuclei in breathing regulation, which are structurally abnormal in pontine tegmental cap dysplasia. We present detailed polysomnography data for 3 patients aged 41 years, 20 years, and 1.5 years revealing significant central sleep apnea. We discuss our experience with managing their sleep-disordered breathing in the setting of multiple cranial nerve palsies and corneal anesthesia, and its relative contraindication of noninvasive positive pressure ventilation treatment.

Citation:

Ju Wang J-D, Doherty D, Ramirez J-M, Chen M. Central sleep apnea in patients with pontine tegmental cap dysplasia treated with supplemental oxygen: a case report. J Clin Sleep Med. 2023;19(4):843–849.

INTRODUCTION

Pontine tegmental cap dysplasia (PTCD) is a very rare neurological condition characterized by a hypoplastic ventral pons with a characteristic dorsal “cap” that protrudes into the fourth ventricle. Although the cause of PTCD is not yet understood, plausible hypotheses include impaired axon guidance (as observed in mice with inactive NTN1 and DCC genes) and neuronal migration.¹

Clinical presentations differ widely, but most patients have some degree of cognitive impairment and prominent cranial nerve palsies. Several studies describe ophthalmological findings in these patients due to trigeminal nerve dysfunction, with symptoms including corneal scarring and hypoesthesia, lagophthalmos, redness, visual disturbances, impaired extraocular movements.^1,2 Treatment usually involves topical artificial tears and surgeries to prevent further eye damage. Other studies also report vestibular and hearing problems^1,3 (sensorineural deafness) and swallowing and feeding difficulties.¹

Respiratory disturbances, including apnea and sleep-disordered breathing (SDB), have been canonically associated with brainstem dysfunction. Previous literature has shown the importance of the pons in establishing the respiratory rate and pattern through connections between neurons from the parabrachial nucleus (specifically the Kolliker-Fuse nucleus next to the superior cerebellar peduncle) and the medullary respiratory groups.⁴ Thus, a disruption in connectivity between these structures might impair breathing control, which would be more evident during the sleep state due to the lack of voluntary respiration. Nonetheless, detailed sleep parameters have not been described in patients with PTCD. As far as we know, only Jissendi-Tchofo et al¹ mentioned central sleep apnea (CSA) and obstructive sleep apnea present in 2 out of 6 patients with PTCD in a case-series of radiographic findings. The prevalence, characteristics, and long-term sequelae of SDB in patients with PTCD are not known.

We present detailed polysomnography (PSG) data for 3 patients with PTCD and discuss our experience managing the complexities of treatment in the setting of cranial nerve palsies, particularly corneal anesthesia that are present in this population and its relative contraindication of noninvasive positive-pressure ventilation (NIPPV) treatment.

REPORT OF CASES

Table 1 provides detailed sleep parameters from all PSG studies performed in our institution.

Table 1.

Respiratory control abnormalities in patients with PTCD based on PSG parameters.

Case	Age (y)	Intervention	SE (%)	Arousal Index events/h	REM (%)	AHI	cAHI events/h	Mean SpO2 (%)	Nadir SpO2 (%)	Mean CO2 (mmHg)	Max CO2 (mmHg)	Time > 50 (%)	RR events/h	CA (s)	PLM events/h
1	39	Room air	71.0	36.2	0.9	28.8	22.4	95.8	82.2	43.4	46.8	0	8.8	<20	0.9
	39	O2 0.5 L/min	71.1	16.8	9	9.8	7.7	97.8	93.3	58.7	69.0	100	10.0	<10	8.6
		O2 1 L/min	46.2	5.7	0	11.4	8.9	96.5	92.6	60.0	67.0	87.7	12.3	<20	10.6
	40	Room air	78.6	7.8	10.8	22.2	15.36	95.2	89.3	54.0	59.7	94.1	11.2	6	20.6
		O2 0.25 L/min	64.0	5.6	0	4.1	3.06	96.7	90.7	57.0	57.0	100	7.0	10	22.5
		O2 0.5 L/min	68.7	4.2	0	NR	NR	97.1	95.5	54.6	65.0	100	9.0	NR	14.7
2	14	Room air	55.2	8.9	5.0	20.5	15	96.7	90.3	42.4	46.7	0	14.7	<20	3
		O2 1 L/min	38.0	12.9	23.4	14.3	5.9	98.7	95.9	43.7	45.0	0	12.0	<20	0
	19	Room air	50.8	12.7	7.8	30	28.6	92.2	85.8	46.8	48.6	0	12.0	18–24	1.9
		O2 0.5 L/min	98.6	7.1	20.5	8.8	5.3	91.5	88	48.8	51.0	0	12.0	8–13	1.8
		O2 1 L/min	92.2	10.2	28.1	8.8	7.4	93.8	89.9	50.7	55.0	55.5	10.0	8–20	0.5
		O2 1.5 L/min	84.4	4.2	35.1	5.3	5.3	94	89.6	50.3	51.0	41.2	8.0	8–10	0
3	1	Room air	53.8	8.2	4.9	13.2	11.6	96.6	91.0	42.4	49.0	0	19.5	<10	3.2
		O2 0.25 L/min	82.6	5.6	14.4	8	6.8	97.4	91.0	42.6	45.0	0	16.0	8–14	4.3
		O2 0.5 L/min	53.2	10.9	0	2.9	2.3	97.8	95.0	41.7	44.0	0	14.0	6–12	7.3

Chronological data from PSG studies for each case. Age represents the patients’ age in years when the study was performed. Time > 50 represents the percentage of time spent above 50 mmHg of CO₂. CA (s) represents the length of the CA events in seconds. AHI = apnea-hypopnea index, CA = central apnea, cAHI = central apnea-hypopnea index, NR = not reported, PLM = periodic limb movement, PSG = polysomnography, PTCD = pontine tegmental cap dysplasia, REM = rapid eye movement, RR = respiratory rate, SE = sleep efficiency, SpO₂ = oxygen saturation.

Patient 1

A 41-year-old female with past medical history of PTCD, bilateral sensorineural hearing loss, cyclic vomiting, bilateral corneal scarring in part from retrograde airflow through abnormal nasolacrimal ducts, dysphagia, scoliosis, seizures, osteopenia, cognitive delay, history of “apnea” requiring cardiopulmonary monitoring and supplemental oxygen (O₂) as a young child, and later diagnosed with “sleep apnea” was referred to the sleep clinic for specialty management of SDB related to PTCD; despite our center being a pediatric center, the adult patient was referred specifically for our pediatric center’s experience with otherwise very rare PTCD.

Detailed medical records from childhood were not available. PTCD was diagnosed based on magnetic resonance imaging (MRI) done at 29 years old, showing a flat ventral pons with a dorsal cap representing abnormally decussating axon pathways, as well as mild cerebellar hypoplasia with abnormal configuration of the superior cerebellar peduncles, hypoplasia of the middle cerebellar peduncles, and absence of the inferior cerebellar peduncles.

The patient underwent PSG at another institution and was diagnosed with severe CSA and bradypnea at 36 years of age. She started treatment with bilevel positive airway pressure (bilevel) in spontaneous/timed mode, reducing the overall apnea-hypopnea index (AHI) from 49.0 to 1.9 events/h. Nonetheless, facial asymmetry, a deviated nasal septum, nasolacrimal duct issues, and poor mask fit caused leak/pressure, which was not tolerated and produced eye irritation and corneal damage. Hence, bilevel treatment was discontinued for fear of further corneal damage and possible vision loss. The first PSG at our institution was performed at 39 years of age, showing an overall AHI of 28.8 events/h, predominantly central events (obstructive AHI [oAHI] of 6.4 events/h), with many lasting > 20 seconds (Table 1). A second split-study PSG was performed at 39 years of age, several months after the initial PSG, revealing that 0.5 and 1.0 L per minute supplemental O₂ via low-flow nasal cannula had beneficial effects, bringing down the overall AHI to 10.0 event/h and reducing the duration of the events (Figure 1). A follow-up split-study PSG at 40 years of age revealed an overall AHI of 22.2 events/h that markedly improved with 0.25 and 0.5 L/minute supplemental O₂, which also improved bradypnea and respiratory stability.

Figure 1. Eight illustrative epochs (4 minutes) of N3 sleep from the PSG performed on patient 1 at 39 years of age.

The top panel shows an abnormal respiratory pattern associated with central apneas (delimited by red boxes), bradypnea and mild desaturations without O₂ supplementation. The bottom panel shows marked stabilization of respiratory pattern and improvement of central apneic events with higher oxygen saturation. PSG = polysomnography.

After the last outpatient visit to the sleep medicine clinic, NIPPV was contraindicated due to the history of corneal desiccation injury from air leak; therefore, continued 0.25–0.50 L/minute nasal cannula O₂ during sleep was recommended.

Patient 2

A 20-year-old male with past medical history of PTCD, prior diagnosis of possible mitochondrial myopathy, cyclic vomiting, corneal ulcerations in part from retrograde airflow injury through abnormal nasolacrimal ducts, bilateral sensorineural hearing loss, and past surgical history of adenotonsillectomy at 1 year of age at a separate institution was referred to the sleep clinic for specialty management of SDB related to PTCD.

He was born at 42 weeks’ gestation and discharged at 7 days of age with ongoing concerns of poor suck, suboptimal feeding, and right facial palsy. The newborn hearing test showed no hearing bilaterally and he underwent a cochlear implant at 7 months of age. Remarkably, the clinical report for the pre-procedure brain MRI did not report brain abnormalities. Throughout the years, the patient developed severe hypotonia with cranial nerve palsies. MRI was repeated at 3 years of age showing hypoplastic pontine tegmentum. This finding was stable and seen in subsequent MRIs at 8 and 14 years old, associated with small brainstem, tissue projection in the dorsal pons, hypoplasia of the cerebellum, and enlargement of the fourth ventricle, diagnostic of PTCD.

The patient underwent several sleep studies and was diagnosed with SDB (Table 1). He was initially treated with bilevel 20/10 cm H₂O support as a toddler; this was discontinued due to recurrent corneal abrasions. The first PSG at our institution was performed at 7 years of age and revealed an overall AHI of 27.1 events/h, with predominantly central events (central apnea index[CAI]: 17.2; oAHI: 9.9). Most central apneas lasted > 20 seconds, with 9 episodes of periodic breathing (PB), which encompassed many of these central events. The respiratory rate was 12 breaths per minute and tachypnea was not observed between central events. A second O₂ titration study, performed several months after the initial PSG, demonstrated improvement in CAI from 17.2 to 4.3 with 1.0–1.5 L/minute nasal cannula O₂, associated with a mild increase in oAHI from 9.9 to 12.8 events/h. Home treatment with 1.0 L/minute supplemental nasal cannula O₂ was established. Subsequent PSG studies at 10, 13, 15, and 19 years old did not show significant qualitative changes (Figure 2).

Figure 2. Two illustrative epochs (60 seconds) of N3 sleep from the PSG performed on patient 2 at 19 years of age.

The top panel shows an abnormal respiratory pattern associated with a 26-second central apnea (delimited by the red box), associated with a 5% desaturation (delimited by the green box) without O₂ supplementation. The bottom panel shows marked stabilization of the respiratory pattern, improvement in central apneic events, and higher O₂ saturation with 0.5 L/minute of O₂ supplementation. PSG = polysomnography.

Similar to patient 1, NIPPV was contraindicated due to a history of corneal injury; therefore, 1.0 L/minute nasal cannula O₂ was recommended during sleep only, since baseline saturations were normal while awake.

Patient 3

A 1.5-year-old female with past medical history of PTCD, global developmental delay, hypotonia, corneal anesthesia, bilateral sensorineural hearing loss, facial paralysis, feeding difficulties, gastrostomy tube dependence, ankyloglossia, and left plagiocephaly was referred to the sleep clinic for specialty management of SDB related to PTCD.

She was born at 39 weeks’ gestation, and delivery was complicated by nuchal cord and respiratory distress requiring brief continuous positive airway pressure support before transitioning to blow-by O₂ (up to 100%) and room air. Of note, normal oxygen saturation level was achieved within 12 minutes after delivery. She was admitted to the neonatal intensive care unit due to poor oral feeding skills, accompanied by facial nerve palsy, corneal anesthesia, and bilateral sensorineural hearing loss. She also exhibited abnormal symmetric startle movements with coordinated extension of both upper and lower extremities. Two electroencephalograms did not reveal seizure activity, and a subsequent brain MRI showed pontine tegmental cap and small cerebellum. A G-tube was placed at 1 month of age due to poor oral motor skills.

She underwent a PSG split study at 1.5 years of age, which revealed primary CSA characterized by PB mostly during rapid eye movement (REM) and transitional sleep with a central AHI (cAHI) of 11.6, and mild hypopneas with a minimally elevated oAHI of 1.7. Events were associated with mild desaturations and no hypoventilation. Most events lasted < 10 seconds and respiratory sinus arrhythmia during PB events ranged from 60 to 100 beats per minute. On 0.25 and 0.5 L/minute supplemental O₂ CSA/PB was attenuated from 11.6 to 6.8 and 2.3, respectively. No differences were seen in mean oxygen saturation (SpO₂), carbon dioxide (CO₂), and event length.

Since the patient did not exhibit significant hypoxic burden or hypoventilation, and most of the sleep events were self-limited without causing sleep symptoms, empiric treatment with supplemental O₂ trials with 0.5 L/minute via nasal cannula was recommended, while monitoring for significant changes in overall health, energy, sleep, development, and growth.

DISCUSSION

The muscular act of respiration includes active inspiration and passive expiration. However, neurologic breathing control is more complex, consisting of (at least) 3 phases: inspiration, postinspiration, and active expiration. These phases are generated by interchanging excitatory and inhibitory signals to the pre-Bötzinger complex, postinspiratory complex, and the retrotrapezoid nucleus/parafacial respiratory group, respectively. These subregions form part of the medullary ventral respiratory column considered to be the generator of respiratory rhythm. The ventral respiratory column, along with additional components in the pons and brainstem, help to generate the signal and subsequent muscular execution of a breath to maintain eupnea.⁵ Previous pathology reports showed that the pons and medulla were both grossly and histologically abnormal in patients with PTCD.⁶ The 3 patients with PTCD presented here illustrate the adverse impact of a structural perturbation in this system. We speculate that the central apnea and bradypnea in our patients are a manifestation of a disturbance in pontine structures that are critical for the central control of breathing and that have been associated with apneas and disturbances in arousal.^7,8 The pons is an important relay for chemosensory stimuli and the hypoplastic pons would increase the gain of the system since more stimulus is needed to relay information from the peripheral and central chemoreceptors to the pons and finally to the medulla to activate the diaphragm through the phrenic nerve.

During wakefulness, ventilation is controlled by chemoreceptor sensitivity and voluntary respiratory drive. In the sleep state, voluntary breathing is abolished; thus, central and peripheral chemoreceptors alone oversee ventilation according to the varying levels of partial pressure of oxygen (PaO₂) and partial pressure of carbon dioxide (PaCO₂). This is best explained by a feedback-controller system. This system consists of a controller (brainstem/pons), a plant (lungs and respiratory muscles) that produces an output (a breath), and sensors (chemoreceptors) that then provide feedback back to the controller. The reference targets are then eupneic gas exchange. Adjustments in output based on varying input constitute one’s loop gain.⁹ Abnormal loop gain occurs when an exaggerated response from any of the components (controller, chemoreceptors, plant) results in ongoing gas exchange perturbations, further perpetuating respiratory disturbances.⁹ Interestingly, the use of O₂ supplementation in patient 1 produced an increase in the mean CO₂ levels and time spent above 50 mmHg of CO₂, which was not observed in the other 2 patients. Although we do not know the exact mechanism behind this change, we hypothesize that, due to the higher baseline CO₂ in patient 1, by stabilizing the respiratory pattern with O₂ we also blunted hypoxic drive, thus further increasing CO₂ retention.¹⁰

The conventional treatment of CSA includes mainly the use of positive airway pressure (continuous or bilevel positive airway pressure), acetazolamide, and nocturnal oxygen therapy¹¹; however, due to the range of contributing etiologies, therapeutic strategies vary considerably and the quality of evidence in the literature is sparse.¹² Most of the studies are done in patients with congestive heart failure and are less pertinent to this case series. Ghirardo et al,¹³ described a cohort of 95 pediatric patients with CSA, of whom 42% were lost to follow-up, 22% underwent neuro- and otorhinolaryngologic surgical intervention, 4% received oxygen supplementation, 3% were treated with noninvasive ventilation, 3% received acetazolamide or caffeine, and watchful observation was indicated for the remaining patients. The CAI decreased significantly in all treatment groups, including those who were just observed over time. Interestingly, patients with encephalopathy and epilepsy experienced higher CAI reductions compared with patients with achondroplasia and Down syndrome. A systematic review of nocturnal oxygen therapy showed an improvement in cAHI among short-term and long-term studies ranging from 24–84% and 34–81%, respectively. One of the short-term studies exhibited a reduction in central events from a median of 33.5 to 5.0 after oxygen therapy.¹⁴ Even though positive airway pressure is still more commonly used as a first treatment modality for CSA, new evidence regarding oxygen therapy and its positive effects may dictate a change in management of patients with CSA.¹⁵

Patients with PTCD illustrate the challenges of CSA treatment in the setting of cranial nerve palsies. Given their trigeminal nerve palsy and retrograde airflow through abnormal nasolacrimal ducts, NIPPV renders their corneas at very high risk for desiccation, which results in corneal abrasions and can lead to blindness. Thus, it is critical that treatment initiation and goals of therapy use a shared-decision–based model¹⁶ that weighs the benefit of better breathing against the risk of blindness. In our case, the families and patients stressed the importance of maintaining eyesight at all costs given the hearing deficits and other comorbidities that already have a negative impact in the patient’s quality of life. Corneal neurotization surgery is a promising intervention to restore eye sensation, lubrication, and ability to heal, which may allow for NIPPV support in the future.¹⁷ As a result, goals of care were based on avoiding hypoxia with low-flow oxygen (nasal cannula 0.25–1.0 L/minute), while foregoing the normalization of ventilatory pattern, knowing that the risk of hypoventilation may still lead to adverse outcomes, such as pulmonary hypertension¹¹ and theoretic risk of sudden death from prolonged central apneas. Fortunately, supplemental oxygen demonstrated partial resolution of CSA/PB, possibly by blunting the associated desaturations and, consequently, stabilizing blood gas fluctuations and reducing loop gain.

While clinical and radiological findings in patients with PTCD are well described in the literature, no studies to date focus on respiratory control in this population. While this report presents detailed PSG findings and therapeutic challenges in this population, more comprehensive data are needed to assess the impact and trajectory of SDB on morbidity, life expectancy, functional outcomes, and mortality associated with PTCD. Furthermore, additional work is required to clarify the optimal treatment strategies and outcomes for SDB in this challenging patient population. Given the involvement of critical respiratory control structures in the brainstem, further study of PTCD may provide key insights into key pathophysiologic pathways involved in human SDB.

CONCLUSIONS

The impact of SDB and its treatment on quality of life in patients with PTCD remains undetermined. Larger case series and repeat PSG studies with oxygen titration are needed to document the spectrum of respiratory control issues, monitor disease progression, and establish treatment guidelines. Standard bilevel therapy is not feasible in many of these patients due to the risk of corneal scarring and blindness.

DISCLOSURE STATEMENT

All authors reviewed and approved this manuscript for submission. Work for this report was performed at the Center for Integrative Brain Research, Seattle Children’s Research Institute, Seattle, Washington. This study was supported by National Institutes of Health P01HL144454 (Dr. Ramierez) and R01HL126523 (Dr. Ramierez). The authors report no conflicts of interest.

ABBREVIATIONS

AHI

apnea-hypopnea index

cAHI

central apnea-hypopnea index

CAI

central apnea index

CSA

central sleep apnea

MRI

magnetic resonance imaging

NIPPV

noninvasive positive-pressure ventilation

oAHI

obstructive apnea-hypopnea index

PB

periodic breathing

PSG

polysomnography

PTCD

pontine tegmental cap dysplasia

SDB

sleep-disordered breathing