Obstructive sleep apnea as a presentation of congenital central hypoventilation syndrome

Abstract

Congenital central hypoventilation syndrome is a rare disorder due to a mutation in the PHOX2B gene, characterized by a failure in autonomic control of breathing with diminished or absent response to hypoxia and hypercapnia, which is most pronounced during sleep. Most patients present from birth with central apneas and hypoventilation, or later in the setting of a physiologic stress. Recent literature in mice with a Phox2b27Ala/+ mutation suggests a predisposition to obstructive apneas likely due to hypoglossal dysgenesis. We report on three patients with obstructive sleep apneas with absent or mild hypoventilation. Our cases propose that obstructive apneas can be the primary presentation in patients who subsequently develop the classic phenotype of congenital central hypoventilation syndrome and emphasize their close monitoring and surveillance.

Citation:

Kagan O, Zhang C, McElyea C, Keens TG, Davidson Ward SL, Perez IA. Obstructive sleep apnea as a presentation of congenital central hypoventilation syndrome. J Clin Sleep Med. 2023;19(9):1697–1700.

INTRODUCTION

Congenital central hypoventilation syndrome (CCHS) is a rare condition due to mutations in the paired-like homeobox 2B (PHOX2B) gene critical for the development of neural crest cells. CCHS is classically defined by alveolar hypoventilation secondary to a failure in the autonomic central control of breathing. Due to dysfunction in respiratory chemosensitivity, these patients show a reduced or absent ventilatory response to hypoxia or hypercapnia with absent signs of respiratory distress, most notable during sleep.^1–3 Most patients present in the newborn period with apnea or cyanosis from birth and need assisted ventilation with failure to extubate and wean off from their respiratory support. Some patients present later with apnea, cyanosis, or hypoventilation associated with exposure to anesthesia or respiratory infections.^1–3 Early identification and initiation of optimal ventilatory support are critical in optimizing outcomes in patients with CCHS. While profound hypoventilation, cyanosis, and central apneas are the hallmark of CCHS, we have observed some patients with obstructive sleep apnea (OSA) as their principal presentation. Madani et al⁴ demonstrated a predisposition to obstructive apneas in mice with the Phox2b27Ala/+ mutation likely due to hypoglossal dysgenesis. We report on three patients with CCHS with varying genotypes, presenting primarily with OSA, who subsequently developed the classic phenotype of CCHS.

This study was approved by the Institutional Review Board at Children’s Hospital Los Angeles.

REPORT OF CASES

Patient 1, who was 9 years of age at the time of writing, was born at term with pregnancy complicated by maternal substance abuse. He was asymptomatic at birth, but around 5 weeks of age was admitted to an outside pediatric intensive care unit for an apparent life-threatening event and found to have irregular breathing with feeding. A pneumogram at the time showed periodic breathing and mild desaturations. Workup with upper gastrointestinal series with small bowel follow-through, pH probe, and electroencephalography were all normal. He was discharged on home apnea monitoring but was readmitted 5 days later due to severe bradypnea as low as 4 breaths per minute and bradycardia as low as 40 beats per minute requiring stimulation. Evaluation included a normal chest X-ray, complete blood count, and basic metabolic panel including an HCO₃ of 24 mm Hg. An echocardiogram (ECHO) was significant for a patent foramen ovale with left to right shunting, and electrocardiogram showed normal sinus rhythm. A respiratory viral panel and urine toxicology were negative. The patient was thus transferred to our center at 7 weeks of age for further evaluation.

During the hospitalization, the patient had a daytime polysomnogram (PSG) showing episodes of periodic breathing with severe OSA and hypoxemia, but no central apneas or hypoventilation (Table 1). Direct laryngoscopy and bronchoscopy were performed, notable for mild laryngomalacia and prolapse of the tongue base with a significant number of desaturations reported during the procedure. Cranial ultrasound and brain magnetic resonance imaging were normal, whereas ECHO was unchanged from prior. He was treated for gastroesophageal reflux manifested by recurrent emesis and was discharged on supplemental oxygen via nasal cannula with continued home apnea monitoring.

Table 1.

Polysomnography findings.

Patient	Age at Time of Study (years)	oAHI (events/h)	CAI (events/h)	Baseline PetCO2 (mmHg)	Maximum PetCO2 (mmHg)	%TST with CO2 spent >50 mmHg	Baseline SpO2 (%)	SpO2 Nadir (%)	% TST with SpO2 < 90%
1	7 weeks	29.5	0.0	36–42	42	0	90–93	84	n/a
	1.75	19.6	5.1	36–50	58	0.8	94–97	78	n/a
	2 s/p AT	1.8	0.6	36–46	56	<25	94–96	92	0
	3 w/ O2	0.4	0.7	48–50	53	1.4	98–99	85	3
	5*	15.8	16.4	28	34	0	95	77	2
2	20*	68.5	4.5	44–46	57	10	92–94	70	20

AASM scoring was used for PSGs performed at Children’s Hospital Los Angeles. *Split-study diagnostic portion. AT = adenotonsillectomy, CAI = central apnea index, oAHI = obstructive apnea-hypopnea index, PetCO₂ = end-tidal carbon dioxide, s/p = status post, SpO₂ = oxygen saturation, TST = total sleep time.

The patient was lost to follow-up and re-presented at age 22 months with apneas during sleep lasting up to 30 seconds that responded to tactile stimulation. At the time, he had a concurrent upper respiratory infection. The patient was noted to have been off supplemental oxygen and apnea monitoring since 5 months of age. His chest X-ray during this hospitalization showed a moderately large cardio-mediastinal silhouette with a prominent pulmonary artery and hazy interstitial opacities (Figure 1). Capillary blood gas was significant for a pH of 7.48, PCO₂ of 30 mmHg, and HCO₃ of 23 mEq/L. An overnight PSG showed severe OSA, mild central sleep apnea, and hypoxemia, without evidence of hypoventilation (Table 1). Due to a history of prolonged recurrent apneas during sleep, CCHS was considered, and PHOX2B gene sequence analysis was performed, which later resulted with a PHOX2B non-polyalanine repeat mutation (NPARM) P.E183*, confirming the diagnosis of CCHS at 2 years of age.

Figure 1. Chest X-ray.

Moderately large cardio-mediastinal silhouette with a prominent pulmonary artery and hazy interstitial opacities.

During this hospitalization, the patient underwent adenotonsillectomy at 22 months due to a large adenoid pad obstructing 90% of the nasopharynx and large endophytic tonsils 3 to 4+ in size. He notably did not have complications from general anesthesia and remained on supplemental oxygen during sleep. Repeat PSG following adenotonsillectomy showed resolution of central sleep apnea, improvement in OSA, and peak end-tidal carbon dioxide (PetCO₂) of 56 mmHg but did not meet the criteria for hypoventilation (Table 1). With the diagnosis of CCHS, routine surveillance for autonomic dysfunction showed no Hirschsprung disease, neural crest–derived tumors, or sinus pauses on ambulatory cardiac monitoring.

PSG at 3 years of age showed further resolution of OSA and central sleep apnea, without hypoventilation (Table 1). However, due to the persistence of sleep-related hypoxemia he required 1 L/minute of supplemental oxygen during sleep. At age 5, repeat PSG showed the re-emergence of central sleep apnea, OSA, and sleep-related hypoxemia, still with no hypoventilation (Table 1). Of note, at this time, he was also found to have an increased body mass index in the 97th percentile. The patient was subsequently placed on noninvasive positive pressure ventilation (NIPPV) with supplemental oxygen. Details of the PSG findings are summarized in Table 1. The patient was previously reported in Kasi et al.⁵

Patient 2, who was 21 years of age at the time of writing, was born at term via emergency cesarean section for fetal distress, requiring subsequent intubation and assisted ventilation. Reports describe a history of bradycardia and a heart murmur from ages 2–3 years old. At 9 years of age, he underwent adenotonsillectomy for large tonsils and moderate-sized adenoids, with a body mass index at the time found to be greater than the 99th percentile. Ophthalmology evaluation at 10 years of age showed abnormally shaped and eccentrically located pupils with poor reactivity to light, concerning for dyscoria. At age 13, he underwent a laparoscopic cholecystectomy due to symptomatic cholelithiasis and was extubated without any complications. At that time, he was noted to have elevated levels of HCO₃, suggestive of hypoventilation (28–34 mEq/L). At age 17, a PSG was performed at an outside hospital after a seizure-like episode. Central sleep apnea was noted, with periodic breathing and oxygen desaturations to a nadir of 63%, predominantly in transitional sleep and worse during supine rapid eye movement (REM) sleep. Half of the central apneic events occurred post-arousal. Obstructive apneas were also noted, with desaturations to 79% with frequent arousals and sleep fragmentation. PetCO₂ ranged from 39 to 43 mmHg while awake and between 38 and 55 mmHg while asleep, with periods of bradycardia noted in the setting of respiratory events. The study was reported to not demonstrate clear sleep-related hypoventilation and he was subsequently treated with continuous positive airway pressure therapy.

At 18 years old, the patient was found to have an entire deletion of the PHOX2B gene as part of family genetic testing. The patient’s history is notable for fecal incontinence, with normal sphincter tone and length on anorectal manometry, making Hirschsprung disease unlikely. He is currently undergoing workup for achalasia in the setting of inability to tolerate oral intake.

At 20 years of age, the patient underwent a split-night PSG, with findings on the diagnostic portion showing severe OSA, mild baseline hypoxemia, and hypoventilation (Table 1). He was subsequently transitioned to NIPPV via a nasal mask. As part of surveillance for CCHS, brain magnetic resonance imaging and electroencephalography were normal, whereas electrocardiogram showed normal sinus rhythm with nonspecific ST segment abnormalities. Ambulatory cardiac monitoring showed no sinus pauses but recorded rare atrial and ventricular ectopic beats without patient-triggered or diary events. An ECHO showed normal biventricular systolic function without evidence of pulmonary hypertension.⁶

Patient 3, who was 10 years of age at the time of writing, was born preterm at 35 weeks via cesarean section due to intrauterine growth restriction and fetal distress. The patient had Apgar scores of 8 and 9 at 1 and 5 minutes, respectively, but soon after developed respiratory distress, initially treated with supplemental oxygen. An initial chest X-ray showed mild respiratory distress syndrome. Around this time, she had a PSG, which showed predominantly obstructive sleep apnea, as well as central sleep apnea and hypoxemia, with oxygen desaturations to ≤ 85%. There was no measurement of end-tidal carbon dioxide obtained during this study. ECHO at the time showed a patent foramen ovale, and a cranial ultrasound was grossly normal. She was treated for gastroesophageal reflux at 4 weeks of age due to emesis after feeding.

Due to worsening respiratory acidosis (PCO₂ = 89 mmHg) with apneas, bradycardia, and desaturations, she was intubated at 1 month of age, placed on assisted ventilation, and transferred to our center. Respiratory acidosis was initially attributed to prematurity; however, because her maternal aunt had a diagnosis of CCHS, PHOX2B gene mutation analysis was obtained, which showed an NPARM p. P82L (c.245C>T), similar to her aunt. She underwent a tracheostomy and was discharged on home mechanical ventilation.

Further evaluation included an electrocardiogram, which showed normal sinus rhythm, and brain magnetic resonance imaging, which was normal. Renal ultrasound showed a duplicated collecting system on the right with mild pelviectasis bilaterally. She has intermittent exotropia but does not have tumors of neural crest origin. Her surveillance ECHOs to date do not show pulmonary hypertension, and her annual cardiac ambulatory monitoring have shown a sinus rhythm with rare atrial and ventricular ectopic beats not associated with patient-triggered events.

At 7 years of age, she was transitioned to NIPPV using a nasal mask and chin strap and underwent laryngoscopy and bronchoscopy with decannulation and closure of her tracheostomy, which she tolerated well. This patient was previously reported in Kasi et al.⁷

DISCUSSION

Our cases demonstrate that OSA can be the initial or predominant respiratory presentation in three patients with CCHS. The OSAs may be severe, with mild to absent central sleep apnea or hypoventilation as seen in cases 1 and 3. In addition, the obstructive and central sleep apneas may improve temporarily following adenotonsillectomy as seen in case 1.

Madani et al⁴ described the preponderance of obstructive and mixed apneas in addition to central apnea in Phox2b27Ala/+ newborn mice. Furthermore, the obstructive and mixed apneas were found to be longer in duration than the central apneas. The authors attributed the obstructive events to dysgenesis and dysfunction of the hypoglossal nucleus, a non–Phox2b-expressing structure that is critical to the maintenance of upper airway patency, in combination with disruption of input from the Phox2b-expressing retrotrapezoid nucleus.

The recognition of OSA as a presenting or concurrent feature in CCHS is of major clinical significance. Due to obesity or prematurity, CCHS may be disregarded as the etiology of a patient’s sleep-related breathing disorder. As seen in cases 1 and 2, our patients’ clinical courses were also notably complicated by obesity, with body mass indexes greater than the 95th percentile, possibly contributing to the re-emergence of their OSA on later sleep studies and further complicating the interpretation of their hypoventilation. Our third case displayed a predominance in OSA and CCHS would not have been considered without knowledge of her aunt’s diagnosis.⁷

Patients with CCHS typically present with central apneas and hypoventilation, particularly during sleep. Most are ventilated via positive-pressure ventilation via tracheostomy with mandatory rates. Diaphragm pacing without tracheostomy and NIPPV are treatment options in older children with CCHS. Diaphragm pacing as a mode of ventilatory support allows for decannulation, such that patients are ventilated without tracheostomy. However, diaphragm pacing without tracheostomy may result in OSA, particularly at a young age, as the synchronous contraction of the upper airway skeletal muscles and contraction of the diaphragm does not occur, predisposing to collapse of the upper airway.^8–10 In the presence of OSA, patients treated with NIPPV may require higher pressure settings than they would otherwise need. Thus, knowledge of a patient’s predisposition to OSA is critical in patient selection for these modes of ventilatory support. The presence or absence of OSA needs to be incorporated into decision making when these options are contemplated. We propose that CCHS be considered in the differential diagnosis of patients with OSA who do not have risk factors, and to test for this disorder when the OSA is unexplained.

Case 1 was reported in a case series describing the variable phenotype of PHOX2B NPARM⁵ and cases 2 and 3 were reported describing the different phenotypes in families carrying the same PHOX2B genotype.^6,7 We chose to reintroduce these cases to emphasize the presence of OSA in patients with CCHS.

There are some limitations that may result in the underestimation of the presence of hypoventilation. First, one patient had a daytime polysomnography (nap study); thus, some period of hypoventilation may have been missed. Second, another patient had no end-tidal CO₂ measurement during their first sleep study, highlighting the importance of CO₂ monitoring in these patients. Some infants and children may have respiratory rates that preclude optimal PetCO₂ data collection during their polysomnography. In these situations, transcutaneous CO₂ monitoring may provide additional information.

In summary, our report shows that OSA can be an initial or predominant finding in patients who subsequently develop the phenotype of CCHS, even in the absence of initial central apneas and hypoventilation. In the setting of other clinical manifestations of autonomic nervous system dysfunction, a high index of suspicion for the diagnosis of CCHS is indicated in those with an initial presentation of OSA, recognizing that central apneas and hypoventilation may develop over time.

DISCLOSURE STATEMENT

All authors have seen and approved the manuscript. Work for this study was performed at Children’s Hospital Los Angeles (CHLA), Los Angeles, California. The authors report no conflicts of interest.

ABBREVIATIONS

CCHS

congenital central hypoventilation syndrome

ECHO

echocardiogram

NIPPV

noninvasive positive pressure ventilation

OSA

obstructive sleep apnea

PetCO₂

end-tidal carbon dioxide

PHOX2B

paired-like homeobox 2B

PSG

polysomnogram