We report an unusual case of an adult patient carrying a germline PHOX2B frameshift mutation and hence was diagnosed with congenital central hypoventilation syndrome. He came to medical attention after the mutation was identified in his daughter who presented with hypoventilation and a neuroblastoma. Although PHOX2B mutations are usually associated with a phenotype of congenital hypoventilation, severe autonomic dysfunction and neural crest tumors, our patient had no complaints at the time of presentation. At polysomnography we found severe positional hypercapnic central sleep apnea, partly responsive to positional therapy. Eventually, he was titrated to noninvasive ventilation with resolution of the central breathing events and, in hindsight, a more refreshing sleep than before. Clinicians working in sleep medicine need to be aware of the variable expression of this rare condition to prevent late cardiorespiratory and neurocognitive complications.
Janssen HC, Vulto-van Silfhout AT, Jongmans MC, van der Hout AH, Overeem S. Severe positional central sleep apnea in an asymptomatic adult with a PHOX2B frameshift mutation. J Clin Sleep Med. 2018;14(8):1427–1430.
Congenital central hypoventilation syndrome (CCHS) is a rare condition in which hypoventilation, predominantly during non-rapid eye movement (NREM) sleep, is the key feature.1 Patients with CCHS lack responsiveness to hypercapnia and hypoxemia and sense no dyspnoea despite increase in pCO2 and decrease in saturation. It is associated with abnormalities of neural crest origin: severe constipation or Hirschsprung disease and neural crest tumors, in addition to other symptoms associated with autonomic nervous system disease such as cardiac arrhythmias, tonic pupils, profuse sweating and problems with controlling body temperature.2 A germline mutation in the paired-like homeobox PHOX2B gene, that codes for a transcription factor important for neural crest differentiation, confirms the diagnosis of CCHS. Patients usually present in the neonatal period, although case reports on late-onset CCHS have been presented in which an older child or even an adult came to medical attention after a respiratory tract infection or general anaesthesia.3 In some patients, symptoms such as morning headache or decreased exercise tolerance were present before the acute medical situation, but were not recognized as relevant.4
Here we present an adult patient with a PHOX2B frameshift mutation who was diagnosed with severe positional central sleep apnea, despite a lack of sleep-related complaints.
REPORT OF CASE
A 36-year-old male patient was referred to our outpatient sleep medicine clinic after his daughter was diagnosed with hypoventilation and a neuroblastoma. A PHOX2B frame-shift mutation was found in the daughter (c.691_698dup, p.Gly234Alafs*78, NM_003924.3) and upon family screening, our patient was diagnosed with the same PHOX2B mutation. The reason for referral was to screen for nocturnal hypoventilation.
He reported no excessive daytime sleepiness, only mild fatigue, which was attributed to stress because of the recent diagnosis of his daughter. There were no insomnia complaints. He had a regular sleep-wake schedule, with a bedtime of 10:30 pm and getting up at 7:00 am. He reported his sleep quality as reasonably good. His wife noticed snoring approximately 3 times/wk and incidental apneas 1–2 times/wk. There was no nycturia, no morning headaches, nor dry mouth upon awakening. There were no clinical signs of autonomic nervous system dysfunction.
He worked fulltime as a consultant. As a child he had suffered from constipation and it was noticed that his pupils had a delayed response to light. Cardiological screening revealed a bicuspid aortic valve. He never smoked and drank on average one glass of wine and six cups of coffee a day. He used no medication. His body mass index was 25.5 kg/m2. Scores on the Insomnia Severity Index (1) and Epworth Sleeping Scale (5) were normal.
Video polysomnography (Figure 1) was done. Because of a technical defect, the end-tidal CO2 measurement failed during the initial recording. Total sleep time was 7 hours 33 minutes, with a sleep efficiency of 90.6%. A fragmented sleep pattern was observed with frequent stage shifts. Severe central sleep apnea (CSA) was present with an apnea-hypopnoea index (AHI) of 51.7 events/h and an apnea index (AI) of 44.7 events/h. There was a marked position dependency, with a supine AHI of 78 events/h during 271 minutes supine sleep and a nonsupine AHI of 12.5 events/h during 182.5 minutes of non-supine sleep. Saturation was 22.4 minutes below 90%. There were no breathing events observed in the supine wake position, although it should be noted that these periods were short. The estimated respiratory rate was between 8–10 per minute in wake, stage N1, N2 and N3 sleep and 10–14 per minute in rapid eye movement (REM) sleep. On occasion snoring was heard, predominantly in supine position. Capillary blood gas during the day revealed: pH 7.45, pCO2 42 mmHg, BE 4.1, and bicarbonate 28.2 mmol/L. Capillary blood gas upon awaking: pH 7.39, pCO2 50 mmHg, BE 3.8, bicarbonate 29.6 mmol/L.
Polysomnography without therapy.
Severe positional central sleep apnea was detected. (top) Trend analysis showing the hypnogram, sleeping position (blue stripes are supine), saturation, and breathing events. (bottom) Detail showing the trend, hypnogram (30 second timeline) and central breathing events on a 3-minute timeline. Note that the breathing pattern does not comprise Cheyne-Stokes respiration.
Polysomnography without therapy.
Rather unexpectedly in this asymptomatic patient, we made a diagnosis of severe positional central sleep apnea with mild hypercapnia upon awakening. After deliberation with the patient, given his mild complaints, we decided to start with positional therapy and to evaluate the effect with a follow-up polysomnography. He was also referred to a pulmonologist for further analysis. Lung function tests were normal. Two months later, a new polysomnography with capnography was performed while the patient used positional therapy with the tennis ball technique.
The video polysomnography with positional therapy (Figure 2) recored a total sleep time of 8 hours 5 minutes with a 95.9% sleep efficiency. The AHI was 19.2 events/h and an AI 12.9 events/h, predominantly central in origin, but also a few mixed and obstructive events. There was snoring. Despite positional therapy, 1 hour 11 minutes of sleep was spent in the supine position, with a supine AHI of 68 events/h; nonsupine AHI was 10.6 events/h. Transcutaneous CO2 during wake was 39.6 mmHg. The highest transcutaneous CO2 value during sleep was 48.8 mmHg. Capillary blood gas in the morning showed a pH of 7.36 and a pCO2 of 53 mmHg.
Polysomnography with positional therapy using the tennis ball technique.
Trend analysis. Note the decreased time in supine sleeping position (blue stripes), with a decreased amount of central breathing events.
Polysomnography with positional therapy using the tennis ball technique.
Given the reduction in the number of central apneas with positional therapy, the patient was motivated to intensify positional therapy with a different (larger) device and a new polysomnography was performed after 4 months. As before, he did not have any complaints and improvement of the nocturnal breathing disorder did not change his daytime functioning.
Video polysomnography with the adapted positional therapy recored a total sleep time of 7 hours 48 minutes and sleep efficiency of 89.5%. AHI was 28 events/h and AI 16.3 events/h, predominantly central in origin. There was snoring. Despite modified therapy, 1 hours 34 minutes was spent in the supine position. Supine AHI was 46 events/h. Saturation was 0.7 minutes below 90%. The transcutaneous CO2 while awake was 52 mmHg, the highest value was 60.5 mmHg during NREM sleep. The capillary blood gas upon awaking showed a pCO2 56 mmHg. In the capillary blood gas during the day the pCO2 was 51 mmHg.
Despite a different positional therapy device, he still slept more than one hour per night in the supine position. Furthermore, pCO2 values did not improve and over the weeks hypercapnia persisted during the day.
We discussed our findings thoroughly with the patient. Although he still did not report any complaints, we referred him to a home ventilation center because of the possible long term cardiopulmonary consequences of this untreated hypercapnic central sleep apnea. He was titrated to nocturnal noninvasive ventilation (NIV) with a Vivo 50 device (Breas Medical AB, Mölnlycke, Sweden): pressure A/C, inspiratory positive airway pressure 14 cmH2O, expiratory positive airway pressure 2 cmH2O, inspiration time 1.0 second, trigger 4, breathing frequency 16/min. While on NIV, polysomnography showed a normalized AHI < 1 event/h. Despite the fact that he did not have any complaints before, he reported feeling more refreshed in the morning since using NIV. Some nights, he did not use the NIV and then noticed more fatigue during the day. Upon using NIV, the nocturnal transcutaneous CO2 decreased to 42 mmHg with a mean saturation of 98.4%. His capillary blood gas during the day revealed a pCO2 of 42.6 mmHg with a bicarbonate of 25.5 mmol/L.
We report on an adult patient with a PHOX2B mutation who only came to medical attention after his daughter was diagnosed with this mutation after development of a neuroblastoma. Polysomnography revealed severe positional central sleep apnea, with mild hypercapnia and hypoxia. This case is unusual in several regards.
More than 90% of patients with CCHS are heterozygous for a polyalanine repeat expansion mutation (PARM), which leads to alanine expansions of between 24–33 on the affected allele. In general, there is an association between the repeat length and the severity of the symptoms. Most mutations occur de novo, but in about 20% of cases the mutation is inherited from an asymptomatic parent who carries a mosaic mutation or from a parent with the 20/25 polyalanine repeat genotype, in which clinical symptoms may vary and are usually milder.1
Although our patient had mild symptoms, he did not have a PARM. The mutation found in our patient is denoted as a non-polyalanine repeat expansion mutation (NPARM), which constitutes about 8% to 10% of the mutations in the PHOX2B gene and are frameshift, nonsense and missense mutations. They are usually associated with a more severe clinical presentation and a higher risk of neuroblastoma, ventilator dependency and Hirschsprung disease, although variable penetrance has been described in some NPARM mutations.5 The specific mutation that was identified in our patient and his daughter (c.691_698dup) has been reported prior in 3 families.6–8 The fact that most PHOX2B NPARM mutations occur de novo while this specific mutation is inherited in several families, underscores the milder phenotype associated with a mutation in this region.
Severe (in terms of the AHI) positional CSA was found in our patient. This is a rather unexpected finding in CCHS. It is known that the sleep-related hypoventilation in CCHS is most prominent in NREM sleep and in the first polysomnography we performed in our patient, central events were virtually absent in REM sleep. However, it appeared more as a consequence of the lateral position in REM sleep than REM sleep per se, since there were also few events in stage N3 sleep in the lateral position. Reports on positional central sleep apnea are scarce and the reason for the position dependency of CSA is not fully understood.9 Supine dependency may be due to a coexisting obstructive component particularly at the end of a central apnea, due to the supine position affecting cardiac function or pulmonary congestion.9 It might be that an obstructive component did influence the positional dependency in our patient, given the mild snoring detected on the PSG.
To our knowledge, this is the first report of a bicuspid aortic valve in a patient with a PHOX2B mutation, which has a prevalence of approximately 1% in the general population. Congenital heart disease in patients with a PHOX2B mutation has only been described in a few patients. Recently, Lombardo et al.10 reported on one patient with a right-sided aortic arch with complete vascular ring and one patient with an aberrant origin of the left coronary artery. One can hypothesize on the dysfunction of the PHOX2B protein and the effect on other transcription factors in the aetiology of the bicuspid aortic valve in this patient. Development of the aortic and pulmonary valves requires contribution from migratory cardiac neural crest cells. Precursor cells of the neural crest differentiate, proliferate and migrate to become differentiated cells in the nervous system, but also in cardiac development.1
Our case broadens the knowledge on the genotypephenotype relation in late onset CCHS. It is important that sleep medicine clinicians working with adult patients are aware of the variable presentation of this rare condition. Early recognition is important to prevent late cardiorespiratory and neurocognitive complications.
All authors have seen and approved the manuscript. The authors report no conflicts of interest.