Congenital Central Hypoventilation Syndrome

Clinical characteristics.

Congenital central hypoventilation syndrome (CCHS) is a rare disorder of respiratory and autonomic regulation. It is typically characterized by a classic presentation in newborns and, rarely, a milder later-onset (LO-CCHS) presentation in toddlers, children, and adults.

Classic CCHS presents in newborns as:

• Apparent hypoventilation with monotonous respiratory rates and shallow breathing either during sleep only or while awake as well as asleep;
• Autonomic nervous system dysregulation (ANSD); and
• In some individuals, altered development of neural crest-derived structures (i.e., Hirschsprung disease) and/or tumors of neural crest origin (neuroblastoma, ganglioneuroma, and ganglioneuroblastoma).

Individuals with CCHS who have been diagnosed as newborns and ventilated conservatively and consistently throughout childhood have now reached the age of 20 to 30 years; they are highly functional and live independently. LO-CCHS manifests as nocturnal alveolar hypoventilation and mild ANSD. Individuals with LO-CCHS who were not identified until age 20 years or older have now reached the age of 30 to 55 years.

Diagnosis/testing.

Diagnosis of CCHS is established based on:

• Clinical findings of alveolar hypoventilation and ANSD in the absence of primary pulmonary, cardiac, or neuromuscular disease, or a causative brain stem lesion that can account for the entire phenotype; and
• Identification of a pathogenic variant in PHOX2B. PHOX2B is the only gene in which mutation is known to cause CCHS.

Management.

Treatment of manifestations: Tracheostomy and home ventilator for individuals requiring ventilatory support 24 hours per day and for infants/children/adults requiring ventilatory support during sleep only. Diaphragm pacing by phrenic nerve stimulation can be considered in ambulatory children requiring mechanical ventilation 24 hours a day and potentially in older children and adults requiring nocturnal ventilation only, though tracheostomy removal for nocturnal diaphragm pacing is not assured. Mask ventilation or negative-pressure ventilation is a consideration in cooperative older children requiring ventilatory support during sleep; however, during intercurrent illnesses more aggressive ventilatory support such as intubation with continuous mechanical ventilation in an intensive care setting may be needed. A cardiac pacemaker may be required for prolonged sinus pauses. Hirschsprung disease is treated in the usual manner. Neuroblastomas are removed surgically; those beyond Stage 1 are treated with chemotherapy. Treatment of other tumors of neural crest origin is based on location and type, though surgical removal is typically recommended.

Prevention of secondary complications: Mask ventilation in the infant and young child is strongly discouraged because it is not adequately stable as a life-sustaining support, with risk for repeated hypoxemia and neurocognitive compromise.

Surveillance: For all individuals with CCHS: at least yearly (every 6 months until age 3 years) comprehensive, multiple-day in-hospital physiologic evaluation to optimize ventilatory support awake and asleep and in varied levels of activity and concentration simulating activities of daily living; yearly 72-hour Holter recording to identify any prolonged sinus pauses; yearly echocardiogram to identify right ventricular hypertrophy or cor pulmonale; yearly hemoglobin, hematocrit, and reticulocyte counts to identify polycythemia; and yearly neurocognitive testing to evaluate the success of artificial ventilation. For children with specific PHOX2B variants placing them at higher risk: evaluate for Hirschsprung disease and tumors of neural crest origin.

Agents/circumstances to avoid: Swimming (asphyxia; death); breath-holding contests (asphyxia; death); alcohol (respiratory depression), recreational drugs (varied effects including death), and prescribed as well as non-prescribed medications/sedatives/anesthetics that could induce respiratory depression.

Evaluation of relatives at risk: Both parents of children with a known PHOX2B pathogenic variant should be tested for the family-specific variant to determine their risk for later-onset CCHS or mosaicism.

Genetic counseling.

CCHS is inherited in an autosomal dominant manner. Most individuals with CCHS are heterozygous for a de novo PHOX2B pathogenic variant; some have an affected parent and up to 25% have an asymptomatic parent who has mosaicism for a PHOX2B variant. Each child of an individual with CCHS has a 50% chance of inheriting the PHOX2B pathogenic variant; the risk to the offspring of an individual with mosaicism is 50% or lower. Prenatal testing for pregnancies at increased risk is possible if the causative variant has been identified in an affected family member. Some families choose to pursue prenatal testing in order to make informed decisions about the pregnancy and, if the pregnancy is continued, allow for a smooth transition to extrauterine life for the affected infant.

Clinical Diagnosis

The American Thoracic Society has issued both an updated statement on the diagnosis and management of congenital central hypoventilation syndrome (CCHS) [Weese-Mayer et al 2010] (full text) and a lay summary [Patwari et al 2010b] (full text).

CCHS is diagnosed in newborns with the following:

• Hypoventilation with absent or attenuated ventilatory response to hypercarbia and/or hypoxemia when awake and asleep
• Generally adequate ventilation while awake and at rest and apparent hypoventilation with monotonous respiratory rate and shallow breathing (diminished tidal volume) during sleep OR apparent hypoventilation while both awake and asleep
• Absent perception of asphyxia (i.e., absent behavioral awareness of hypercarbia and/or hypoxemia) and absent arousal from sleep with development of physiologic compromise secondary to hypercarbia and/or hypoxemia
• No evidence of primary neuromuscular, lung, or cardiac disease or identifiable brain stem lesion that could account for the full constellation of signs and symptoms including autonomic nervous system dysregulation (ANSD)
• Presence of a CCHS-related PHOX2B variant
• Symptoms of ANSD including but not limited to severe breath-holding spells; lack of physiologic responsiveness to the challenges of exercise and environmental stressors; diminished pupillary light response; esophageal dysmotility; severe constipation even in the absence of Hirschsprung disease; profuse sweating; reduced basal body temperature; and altered perception of anxiety

Later-onset CCHS (LO-CCHS) is diagnosed in individuals with the following:

• Same criteria as described above for CCHS of the newborn but with presentation after one month of life, often occurring in later childhood or adulthood

Molecular Genetic Testing

Gene. PHOX2B is the only gene in which mutation is known to cause CCHS.

The two major types of PHOX2B variants observed in CCHS are polyalanine repeat expansion mutations (PARMs) and non-polyalanine repeat expansion mutations (NPARMs).

Polyalanine repeat expansion mutations (PARMs)

PHOX2B has two polyalanine repeat regions in exon 3, the second of which is the region of primary importance in CCHS. This polyalanine repeat comprises any one of four codon combinations – GCA, GCT, GCC, or GCG – as each one encodes the amino acid alanine. The term "GCN" has been used to designate these four codons.

Allele sizes. Allele sizes and categories are summarized here; see also Weese-Mayer et al [2010].

• Normal alleles. The unaffected individual has 20 alanines (GCN repeats) on both PHOXB alleles in the repeat region of exon 3.Though benign variants of 9, 13, 14, and 15 GCN repeats have been reported [Amiel et al 2003, Weese-Mayer et al 2003, Toyota et al 2004], individuals with alleles of this length have not been studied systematically to confirm they are entirely normal without any control of breathing deficit or autonomic dysregulation.
• Mutable normal alleles. Currently, this category of alleles is not known to occur in this disorder
• Reduced penetrance alleles. Individuals heterozygous for 24 alanine repeats (e.g., genotype 20/24) and a subset of individuals heterozygous for 25 alanine repeats (e.g., genotype 20/25) may have a very mild phenotype such that diagnosis is delayed and/or not manifest except when exposed to respiratory depressants or severe intercurrent pulmonary illness [Repetto et al 2009]. Rarely a small NPARM will also have variable penetrance [Berry-Kravis et al 2006].
• Full penetrance alleles. Individuals with 25 alanine repeats who present in the newborn period (e.g., genotype 20/25), and those heterozygous for 26 to 33 alanine repeats (e.g., genotype 20/26 to 20/33) [Weese-Mayer et al 2003, Weese-Mayer et al 2010]. The largest known repeat length is 33 alanines.
• Alleles of uncertain significance. Only one individual with such a small expansion allele has been described [Toyota et al 2004], in a study on schizophrenia; no clinical information relevant to CCHS phenotype is known about the individual.

Non-polyalanine repeat expansion mutations (NPARMs)

PHOX2B variants that are not specifically polyalanine expansions, including sequence alterations outside of the polyalanine repeat and frameshift variants affecting the region encoding the polyalanine repeat, are typically small out-of-frame deletions or duplications of approximately one to 38 nucleotides.

Note: Details of these variants from many published reports are summarized in Berry-Kravis et al [2006] and Weese-Mayer et al [2010].

Though individuals with NPARMs typically have a more severe phenotype than most individuals with PARMs, on rare occasion a small frameshift variant could have reduced but variable penetrance in a given family [Berry-Kravis et al 2006].

PHOX2B deletions ranging from 6,216 base pairs (involving only PHOX2B exon 3) to 2.6 megabases (involving all of PHOX2B and 12 other genes) have been observed in a small cohort of individuals with clinical findings that may include alveolar hypoventilation or Hirschsprung disease [Jennings et al 2012]. Further study is necessary to elucidate the relationship between PHOX2B haploinsufficiency and the CCHS phenotype [Jennings et al 2012].

Clinical testing

• Targeted analysis for pathogenic variants (fragment length analysis). This test, referred to as the PHOX2B Screening Test [Weese-Mayer et al 2010, Weese-Mayer et al 2012], amplifies the region encoding the polyalanine repeat and determines the polyalanine repeat length. Specifically, it detects the polyalanine repeat expansion mutations (PARMs) observed in 92% (185/201) of individuals with CCHS as well as the large (35- and 38-bp) deletions, and some of the small out-of-frame deletions or duplications [Berry-Kravis et al 2006]. Thus, the PHOX2B Screening Test identifies pathogenic variants in approximately 95% of individuals with CCHS. In addition, it is the only clinically available test to identify low-level somatic mosaicism [Jennings et al 2010].
  Note: Small out-of-frame deletions or duplications change the expected length of the PCR fragment and, thus, can also be detected by fragment length analysis; however, the identification of the precise nucleotide changes and confirmation of a frameshift require subsequent (sequel) sequence analysis.
• Sequence analysis. Approximately 8% (16/201) of individuals with CCHS have a PHOX2B missense, nonsense, frameshift, or stop codon variant, including frameshifts in the polyalanine region described above. As noted above, a subset of these NPARMs are detected by the PHOX2B Screening Test.
• Deletion/duplication analysis. MLPA analysis can be used to detect deletions of the entire PHOX2B gene (although the clinical significance of these whole-gene deletions is unclear), or a single or multiple exons (expected to cause CCHS). These variants can be missed with sequencing and targeted analysis for pathogenic variants.

Testing Strategy

To confirm/establish the diagnosis in a proband, the American Thoracic Society Statement on CCHS suggests step-wise PHOX2B testing in persons meeting clinical diagnostic criteria; see Figure 1, Weese-Mayer et al [2010] (full text), Weese-Mayer et al [2012].

Figure 1

(A) Algorithms to determine when and what type of PHOX2B genetic testing should be performed in various clinical scenarios in which CCHS and LO-CCHS are suspected or confirmed (B) Algorithm to determine when and what type of PHOX2B genetic testing should (more...)

1.

Targeted analysis for pathogenic variants (fragment analysis; Screening Test) should be performed to identify the following [Weese-Mayer et al 2003, Berry-Kravis et al 2006, Weese-Mayer et al 2010, Bachetti et al 2011, Jennings et al 2012, Weese-Mayer et al 2012]:

• All the CGN polyalanine repeat expansion mutations (PARMs)
• The 35-bp and 38-bp NPARM recurrent out-of-frame deletions in the coding region involving the polyalanine repeat region, which cause a frameshift and obliteration of the polyalanine repeat sequence
• Low-level mosaicism for both PARMs and NPARM deletions

2.

If no pathogenic variant is identified with the Screening Test, perform sequence analysis of the entire PHOX2B coding region and intron-exon boundaries.

3.

If no pathogenic variant is identified and if clinical suspicion is high, perform deletion/duplication analysis to determine if an exon or whole-gene deletion is present.

Note: This third step of testing became available after the American Thoracic Society Statement of 2010 was published [Weese-Mayer et al 2010].

Predictive testing for at-risk asymptomatic adult family members requires prior identification of the pathogenic variant in the family.

• Parents of a proband who has the 20/24 genotype or the 20/25 genotype (i.e., 20 CGN repeats on one allele and 25 CGN repeats on the other allele) should be tested for the PHOX2B pathogenic variant with the Screening Test to determine if they are at risk for later-onset CCHS (LO-CCHS).
• Parents of a proband with a longer PARM (genotypes 20/26-20/33) should be tested for a PHOX2B pathogenic variant with the Screening Test to determine if they have somatic mosaicism for their child's identified variant.
• Parents of a proband with a 35-bp or a 38-bp deletion (NPARM) should be tested for a PHOX2B pathogenic variant with the Screening Test to determine if they have somatic mosaicism for their child's identified variant.

Note: Germline mosaicism in a parent of a proband is rare and cannot be identified with molecular genetic testing of leukocytes or tissues other than germ cells [Rand et al 2012].

Prenatal diagnosis and preimplantation genetic testing for at-risk pregnancies require prior identification of the pathogenic variant in the family.

Clinical Description

Congenital central hypoventilation syndrome (CCHS) represents the extreme manifestation of autonomic nervous system (ANS) dysregulation (ANSD), with a hallmark of disordered respiratory control [Weese-Mayer et al 2010].

Classic CCHS is characterized by adequate ventilation while the individual is awake and apparent hypoventilation with monotonous respiratory rates and shallow breathing (diminished tidal volume) during sleep. More severely affected individuals with CCHS hypoventilate both when awake and when asleep [Weese-Mayer et al 2010]. Children who hypoventilate both when awake and when asleep typically present in the newborn period, as do the vast majority of children who hypoventilate only when asleep. The salient respiratory and cardiac findings of CCHS are summarized in Table 2.

Autonomic nervous system dysregulation (ANSD) [Marazita et al 2001, Weese-Mayer et al 2001]. As would be expected in consideration of the key role of PHOX2B in development of the autonomic nervous system [Howard et al 2000], children with CCHS have manifestations of ANSD (Table 2). Table 3 highlights neuropathologic and neuroimaging findings of individuals given a clinical diagnosis of CCHS (many of whom unfortunately did not undergo confirmatory PHOX2B molecular genetic testing).

Many successfully ventilated individuals with CCHS are now in their 20s, suggesting the potential for a normal life span. The cause of death in individuals with CCHS is usually related to suboptimal ventilatory support or involvement with substances that could affect judgment or ventilation [Chen et al 2006]. Development of asystoles is another potential cause of sudden death in CCHS [Gronli et al 2008] among individuals with a prolonged R-R interval who have not received a cardiac pacemaker [Antic et al 2006] or in individuals who are not rigorous about monthly cardiac pacemaker assessment (e.g., the battery life is depleted or the pacemaker malfunctions).

Neurocristopathy (i.e., maldevelopment of neural crest-derived structures) including Hirschsprung disease and congenital absence of parasympathetic intrinsic ganglion cells of the distal hindgut are present in 16%-20% of individuals with CCHS. The risk of Hirschsprung disease is highest in children with NPARMs and with the longer PARMs. Hirschsprung disease typically presents in the newborn period, although it has been diagnosed later in infancy and early childhood (see also Genotype-Phenotype Correlations).

Tumors of neural crest origin including neuroblastoma, ganglioneuroma, and ganglioneuroblastoma, are observed overall in 5%-6% of children with CCHS [Trochet et al 2005b, Berry-Kravis et al 2006]. The risk of a neural crest tumor is highest in children with NPARMs (~50% will develop a neuroblastoma), and rare among children with PARMs (low but apparent risk is in those with 20/29-20/33 genotypes). The tumors can present at variable ages: neuroblastoma typically before age two years; ganglioneuromas later as incidental findings. Tumor-related deaths are uncommon (see also Genotype-Phenotype Correlations).

Later-onset CCHS (LO-CCHS) with mutation of PHOX2B is characterized by alveolar hypoventilation during sleep and symptoms of autonomic nervous system dysregulation (ANSD); however, onset is after the first month of life with diagnosis in later infancy, childhood, or adulthood.

LO-CCHS results from reduced penetrance of certain PHOX2B variants: for example, compound heterozygosity for the normal 20 CGN and the abnormal 24 CGN alleles, compound heterozygosity for 20 CGN and 25CGN, and (rarely) a small NPARM or homozygosity for an allele coding for 24 CGN alanine repeats.

LO-CCHS needs to be considered in:

• Individuals who do not necessarily have the characteristic CCHS phenotype, but do have the following:
  • Apparent life-threatening events and cyanosis during sleep
  • Recurrent severe pulmonic infections with related hypoventilation
  • Unexplained seizures
  • Respiratory depression after anti-seizure medication, sedation, or anesthesia
  • Unexplained neurocognitive delay with any history of prior cyanosis
  • Unexplained nocturnal hypercarbia and hypoxemia
  • Unresolved central alveolar hypoventilation after treatment for obstructive sleep apnea
  • Seeming unresponsiveness to conditions of apparent hypercarbia or hypoxemia (prolonged underwater swimming, pneumonia)
• Infants and children who die suddenly and unexpectedly ("SIDS" and "sudden unexplained death of childhood [SUDC]"), especially if there is a family history of CCHS

See ATS Statement for more details.

A growing number of individuals have been reported in the literature with LO-CCHS and a confirmed PHOX2B PARM or NPARM. On occasion only one family member is described, but more typically several family members in multiple generations are described. These studies emphasize the importance of adult care providers obtaining a family history which includes whether the individual has a child with a genetic disorder (e.g., CCHS).

The physician with a heightened clinical suspicion of LO-CCHS who orders prompt molecular genetic testing ofPHOX2B will quickly make the diagnosis and avert potentially life-threatening decompensation as well as the risk for neurocognitive compromise. The clinician should inquire about whether the individual has a history of hypoventilation temporally related to past anesthesia or sedation exposure, delayed "recovery" from a severe respiratory illness, and/or unexplained seizures or neurocognitive impairment.

Genotype-Phenotype Correlations

Respiratory. A correlation between the PHOX2B polyalanine repeat expansion mutation (PARM) length and the severity of the respiratory phenotype and associated symptoms has been observed [Weese-Mayer et al 2003, Matera et al 2004, Berry-Kravis et al 2006].

ANSD. Association between the PARM length and quantitative ANSD traits (i.e., number of ANSD symptoms and number of affected systems as described in Weese-Mayer et al [2001] and Marazita et al [2001]) has been investigated [Weese-Mayer et al 2003].

A significant association was observed between:

• PARM length and number of ANSD symptoms (p=0.02), but not number of ANSD-affected systems (p=0.13);
• PARM length and daily duration of required ventilatory support (p=0.003).

The type of PHOX2B variant and the length of PARMs determine severity of ANSD. Increasing PARM length is associated with increasing frequency of organ system-specific physiologic ANSD.

Hirschsprung disease. Individuals who are heterozygous for 20/27 genotype or longer PARMs are at greatest risk for Hirschsprung disease. Nearly all individuals with NPARMs have Hirschsprung disease [Trochet et al 2005b, Berry-Kravis et al 2006].

Tumors of neural crest origin. Individuals with NPARMs have a greater risk of developing a tumor of neural crest origin than those with PARMs. Likewise, individuals with the longest PARMs are at an increased risk (albeit lower than the risk of those with NPARMs) of developing a tumor of neural crest origin [Trochet et al 2005b, Berry-Kravis et al 2006]. Prevalence of tumors of neural crest origin varies by type of PHOX2B variant with report of individuals with PARM genotypes of 20/29 and 20/33 only and in NPARMs [Amiel et al 2003, Weese-Mayer et al 2003, Trochet et al 2005b, Weese-Mayer et al 2010].

Cardiac arrhythmia. A positive correlation between longest R-R interval and PARM length was identified in individuals with the three most common PHOX2B genotypes: compound heterozygosity for the following number of GCN repeats (20/25; 20/26; 20/27).

Specifically, the risk for prolonged sinus pauses and the need for a cardiac pacemaker are increased in individuals with PARMs of 20/26 and 20/27 as compared to 20/25 [Gronli et al 2008]. Likewise, a positive correlation between number of children for whom a cardiac pacemaker was recommended and PARM length was identified [Gronli et al 2008].

Facial features. The significant negative correlation between PARM length and four anthropometric measures (mandible breadth, nasolabial angle, lateral lip height, and mandible-face width index) decreases as the PARM length increases [Todd et al 2006b].

Dermatoglyphic pattern. No significant association was found between the PARM length and dermatoglyphic patterns [Todd et al 2006a]. However, an increase in arches among girls and an increase in ulnar loops among boys were reported.

Pupillary response to light. See information in Table 2.

Penetrance

Penetrance for the PHOX2B polyalanine repeat expansion appears to be high. Amiel et al [2003], Sasaki et al [2003], Weese-Mayer et al [2003], Matera et al [2004], and Berry-Kravis et al [2006] found no controls with a PHOX2B polyalanine repeat expansion.

However, the recent identification of CCHS in adults and young children (but not infants) with the 20/24 genotype and the 20/25 genotype indicates reduced penetrance in early childhood for this specific genotype. This also appears to be true for a small subset of the NPARMs [Berry-Kravis et al 2006].

Anticipation

Limited data suggest that the polyalanine expansion in PHOX2B is meiotically stable. Many reports have consistently documented a stable number of repeats during parent-to-child transmission, including instances of parental mosaicism for the expansion [Weese-Mayer et al 2003, Trochet et al 2005b, Weese-Mayer et al 2005, Antic et al 2006]. In all instances in which the PHOX2B polyalanine repeat expansion was transmitted from a parent with CCHS to a child with CCHS or from a mosaic parent to a child with CCHS, no change was observed in the number of repeats (i.e., parent and child had mutated alleles of the same size).

Nomenclature

The appropriate nomenclature for this disorder is congenital central hypoventilation syndrome (CCHS).

A literary misnomer, "Ondine's curse," has been used in the past. In the German folk epic [Sugar 1978], the nymph Ondine falls in love with a mortal. When the mortal is unfaithful to Ondine, the king of the nymphs places a curse on the mortal that makes him responsible for remembering to perform all bodily functions, even those that occur automatically such as breathing. When the mortal falls asleep, he "forgets" to breathe and dies. Because Ondine was not the one who cursed the mortal, individuals with CCHS do not forget to breathe, and individuals with CCHS are not "cursed," the term "Ondine's curse" is a misnomer and should be discouraged.

Haddad syndrome refers to the co-occurrence of CCHS and Hirschsprung disease; the term is not widely used.

Prevalence

With the introduction of clinically available molecular genetic testing for PHOX2B in 2003, it has become apparent that CCHS is no longer as rare as previously considered. Current estimates of at least 1,000 individuals worldwide [Weese