Introduction

Congenital central hypoventilation syndrome (CCHS) is a rare autonomic nervous system disorder. The characteristic abnormal ventilatory response to progressive hypercapnia and sustained hypoxia is usually identified in neonates. The typical respiratory pattern, with reduced tidal volumes and persistent hypoventilation, worsens during sleep [1–5]. Other manifestations of dysautonomia may include arrhythmias, ophthalmologic signs, Hirschsprung’s disease and neuroblastoma [6].

This syndrome was first described by Mellins in 1970 [7]. However, it was not until 2003 that its etiology was found to be genetic. More specifically, the syndrome was found to be caused by heterozygous mutations within the paired-like homeobox 2B gene (PHOX2B) [8, 9]. This gene is expressed uniquely in the nervous system, in neurons that control the cardiovascular, gastrointestinal and respiratory systems, and codifies for the protein PHOX2B which is a homeodomain transcription factor.

The identification of a PHOX2B gene mutation is currently required to confirm the diagnosis. More than 90 % of the mutations presently known are heterozygous duplications within the second polyAla stretch (polyalanine repeat mutations (PARMS)) located in exon 3 of the PHOX2B gene in the 4p12 region, which lead to in-frame alanine expansions of 5 to 13 alanines [2, 8, 10–12]. Missense, nonsense and frameshift PHOX2B mutations (non-polyalanine repeat mutations (NPARM)), as well as large deletions in the PHOX2B locus explain most of the remaining cases [13]. These mutations usually occur de novo, but some can be inherited from a parent harbouring a constitutional or somatic mutation [12–16]. Parental genetic studies are thus mandatory for appropriate genetic counselling [1, 2]. Correlations between the genotype and the phenotype have been drawn, among PARM and NPARM mutations [2, 9, 10, 12, 17].

Most of the known CCHS cases have been diagnosed during the neonatal period. However, due to increasing awareness and more available and accurate genetic diagnosis, more patients are now diagnosed in late infancy or adult age, with NPARM mutations. Nonetheless, CCHS is still well underdiagnosed [1, 11, 18–21].

In order to contribute to a better knowledge of this syndrome, the authors herein present the clinical case of a child with CCHS. The patient was diagnosed in her second year of life, following several episodes of hypoventilation during respiratory infections.

Clinical case

A 4-year-old girl was a firstborn child of a young and non-consanguineous couple, with no relevant medical family history. The pregnancy was uneventful, with eutocic delivery at 39 weeks of gestation. Somatometry at birth was normal for gestational age, and there were no neonatal events. Somatic growth and development were normal in her first year of life.

At 9 months of age, she was admitted to the Neonatal and Pediatric Intensive Care Service (SCINP), at the Centro Hospitalar do Porto, with respiratory syncitial virus bronchiolitis. A respiratory failure with hypercapnia (pH 7.22, PaCO2 109 mmHg, PaO2 71 mmHg) led to a 5-day period of mechanical ventilation, and 14 days later, she was discharged home. Whilst on mechanical ventilation, the patient’s PaCO2 was above 60 mmHg on several occasions. PaCO2 was 43.2 mmHg after weaning from ventilation and 36 mmHg on discharge.

At 11 months, she was admitted again to the SCINP with a metapneumovirus pneumonia, with respiratory acidosis (pH 7.23, PaCO2 108 mmHg, PaO2 128 mmHg with 0.5 L/min supplemental O2) and a right upper lobe atelectasis. Cellular and humoral immunologic studies showed normal results. The patient was under mechanical ventilation for a 15-day period and was discharged after 19 days, with a PaCO2 of 44 mmHg. No alterations were found on a flexible bronchoscopy performed at 12 months of age.

A third admission to the SCINP occurred at 13 months of age, again with a respiratory infection and progressive hypercapnia (pH 7.19, PaCO2 94 mmHg, PaO2 95 mmHg). A superficial respiratory pattern with apneas during sleep and fluctuant blood gas values was observed, with lower CO2 values when awake. The hypothesis of CCHS was considered and bilevel positive airway pressure (BiPAP) non-invasive ventilation was initiated. This led to both clinical and blood gas improvement (Table 1).

Table 1 Blood gases values before and after implementation of non-invasive ventilation
Full size table

Additional investigations were undertaken, including ophthalmological and neurological examinations, brain MRI, urinary cathecolamines, creatinekinase and aldolase levels, abdominal ultrasound, electrocardiogram and echocardiography. All results were normal.

In order to sequence the PHOX2B gene, a blood sample was taken with informed consent and DNA was extracted according to standard protocols. We screened the coding sequence of the PHOX2B gene by direct DNA sequencing performed by the fluorometric method (BigDye Terminator Cycle Sequencing kit, Applied Biosystems, Warrington, UK) as previously described [6]. Mutation nomenclature is given using the HGVS notation, and the reference sequence used is NM_003924.2 in GenBank. Nucleotidic numbering is based on 1 as the A of the ATG translation codon. We identified a heterozygous insertion of an adenine at position 23 (c.23dupA), leading to a premature stop codon (p.Y8X) (Fig. 1). The mutation was considered to have occurred de novo as it was not detected in DNA extracted from blood of both parents.

Fig. 1
figure 1

Sequence chromatogram of PHOX2B exon1 showing the heterozygous c.23dupA, leading to a premature stop codon (p.Y8X)

Full size image

The patient was discharged after 17 days on mask non-invasive ventilation with BiPAP and backup respiratory rate while asleep.

At 15 months of age, we performed an overnight polysomnography at the Sleep and Ventilation Laboratory in Hospital Pediátrico de Coimbra—Centro Hospitalar e Universitário de Coimbra. Data were recorded by an Alice® 5 Diagnostics System (Healthdyne Technologies, Respironics). Sleep stages were documented by an electroencephalogram, respiratory movements by impedance plethysmography, airflow by thermistor, nasal pressure by nasal cannula, arterial oxygen saturation by pulse oximetry, snoring by tracheal microphone, and mental and leg movements by electromyogram. End-tidal CO2 (ETCO2) was monitored by a CO2SMO® equipment (Respironics). The first part of the sleep study was performed on spontaneous ventilation. During the second part of the study, ventilation parameters were adjusted in order to obtain the best gas values and sleep parameters. Manual scoring and classification was performed according to the “Children Rules of the American Academy Sleep Medicine Manual for Sleep Scoring and Associated Events,” 2007 [22], except for the CO2 criteria for hypoventilation. A level of CO2 ≥ 50 mmHg was adopted attending to software equipment.

During the sleep study, there were no apneas. However, a hypoventilation pattern was registered: a small amplitude wave pattern with ETCO2 values ≥50 mmHg registered during 99 % of total sleep time with a maximum of 65 mmHg and an oxygen saturation lowered to 74 % (Fig. 2). There were frequent gasps followed by very slight recovery of peripheral O2 saturation.

Fig. 2
figure 2

Sleep study. Epoch of 30 s displaying hypoventilation with desaturation (solid arrow) and hypercapnia (arrow with broken lines) during non-REM sleep (N2)

Full size image

Screening test SG Skills II performed at 15 months of age showed a psychomotor development in accordance with the patient’s chronological age. However, there was slight hypotonia and slowing in execution performance. The patient was included in a rehabilitation program, and later evaluations showed recovery, with no cognitive, motor or coordination deficits. In the meantime, she repeated the cardiology evaluation, at 17 and 24 months, with normal echocardiography and 24-h Holter.

At 3.5 years, another polysomnography was done for titration of ventilation parameters, in which CO2 was measured transcutaneouly (TcCO2) by a TCM Tosca®, Radiometer. During this study, hypoxemia continued to be observed when falling asleep and during non-REM sleep (Fig. 3), although apparently less hypoxemia and lesser hypercapnia (Table 2). There were no alterations of respiratory or heart rates in response to SpO2 decline or TcCO2 augmentation.

Fig. 3
figure 3

Sleep study. A period of 5 min during which transition from wakefulness (D) to deep sleep (non-REM) is accompanied by rapid desaturation without increasing respiratory rate

Full size image
Table 2 Polysomnogram data at 15 months (P1) and at 3.5 years old (P2)
Full size table

The patient continues to be supported with non-invasive ventilation during sleep, with no further admissions. Her daytime ventilation continues to be normal. She has no facial deformation, and her physical growth is regular on the 75–90 percentile.

Discussion

Patients with CCHS present with abnormal respiratory control, with diminished sensitivity to hypercapnia and hypoxia [4, 5, 10, 11, 21]. Late-onset central hypoventilation is often triggered by a respiratory infection, in which respiratory demands increase. Significant differences in blood gas values between sleep and wakefulness states are usually observed. Butin et al. report a situation of hypercapnic coma in a 9-year-old girl during a respiratory infection caused by Mycoplasma pneumoniae (Table 3). After 4 days on mechanical ventilation, attempts to wean the patient from non-invasive ventilation were unsuccessful [15]. Cohen-Cymberknoh et al. also describe the case of a 12-year-old girl, with a diagnosis of CCHS following the difficulty in weaning from mechanical ventilation during pneumonia [23]. In our case, as in the one described by Parodi et al., repeated need for mechanical ventilation during common infections led to investigation of an underlying disease. Immune deficiency and anatomical respiratory tract malformation were excluded before late-onset central hypoventilation was diagnosed [24].

Table 3 Literature data concerning late-onset CCHS patients
Full size table

Another possible trigger for CCHS is the use of drugs with depressive effect on arousal. Mahfouz et al. present a 6-year-old girl who had a difficult recovery from general anaesthesia due to hypoventilation [21]. Chronic manifestations like polycythemia, cardiomyopathy, morning headaches or daytime sleepiness may be the diagnostic clue in older patients [19, 20].

Regarding the most common PARM mutations in exon 3 of the PHOX2B gene, a correlation between the genotype and the phenotype was shown. Late-onset presentations are related to shorter polyalanine expansions. Trochet et al. report 17 cases of late-onset CCHS, 15 of which with a short polyalanine expansion, 1 with a missense mutation in exon 2 and one with a deletion in exon 3 [17]. Other NPARM mutations have also been reported [12, 24, 25].

The mutation in exon 1 found in our case had not yet been described. However, Trochet et al. have reported a mutation in exon 1 also resulting in the premature stop codon c.42C>A, p.Y14X, in a boy diagnosed with CCHS at 5 weeks of age [14]. Interestingly, the mutation has been shown to escape nonsense-mediated decay and to lead to a shorter protein through re-initiation of transcription at methionine 18 or 21. A similar consequence is highly likely for the Y8X mutation. Thus, these mutations should not be considered as plain loss-of-function mutations. Finally, as for short alanine expansions, clinical consequences can be variable, with central hypoventilation either congenital or of late onset as in the case of our patient. Such variability could be due to the effect of unknown modifier gene(s).

Polysomnography is important in CCHS patient characterization and evaluation [5]. Our patient has regular assessments of respiratory function both awake and asleep, and ventilator support is adjusted accordingly. We hope this follow-up and home SpO2 monitoring can prevent serious hypoxia and hypercapnia, which may impact on survival, cardiovascular function and central nervous system development [16, 26].,Our patient has no evidence of other systems’ impairment so far. We would venture to say that she seems to have improved a little bit considering the better results of SpO2 and CO2 on the polysomnography at 3.5 years (Table 2). Regular evaluations are essential for identification of state modifications and intervention.

This case illustrates the difficulties in late-onset CCHS diagnosis, where clinical manifestations are often non-specific and highly variable. Early diagnosis has great importance in order to prevent and/or anticipate situations with risk of decompensation. Moreover, such rare cases increase our knowledge about CCHS genotype-phenotype correlations. CCHS is known until now as a lifelong disease, but there may be some improvement with growth, although still requiring attention in stressful situations.